Patent Application
20240281478
There are several ways in which social networks can leverage searching for social media content (e.g., picture or video content). In addition to providing users a means for searching and filtering through social media content based on keywords (e.g., content relating to people, events and topics relevant to them), search features can permit a social network to create user views and dashboards, grouping of social media content, curation of media social media content, and extraction of topics from social media content.
Various ones of the appended drawings merely illustrate some embodiments of the present disclosure and should not be considered as limiting its scope. The drawings are not necessarily drawn to scale. To easily identify the discussion of any particular element or act, the most significant digit or digits in a reference number refer to the figure number in which that element is first introduced, and like numerals may describe similar components in different views.
The description that follows includes systems, methods, techniques, instruction sequences, and computing machine program products that embody illustrative embodiments of the disclosure. In the following description, for the purposes of explanation, numerous specific details are set forth in order to provide an understanding of various embodiments of the inventive subject matter. It will be evident, however, to those skilled in the art, that embodiments of the inventive subject matter may be practiced without these specific details. In general, well-known instruction instances, protocols, structures, and techniques are not necessarily shown in detail.
Among other things, embodiments of the present disclosure improve media content (e.g., social media content) search functionality and curation of media content (e.g., social media content). For instance, various embodiments described in this document provide features that can present media content items in the form of dynamic collection of media content items (e.g., stories) upon a user clicking and typing into a search bar. In another instance, various embodiments described herein improve media content search functionality by ranking user facing search features using input signals.
In the example shown in
The network 106 may include, or operate in conjunction with, an ad hoc network, an intranet, an extranet, a virtual private network (VPN), a local area network (LAN), a wireless LAN (WLAN), a wide area network (WAN), a wireless WAN (WWAN), a metropolitan area network (MAN), the Internet, a portion of the Internet, a portion of the Public Switched Telephone Network (PSTN), a plain old telephone service (POTS) network, a cellular telephone network, a wireless network, a Wi-Fi® network, another type of network, or a combination of two or more such networks. For example, a network or a portion of a network may include a wireless or cellular network and the coupling may be a Code Division Multiple Access (CDMA) connection, a Global System for Mobile communications (GSM) connection, or other type of cellular or wireless coupling. In this example, the coupling may implement any of a variety of types of data transfer technology, such as Single Carrier Radio Transmission Technology (1×RTT), Evolution-Data Optimized (EVDO) technology, General Packet Radio Service (GPRS) technology, Enhanced Data rates for GSM Evolution (EDGE) technology, third Generation Partnership Project (3GPP) including 3G, fourth generation wireless (4G) networks, Universal Mobile Telecommunications System (UMTS), High Speed Packet Access (HSPA), Worldwide Interoperability for Microwave Access (WiMAX), Long Term Evolution (LTE) standard, others defined by various standard setting organizations, other long range protocols, or other data transfer technology.
The messaging server system 108 provides server-side functionality via the network 106 to a particular messaging client application 104. While certain functions of the messaging system 100 are described in this document as being performed by either a messaging client application 104 or by the messaging server system 108, it will be appreciated that the location of certain functionality either within the messaging client application 104 or the messaging server system 108 is a design choice. For example, it may be technically preferable to initially deploy certain technology and functionality within the messaging server system 108, but to later migrate this technology and functionality to the messaging client application 104 where a client device 102 has a sufficient processing capacity.
The messaging server system 108 supports various services and operations that are provided to the messaging client application 104. Such operations include transmitting data to, receiving data from, and processing data generated by the messaging client application 104. This data may include, message content, client device information, geolocation information, media annotation and overlays, message content persistence conditions, social network information, and live event information, as examples. Data exchanges within the messaging system 100 are invoked and controlled through functions available via user interfaces (UIs) of the messaging client application 104.
Turning now specifically to the messaging server system 108, an Application Program Interface (API) server 110 is coupled to, and provides a programmatic interface to, an application server 112. The application server 112 is communicatively coupled to a database server 118, which facilitates access to a database 120 in which is stored data associated with messages processed by the application server 112.
Dealing specifically with the Application Program Interface (API) server 110, this server receives and transmits message data (e.g., commands and message payloads) between the client device 102 and the application server 112. Specifically, the Application Program Interface (API) server 110 provides a set of interfaces (e.g., routines and protocols) that can be called or queried by the messaging client application 104 in order to invoke functionality of the application server 112. The Application Program Interface (API) server 110 exposes various functions supported by the application server 112, including account registration, login functionality, the sending of messages, via the application server 112, from a particular messaging client application 104 to another messaging client application 104, the sending of electronic media files (e.g., electronic images or video) from a messaging client application 104 to the messaging server application 114, and for possible access by another messaging client application 104, the setting of a collection of media content items (e.g., story), the retrieval of a list of friends of a user of a client device 102, the retrieval of such collections, the retrieval of messages and content, the adding and deletion of friends to a social graph, the location of friends within a social graph, opening and application event (e.g., relating to the messaging client application 104).
The application server 112 hosts a number of applications and subsystems, including a messaging server application 114, an image processing system 116, a social network system 122, a search system 124, and a ranking system 126. The messaging server application 114 implements a number of message processing technologies and functions, particularly related to the aggregation and other processing of content (e.g., textual and multimedia content including images and video clips) included in messages received from multiple instances of the messaging client application 104. As will be described in further detail, the text and media content from multiple sources may be aggregated into collections of media content items (e.g., called stories or galleries). These collections are then made available, by the messaging server application 114, to the messaging client application 104. Other processor and memory intensive processing of data may also be performed server-side by the messaging server application 114, in view of the hardware requirements for such processing.
The application server 112 also includes an image processing system 116 that is dedicated to performing various image processing operations, typically with respect to electronic images or video received within the payload of a message at the messaging server application 114.
The social network system 122 supports various social networking functions services, and makes these functions and services available to the messaging server application 114. To this end, the social network system 122 maintains and accesses an entity graph 304 within the database 120. Examples of functions and services supported by the social network system 122 include the identification of other users of the messaging system 100 with which a particular user has relationships or is “following”, and also the identification of other entities and interests of a particular user.
The application server 112 is communicatively coupled to a database server 118, which facilitates access to a database 120 in which is stored data associated with messages processed by the messaging server application 114.
Some embodiments may include one or more wearable devices, such as a pendant with an integrated camera that is integrated with, in communication with, or coupled to, a client device 102. Any desired wearable device may be used in conjunction with the embodiments of the present disclosure, such as a watch, eyeglasses, goggles, a headset, a wristband, earbuds, clothing (such as a hat or jacket with integrated electronics), a clip-on electronic device, or any other wearable devices.
The search system 124 performs search operations on social media content as described and illustrated in this document.
According to various embodiments, a search feature presents the social media search results as cards that appear in a view of a graphical user interface (“user interface”). In particular, the search feature can comprise a pre-typing view, post-typing view, or both. The pre-typing view appears on a user interface in the absence of user parameters for executing a social media search (e.g., a search box displayed on the user interface is left blank), while a post-typing view appears on the user interface in the presence of a user parameter for executing a social media search (e.g., appears as soon as the user begins to enter characters of a search query into the search box). The user interface may be configured such that when the post-typing view appears, the pre-typing view is replaced by the post-typing view, and when the pre-typing view appears, the post-typing view is replaced by the post-typing view. Social media search results presented in the user interface may be consistent across different types of searches (e.g., camera, stories, chat, live stories, events, etc.).
The pre-typing view can comprise such views as: special occasion (birthdays, breaking news, etc.) cards view; an emoji carousel view; an enhanced quick chat carousel view; a topics carousel view; a live stories carousel view; an around me carousel view; media content item (e.g., a digital photo, image, video, etc.) tabs view that can include, for example, highlights, shows, sports, fashion, pets, attractions, and neighborhoods; a discover carousel view; user tabs view; or a combination thereof. The pre-typing goal can, for example, facilitate quick access to best friends on the social network, and access to the best social media relating to (e.g., media content items from) popular topics, events, and places.
As used herein, a snap may comprise a user-captured or user-enhanced (e.g., text, visual effects, etc.) digital photo or video. As also used herein, a story or a gallery is a collection of media content items.
The post-typing view can comprise such views as: a search-as-you-type view; a hero card best matching query view; a view showing the result set based on disambiguation of a query; a view for more stories for hero card; a related searches view; a related people view; or some combination thereof. The pre-typing goal can, for example: bucket the best social media (e.g., media content items) into the Hero card; surface Hero cards as quickly as possible while typing; focus on good experiences within core verticals (specific places, people, topics, broad places, events, and categories); and provide a path to more results/similar searches if available.
The ranking system 126 performs search operations on social media content as described and illustrated in this document.
According to various embodiments, ranking social media content as described in this document can be utilized with: dynamic story generation; emoji carousel generation; topics generation and ranking; live stories ranking; post-type search results; around me carousel generation; media content item tabs table generation; more stories section generation; related search section generation; query suggestion; breaking news section generation; people searching; related people section generation, and spam and abuse moderation.
Inputs that might be utilized by the ranking include, without limitation: bucketed friend count; bucketed follower count; number of media content items (e.g., snaps) sent over a period of time (e.g., last 28 days); number of media content items received over a period of time (e.g., last 28 days); number of stories posted over a period of time (e.g., last 28 days); total number of views for stories posted over a period of time (e.g., last 28 days); total time of views for stories posted over a period of time (e.g., last 28 days); bucketed number of days since account creation; location of the media content item; time of the media content item; caption of the media content item; whether a media content item is a taken using the front-facing camera; and visual labels associated with the media content item; story telling score of the media content item; creative quality score of a media content item; and media quality signals.
The visual labels associated with a media content item comprises a list of labels/annotations describing the objects in the media content item (e.g., both images and videos) using the trained visual annotation model. The story telling score of the media content item comprises a score (e.g., between 0, 1) that is assigned to each media content item. The score may be a function of: number of media content items from other users posted around the time and location of the media content item; number of media content items from the same user posted around the time and location of the media content item; the descriptiveness of the media content item (e.g., whether it is a video and whether it has caption); and the visual representativeness of media content item given the media content items coming from the same location. The creative quality score of a media content item comprises a score (e.g., between 0 and 1) that is a function of the normalized view time per view on a user's “Recent Updates” media content items combined with the total number of views on the user's “Recent Updates” media content items. All “Recent Updates” stats may be computed based on a period of time (e.g., last 28 days of data). The media quality signals may comprise image media content items and video media content items. For image media content item, the media quality signals are the normalized mean of a media content item's gray scaled pixel values, the entropy of a media content item's gray scaled pixels, and the size of the media content item (in bytes). For video media content items, the media quality signals are the percentage of traceable key point features across video frames, the percentage of differences in pixel values in consecutive frames, and the percentage of spikes of differences in pixel values in consecutive frames.
For some embodiments, dynamic stories are algorithmically-generated stories built from media content items submitted to in real-time. This may be accomplished over multiple stages, as illustrated by
For instance, the logic for the quick chat section can rank results by: best friends; recent conversations (e.g., media content item or chat interactions, exclude new friends with no interactions); and bi-directional friendships. In the quick chat section, a user can tap on a card to enter a chat, double-tap on a card to media content item, or press and hold a card for a mini profile for the user represented by the card.
The logic for the new friends section can list results by most recent new friends added by the user (e.g., newest friend at top). This section can include my most recent two-way friends that the user has made within the last 30, and if there are fewer than 5 friends made within the last 30 days, include up to 5 of the most recent friends made from beyond 30 days.
The logic for the quick add section can list people that the user may likely want to quickly add as friends, and may list the people according to a rank (e.g., most likely to least likely).
The logic for the contacts section can list other users that are correspond to contacts in a user's address book (e.g., an address book local to the user device but external to the social network client on the user device). Below the other users in the address book, the user may see people in the user's address book who they can invite to the social network platform. If a user does not have any people in his or her address book, he or she may not see the contacts section.
The set of tabbed subsections presented to the user may be static, dynamic, or a combination of both. For instance, the tabbed subsections may be dynamic such that a user does not always see the same ordering or default tabbed subsections when they horizontally scroll through the tabs. The ordering may change depending on context, so for example if it is the weekend, a “Parties” tabbed subsection may be presented and include social media content relating to upcoming, on-going, or recent parties. The neighborhood tab may be presented if a user is in a metro area that is large enough to have neighborhoods (e.g., Los Angeles or New York). The attractions tab may be presented if a user is in a region that is large enough to have local attractions.
Starting from the left screenshot in
For various embodiments, a successful search refers to a situation where the search results meet a confidence threshold that permits a post-typing view to present a vertical (a vertical section presenting search details) that in addition to just a results section (e.g., hero card section), includes/is fully expanded to show multiple sections that present social media content related to the social media content provided in the results section (e.g., a more stories section and a related search section that directly tie to the results section). Example verticals that can appear in a post-typing view are described in this document with respect to
For some embodiments, if a user taps on the hero card during a weekend (e.g., night), the content would be fresh and show what it is like at the specific location (e.g., at “The Bungalow”) right now. Alternatively, if the user taps on the hero card on a Monday, the hero card will show the user the best social media content relating to the specific location from the weekend. In some instances, the events vertical can list social media content from an entire duration of an event, while if a specific location (e.g., bar or club) has an amount of recent social media content (e.g., social media content from the last few hours) that exceeds a certain threshold, the user may be presented with social media content exclusively from a recent period (e.g., only the last few hours).
As shown in
As shown in
As shown in
As shown in
As shown in
As shown in
For
Screenshot 2602 illustrates where there are no results for exact query and only one result is presented. Screenshot 2604 illustrates where there are no results for exact query and multiple results are presented. Screenshot 2606 illustrates where there are no results for exact query and there is confidence that the search query contains a typo. Screenshot 2608 illustrates where a user is searching for a concept (“ink”) and result contain the concept sought and a specific place also having the same name as the concept. Screenshot 2610 illustrates where a user is searching for a specific place or broad location (“nopa”) and the result are provided. Screenshot 2612 illustrates where a user is searching for a specific topic, person, band or event (“spoon”) and the mix results are provided. Screenshot 2614 illustrates where a user is searching for a specific place (“eggslut”) and the result indicate there are multiple locations near the user. Screenshot 2616 illustrates where a user is searching for a topic story (“mcdonalds”) and the results provide a topic story and multiple specific locations.
In
Flow of the Dynamic Story Creation stage obtains a set of Joins from Joins Bigtable. Flow comprises a set of data flow pipelines each of which takes in Joins from over a period of time (e.g., past 24 hours). Based on the Joins and side data models, Flow generates dynamic stories. For some embodiments, Flow runs clustering algorithms (Clusterers Service) over different dimensions, such as geolocation, time, similar caption terms, and the like. The clustering results from Clusterers Service are stored in a database, which is represented in
For some embodiments, Clusterers Service comprises a service that exposes multiple remote procedure calls (e.g., gRPC) for different clustering tasks. As shown, Clusterers Service is implemented as a service running in GKE. Clusterers Service may be implemented as a Python service. Additionally, Clusterers Service may operate based on stored models (e.g., GCS), utilize API calls to Place Server to compute place names for generated stories, and datastore calls for curation data related to clusters and stories.
For some embodiments, ElasticSearch comprises an index for serving stories, and may store one index per a Flow type (e.g., Topic, Event, and Places Flow as described in this document). For a given Flow, the index may be swapped at each run, resulting in all dynamic stories generated on previous runs of the given Flow being removed and replaced with dynamic stories generated from the latest run of the given Flow. For some streaming and incremental Flows (such as Places Flow), the index may not be swapped and a single long-running index is maintained where dynamic stories are added and removed over time.
Various embodiments may maintain a separate ElasticSearch cluster for indexing where the separate ElasticSearch cluster stores Cluster outputs (e.g., rather than dynamic stories) produced by Flows. Such a separate ElasticSearch cluster may be utilized to store a period of Joins (e.g., past 24 hours) to provide search and analysis functionality for indexing pipelines, curation workflows, debug tasks, or some combination thereof.
For some examples, the architecture of ElasticSearch comprises a cluster deployed as a GKE cluster with dedicated node pools for data nodes, client nodes, and master nodes. Client nodes may expose a NodePort Service for the HTTP endpoint and the ElasticSearch Transport endpoint. To dynamically discover the virtual machines that participate in the NodePort Service, GOOGLE Internal LoadBalancer may be utilized, which provides an internal frontend IP address and port forwards to the IP addresses of the cluster.
As also shown, Annotator comprises Media Processor (MP), which runs multiple machine intelligence data flow graph models (e.g., TensorFlow™ models) on media content item content to compute labels (e.g. stadium, concert, basketball) and metrics (e.g. media quality, shakiness score). The data flow models may be trained offline and loaded at startup time. SMP may also compute the best thumbnail frame for a media content item including video media. SMP may be implemented as a C++ service operating on GKE. SMP exposes a remote procedure call (RPC) interface, which may only be available to internal services of the indexing infrastructure.
A user features of Annotator can run as a plugin inside of Joiner. This user features annotator reads user statistics from a datastore and denormalizes them into a Join. The statistics are computed by a separate pipeline by reading data from FSN and SC-Analytics and storing them on a datastore in the indexing infrastructure.
As noted in this document, Flow supports creation of dynamic stories from a set of Joins and implemented as a dataflow pipeline. For some embodiments, Flow scans a given time window of Join updates (e.g., 1 day, 30 minutes, etc.). Flow bins the Joins based on a set of selection criteria (e.g., media content items for the same geo region). Flow performs an external remote procedure call (RPC) request Clusterer service, which returns clusters computed (by the clustering algorithm) from binned Joins. Flow then writes the output (clustering results) to Cluster Bigtable and to an index of a search service, ElasticSearch.
The output of the Clusterers Service is a set of Cluster protos, which can contain all information about the clusters that were created by Topics Clusterer. The set of Cluster protos are stored on Cluster Bigtable. For serving Topic dynamic stories to search clients in ElasticSearch, the set of Cluster may be transformed into a set of dynamic story (DynamicStory) protos, which contain the necessary fields for serving stories of the set of Cluster. This set of DynamicStory protos may be written to the index of the ElasticSearch.
Flows to create other types of stories (other than topic stories) may be implemented similar to Topics Flow. For example, Events Flow creates event stories based on geographical and time proximity. Events Flow may have a flow similar to Topics Flow with the modification of binning Joins into regions and performing single clustering calls per a region.
In another example, MetaStories Flow creates stories from other stories. MetaStories Flow may have a flow similar to Topics Flow with the modification of that the input is no longer Joins, but rather Clusters, and Clusterer Service is called for metastories.
PlaceStories Flow receives (via GOOGLE Cloud PubSub) a message which triggers the dynamic story creation for a given Place. First the Place Bigtable is queried to retrieve all Join Ids that have been assigned to that Place, and then the actual Joins are retrieved from the Joins Bigtable. Clusterer Service is called to form Clusters for the Place and results are stored as described with respect to Topic Flow.
For some embodiments, a Suggest Request is triggered every X characters or second, and a client (e.g., mobile client) makes a request to the Query Suggest endpoint in the FSN. The FSN may perform client authentication and passes the request through to Suggest Service by issuing a URL fetch request. Suggest Service does an in-memory search using the user query against an APACHE Lucene™ index created from suggestions generated by a Flow pipeline (e.g., Topics Flow). The top set of query suggestions are returned to the FSN and passed through to the client to show the user. Suggest Service may operate as a separate service with its own in-memory APACHE Lucene™ index to avoid latency sensitive issues. Also, the Lucene™ index may be regenerated periodically by reading the latest output from a Flow pipelines.
With respect to the search stack, a user queries from a client (e.g., mobile device client) are sent to the search stack (via the FSN as a Search Request). The Search Request is passed through to Ranking Service, which may be a Python HTTP server running on GAE Flex. Ranking Service may run multiple producers based on the user's query and location to create the different sections for the user. Each product (e.g., EmojiCarrousel) issues queries against the indices of ElasticSearch for the dynamic stories that were generated by the indexing pipeline (e.g., Flow). Dynamic stories returned by the ElasticSearch are ranked, deduped across producers, and then combined into a final Search Response and sent back to the client (via the FSN). The FSN takes the Search Response from Ranking Service and passes it through to the client.
For various embodiments, search features described in this document utilize session logging. For instance, for a given query event, the following fields could be logged: language preference, which provides a comma-separated list of a user's device-level language preferences; query_is_suggestion, which indicates whether a query comes from a search suggestion; and source, which identifies the page source of a search session. In another instance, for result sections that return dynamic stories, events are be logged by a search backend. In yet another instance, a fields keep track user actions with respect to dynamic stories (e.g., skip to next, go to previous, exit story view, etc.), and track physical gestures of user action (e.g., swipe, tap, tap & hold, etc.) when viewing a dynamic story.
As used in this document, dynamic story can comprise a grouping of story media content items based on a topic (e.g., shares an event or a place or other theme). A curator can review a dynamic story (e.g., fetched from Ranking Service) and decide whether to add/remove/make a change to the story. Curated Data comprises corrections and other changes produced by a curator after she/he reviews a dynamic story within the curation tool. A context card comprises a small card user interface that shows a name, an image, and a description.
For some embodiments, the curation tool comprises a web frontend component, a data storage system, a media access component, and a data enrichment system. The web frontend provides a curator (e.g., via a web browser) with a graphical user interfaces (e.g.,
Components of curation data can include, without limitation: caption or title of a dynamic story; place name associated with the story; entities associated with the story (e.g., performer for a concert story, or the politician for a rally story); information about the associated entities (e.g., name, user identifier, official website link, associated Wikipedia page link; link to a picture of the entity (e.g., posted on Wikipedia page). Components of curated data can also include, without limitation: Boolean indicating whether a story contains media content items from multiple locations/events (e.g., a mixed story); Boolean indicating whether a story has captions predominantly from a non-English language; Boolean indicating whether a story being offensive; story rating for how interesting the story is (e.g., discrete values such as {awful, bad, ok, good, love it}); emoji character(s) associated with the story (e.g., curator choose which emoji is best associated with a story); cover media content item id (e.g., curators can change the cover media content item if they feel the cover media content item could be better); removed media content item ids (for media content items removed from the story because, for instance, they are low quality {filled with unappealing media content items that do not form a clear story}, offensive or blurry etc.); highlight media content item ids (e.g., up to 10 media content item ids that represent the highlights of the story); and keywords associated with a story (e.g., curator provide a few keywords that are relevant to the story-based on media content items and not restricted to captions); Boolean indicating relevance of sub stories to main story; newsworthiness (e.g., {not news, local, city, state, country, global}); and related news URLs.
To facilitate certain features, the curation tool utilizes APIs (e.g., GOOGLE APIs), for obtaining location information, knowledge graphs, and performing web searches. Such APIs can enable curation tool to present, for example, locations, related entities, and news related to entities in a dynamic story. The curation tool can also utilize data extracted from public sources to find attributes of an entity (e.g., like stadium capacity), which might be useful to present in a context card.
For various embodiments, curated data generated by the curation tool is maintained separate from user data, no media is stored with the curated data, and no curated is copied or stored by the curation tool. Through the curation tool, curated data can be used to generate context cards, apply changes provided by curators to other similar dynamic stories, and collect store changes to dynamic stores from curators. In this way, curators can implement corrections and add enrichment (e.g., add title, place name, selection of representative media content items, links to news, etc.) to a dynamic story. Curation may be scheduled to take place after media content items are selected algorithmically to create stories.
With respect to a dynamic story, information of interest for curation may include, without limitation, an associated place, associated people, associated organizations, associated events, and the like. Dynamic story metadata can include, without limitation: caption or title correction, where captions may be indirectly derived from user captions on media content items; dynamic story category (e.g., concerts/games/fashion/food/politics); place name correction, where place name may be derived from GPS info on user media content items and Google Places API; place bounding box correction (e.g., outline the boundary of the place on the map); entity metadata, which is information about a specific entity (e.g., Kendrick Lamar) and may be from public sources (e.g., like Wikipedia); place metadata, which is information about a specific entity (e.g., Staples Center) and may be from public sources public sources; entity-story association, which can tie an entity (e.g., Kendrick Lamar) to a specific dynamic story (e.g., he/she is performing there, or is in the audience, or is a producer of the show etc.); curator information (e.g., name); and time curation was submitted.
With respect to some embodiments, curating dynamic stories by the curation tool involves: the title of the story; place name; bounding box for the place; whether it is a mixed story; whether the story is in non-English; whether the story contains offensive material; whether the story is low quality; emoji characters associated with the story; ratings for how interesting the story is; or context cards for places/people/organizations/events related to the story.
For various embodiments, the curation tool is utilized to curate concept stories (“chatter” stories) around a concept or topic rather than an event that happens at a specific time or place. For instance, a concept story for “Snow” could show media content items from all over the United States on dramatic/interesting snow maps. Curation with the tool may involve: the title of the story; whether the story is in non-English; whether the story contains offensive material; emoji characters associated with the story; ratings for how interesting the story is; remove low quality media content items; pick cover media content item; pick highlight media content items; keywords for the story; or context cards for places/people/organizations/events related to the story.
For some embodiments, the curation tool is utilized to curate concept topicality, whereby a curator is shown media content items for “sub stories” of a main story and asked media content items are relevant to the topic of the main story.
Additionally, for some embodiments, the curation tool is utilized to curate breaking news, whereby curation involves: the headline of the story; place name; bounding box for the place; whether it is a mixed story; whether the story is in non-English; whether the story contains offensive material; whether the story is low quality; newsworthiness; keywords for the story; context cards for places/people/organizations/events related to the story; or related news URLs.
In
For example, geo-time clusters are created. Geo-time clusters can capture collections of media content items (e.g., stories) that are mostly around an event and are cohesive in time and location. A range of similarity signals on top of time/location may be taken into account to assure media content items that are clustered tell the same collection of media content items (e.g., story). These similarity signals include, but are not limited to: visual similarity; caption (salient term) similarity; topic similarity (e.g., two (sets of) media content items that are fashion media content items and are close by are more likely to be clustered together); place similarity (e.g., similarity of place entity assigned to two (sets of) media content items is considered a strong signal for considering clustering them together); and On Demand GeoFilter (ODG) similarity.
In another example, topic (“chatter”) clusters are created. Topic clusters can be cohesive around a theme or topic but are not necessarily tied to a specific time and location. In some instances, topic clustering is accomplished in two steps. In the first step, single topic clustering is performed by clustering all media content items (e.g., in the past 24 hours) containing a topic for each distinct topic (e.g., currently identified as a distinct significant caption term or visual label but expandable to any rule expressed in terms of features of the media content items). A score profile is created during the first step. Components of the score profile can include, without limitation, the independence of topic (reflecting how sufficient the topic is to describe the majority of media content items); the popularity of the topic in terms of the number of posted media content items that are relevant to the topic; the freshness of the topic measured using the age distribution of media content items about the topic; the novelty of the topic measured by comparing the volume of fresh media content items about this topic compared to the historical volume for this topic; the globalness of the topic that reflects how geographically spread the media content items relevant to the topic are; and the quality and descriptiveness of the collection of media content items (e.g., story) for the topic that reflects the quality (e.g., including media quality, user/creator quality, etc.) of media content items shown for that topic and how well they tell the collection of media content items (e.g., story) for that topic (for e.g. video media content items are generally considered more descriptive than image media content items). Based on all the score components, an overall score is generated to represent how interesting the topic cluster is. The score components along with the overall score are used in merging similar topics during the second step and also for ranking topic collections of media content items (e.g., stories) in user interface presented to users (e.g., the topics carousel).
In the second step, similar topics are merged. Depending on what topics are trending, the trending topic collections of media content items (e.g., stories) can be correlated and in some cases redundant. Hence, in the second stage of topic cluster generation we merge similar topics together. Given the cost of topic clustering/merging, for some embodiments, full merging is only performed for the top topic clusters (based on the score computed in the first step) and join the other topic clusters to the merged top clusters. The merging of two topic collections of media content items (e.g., stories) can be done based on a variety of similarity criteria including, without limitation: synonym match between two topics; ngram similarity between terms associated with two topics; context similarity between based on the aggregate salient terms of media content items associated with two topics; context similarity based on similarity between users posting media content items about the two topics; and language mismatch is used as a negative similarity signal.
At block 3804, initial dynamic collections of media content items (e.g., stories) are created from the clusters. This may be accomplished by finding collections of media content items (e.g., stories) within the cluster and inferring attributes of the collections including place name, caption, topic, highlight, repeated-ness, and the like. Given clusters of media content items (from block 3802), during block 3804 the method 3800 finds the collections of media content items in each cluster and determines what each collection of media content items (e.g., story) is about. For various embodiments, block 3804 involves: identifying sub-collections (e.g., sub-stories) within the cluster (e.g., based on time, geo, visual, and cohesiveness); breaking the over-clustered events into multiple collections of media content item (e.g., stories); finding a caption for each collection (e.g., story); find a place name for each collection; finding a category of the collection (e.g., concert, game, and fashion); creating a highlight for each dynamic collection; and determining whether a collection is a repeated happening. Creating highlight for each dynamic collection involves: finding the highlight at the order viewable by the user; and finding a hero (e.g., cover media content item) for the highlight. The highlight is selected based on scoring, which may have the following factors/criteria considered: cohesive and representative of the collection; has high collection telling score; has high creative quality score; generally preferred to be a video media content item rather than an image (e.g., with the exception of the collection being image seeking); and does not contain low quality media quality relative to all the candidates for highlight selection. Media quality can be calculated from the input signals described in this document and may be generally aimed to remove black/dark, very bright with no visual content and shaky media content items. Find a hero for the highlight may consider the following factors/criteria: has a high score in highlight scoring; representative in visual features relative to what the collection is about; not a selfie media content item; and does not have a caption on the media content item.
At block 3806, initial dynamic collections are post-processed and reconciled, which generated final dynamic collections. For some embodiments, post-processing comprises: merging collections with similar caption or place; de-duping collections that are too similar (e.g., shared media content item volume, or caption/place similarity); filtering bad clusters (e.g., based on no place name, no collections, and topic based filtering); detecting ancestors (e.g., by generating stable ids to establish continuity between runs [deduping with the past] and keeping a running history of the cluster); and aggregating stats, such as generated cluster sources and sizes (to track pipeline health) and top salient terms and concepts.
With respect to some embodiments, ranking emojis in the listing of emoji pills is accomplished by two components: an offline (e.g., at story building time) computation of a list of relevant emojis for each story; and an online (e.g., when producing the pre-type search page), generation of an emoji listing (e.g., carousel) based on the most recent set stories (e.g., K stories). The offline computing of the list of relevant emojis may comprising using a set of signal including, without limitation: story captions (e.g., text captions from the media content items in the story) and the emojis the users typed in their captions; top salient story terms; story category; and derived aggregate data, such as term IDF tables, an emoji to terms model, and a term to emoji model. With respect to the online generation of an emoji listing, for each story in the most recent set of stories, the following can be retrieved: story id; display caption; top story salient terms; geographic location (e.g., lat/long); raw media content item count; emoji annotation; and story timestamp (e.g., 95-percentile story timestamp). Emojis are extracted from all the stories retrieved, de-duped (for the similar ones), and ranked based on freshness and volume of the underlying stories. For each emoji extracted, the caption of the best ranked story (e.g., by volume and freshness) that has that emoji is assigned as a search query associated with the emoji. In some instances, the top set of emojis (e.g., top 15) are presented to a user in an emoji carousel.
According to various embodiments, dynamic stories are generated for specific places and large locations, merged from multiple data sources in real time for media content item submitted live, and indexed to support place and location various sections described in this document: pre-typing views, such as current place, around me carousel; and post-typing cards, such as for queries of specific places or larger areas like neighborhood, city, universities, and the like. A backend of building dynamic stories for places may be shared across all the views, and a ranking module (e.g., light weighted) may be applied to each view individually. Core components of the backend comprises, for example: media content item-to-place assignment; place story generation; location story generation; and online ranking. Media content item-to-place assignment assigns each media content item (e.g., that is summited to live) to the best matched place using a probabilistic model of a mixed of signals in a real-time data pipeline. Place story generation regenerates a dynamic story using latest media content items for each place assigned to the media content items. Location story generation mixes dynamic stories from local places, current events and trending topics, and may do so for larger areas where media content items are refreshed frequently. For both pre-typing/post-typing views, a ranking module is applied to all the relevant stories in the context (e.g., plus additional place data, from Google API, etc.) to select the best card to show in real-time.
With respect to media content item-to-place assignment, each media content item submitted to live may be taken together with other media content items within a fixed time window, and may be assigned to the best matched place using a probabilistic model. The candidate places can be from multiple data sources, including: GOOGLE Place Response passed by mobile clients using GOOGLE's Place Detection API; GOOGLE Place Response from querying GOOGLE Geocoding/Reverse Geocoding APIs and GOOGLE Place API, in the backend; a geo-indexed Open Street Map data; existing geo-filter data; and geo-fences data through third party partnerships. For some embodiments, the assignment is accomplished through a probabilistic approach, which utilizes a mixed of cues or signals. Example signals can include, without limitation; geo signals; temporal signals, text signals, visual signals, and quality signals.
Geo signals can comprise associated GPS data with each media content item, and the known lat/long of each place candidate. The accuracy and quality of GPS data may be similar to the error range to many small places in interest. The distances of a list of candidates within a relative larger range may need to be taken into consideration in a statistical fashion. Temporal signals can comprise the opening hour, and historical temporal activity pattern, of a place, which can be used to match against the media content item based on its location timestamp. Text signals can comprise place name and other text properties (like description, salient term profile, etc.) which can be used to match the text properties of the media content item, including caption, media content item labels learned by deep learning models from image content. Visual signals can comprise simplified visual models, such as indoor/outdoor/vehicle, which can help to match certain place types which has obvious indoor property. An alternative approach can use all previous annotated media content items to learn visual features of the place, and use it to compute the distance from the visual features of the specific place. Quality signals can comprise the quality of the place like ratings, previous activity, daily volume, and the like, which can indicate the media content item has a higher chance of being taken there. A prior of probability on each place may be pre-computed offline using foot traffic data (e.g., user activity geo-temporal map).
With respect to place story generation, a dynamic story of the place is generated using the media content items that have been assigned to it. For some embodiments, dynamic stories are constantly regenerated for places when fresh media content items are submitted and assigned to them. The ranking algorithm that selects media content items for the place dynamic story may consider: freshness of the media content items (e.g., for specific places like bars, the story will give users a sense what is happening right now at the place); events at the place (e.g., latest events (with lots of media content items at similar time) will be included in the dynamic story); quality of the media content items (e.g., a flexible quality measure, that combines user score, media score and other factors, may be used to filter low-quality media content items); and cohesiveness of the story (e.g., the algorithm tries to balance between the vibe and theme of the place). Additionally, a place profile (e.g., including salient terms, visual features, etc.) can be used to select representative media content items based on historical media content items.
With respect to location story generation, a dynamic story for a larger location, like city, neighborhood, university, etc. is generated by combining sub-stories from local specific places, live events, and topical events. For each dynamic story, a place is annotated using its geo bounding box. The location story data pipeline may constantly aggregate all the place stories and latest live events and topical stories according to the neighborhood and city info of the place. Sub-stories may be selected to show the vibe and live atmosphere in the location, by a ranking module that takes into account various metrics of each story: volume, freshness, story quality, etc.
With respect to online ranking, for each view supported by place stories, a ranking module is applied online based on retrieved candidate stories. For instance, for current place, the nearest place story (a list of nearest place stories) is be retrieved using the user's current location, which will be compared to GOOGLE Current Place API to determine which card type to show. For around me carousel, a batch of candidate nearby place stories, and event stories are ranked based on their scores, including freshness, quality, volume, and the like. For post-typing queries, the place/location stories are queried when a user types the keyword. The candidate stories can be scored to determine if the location/place card, which contains the place/location story, should show up.
As noted in this document, a post-type result is generated while the user is typing in the search box. When the user starts typing, a list of stories that are topical for queries starting or very similar to what the user has typed begin to show up. The results may be rendered in different forms including, for example: a story list; a hero card with see more button; or a stand-out hero card. What queries are shown can depend both on the similarity of query to user input and also significance/quality of the story topical for the query. In some instances, a hero card appears when there is no ambiguity about the intent of the user, and the shown result is a single story that includes the highlights of the most relevant stories around the query in a ranked order. The ranking of story highlights within the hero card is accomplished based on a variety of signals including, without limitation: centrality and topicality of the query for the story; proximity; freshness; popularity (e.g., reflected by the number of media content items in the corresponding story); and quality including the media quality, diversity of media content items and descriptiveness of media content items.
The types of queries and types of more stories presented in the section may be based on the hero type used to present the results. For example, more stories presents live events query stories for the neighborhood/city, other nearby live events, earlier events in the same venue. More stories section presents concept query stories with similar feel like. So, for instance, a hero card presenting ski would have a more stories section presenting snowboarding, a hero card presenting winter would have a more stories section presenting snow fall, and a hero card presenting Christmas would have a more stories section presenting gift unwrapping. If the concept has local significance, other locally significant stories may be presented by more stories, such as a hero card presenting Santa Con would have a more stories section presenting Santa Con SF. More stories section presents place query stories to show stories from same place type (e.g., bars/restaurants/clubs), which can include stories about the neighborhood and city as well as story about nearby attractions. More stories section presents location query stories about popular places and attractions in the location as well as stories about sister locations (e.g., Venice presents Santa Monica Pier).
A related search section involves queries which can be viewed as pivots for exploration around a different dimension from the current search query. The related search section queries may be complementary to the queries being performed for the more stories section. While a more stories section provides peripheral information related to the user query in the same context in the form of a story, the related search section queries can provide user a way to pivot into a new context in the form of new query. According to some embodiments, the signals and inputs used for creating the related searches section are similar to that of more stories.
Query suggest provides suggestion as user type in queries, based on both user context and our index of stories. Through autocomplete, user efforts to type a whole search query can be reduced, and the user can also be guided to search the best content. The index for the query suggest is based on, for example: information from stories generated from other pipelines that have high probability to become a query, such as location, captions etc.; and top queries from logs. On the serving side, the context of client is used to provide the best guess of a user intended search query. The context includes, for example: the partial query user is currently typing; the Location, and the language.
Breaking news section provides a set of dynamic stories intended to capture newsworthy events as they happen and, as such, it may not appear in every pre-type search. Breaking news may only display when there is a sufficiently important story currently ongoing or in the recent past. Breaking news section is created through a pipeline similar to the one that creates dynamic stories for the live story carousel. Like for live story, the pipeline runs the exact same logic on a smaller set of input media content items. Unlike live story, the pipeline for breaking news may run on smaller window of time (e.g., last 30 minutes of media content items).
For breaking news section, media content items from submit-to-live come as input to the story generation pipeline, and these media content items have captions, location, etc. From these media content items, stories are generated, which are groupings of media content items associated with a location, a descriptive caption, a cover media content item, or a selection of media content items to show to the user. For ranking, the stories are assigned a ‘newsiness score,’ which captures both the relevance and significance of the story. By the newsiness score, stories can be distinguished between where users appear to be discussing a newsworthy topic, as well as those that have sufficient media content item volume and diversity of contributing users. According to some embodiments, in order to show a story in the breaking news section, its newsiness score must exceed a threshold. Among those stories whose scores exceed the threshold, the ordering is decided based on the newsiness score, the story freshness, and the proximity of the story to the user. From the client perspective, the breaking news sections presents stories in similar many as other dynamic stories in the client. As such, stories in the breaking news section can have a caption algorithmically generated from the user captions, a cover media content item, and a list of media content items to display when the story is selected.
People search and related people sections provide stories from popular users. A popular user may be defined as an account with public story privacy and at least five thousand followers. Popular users may have names, categories, and associated locations curated. According to some embodiments, a search index is created for each popular user/account that contains the following: username; display name; any names for popular users curated; categories the popular user is associated with (e.g., NBA, hip hop, fashion, yoga) and which may be curated; associated locations curated; and official story indicator (e.g., emoji) if applicable. A popular user may appear in the search results for a match on: any of the above names with no minimum prefix length; any of the above categories; or on any other popular user that is contained in their same category.
At block 4504, the particular media content item is associated with a cluster based on the set of annotations generated for the particular media content item at block 4502. For some embodiments, the cluster groups together certain media content items, within the plurality of media content items, based on a relationship between at least one annotation of each of the certain media content items. The relationship may relate to, for example, geographical proximity, time proximity, similarity of caption terms, similarity of a place, similarity of a visual feature, similarity of an event, similarity of a metric, or similarity of a user statistic. For some embodiments, the relationship between the annotations of two or more media content items reflects the cohesiveness of those media content items with respect to one or more dimensions (e.g., time, space, visual features, topics, etc.). At block 4506, a collection of media content items (e.g., story), associated with the cluster, is determined (e.g., identified) from the plurality of media content items. At block 4508, a set of attributes is determined, for the collection of media content items, based on at least one annotation of a media content item included in the collection of media content items (determined at block 4506). At block 4510, the collection of media content items is stored with the set of attributes determined at block 4508.
Determining the collection of media content items from the plurality of media content items may comprise determining a plurality of collections of media content items. According to some embodiments, the plurality of collections is reconciled by merging at least two collections of media content items, in the plurality of collections, based on a similarity between at least one attribute of each of the at least two collections of media content items. For some embodiments, the plurality of collections is reconciled by removing a particular collection of media content items, from the plurality of collections, based on a similarity between at least one attribute of the particular collection of media content items and another collection of media content items in the plurality of collections. For some embodiments, the plurality of collections is filtered to produce a filtered plurality of collections of media content items that includes media content items having at least a minimum set of attributes.
For various embodiments, determining the collection of media content items from the plurality of media content items comprises selecting a highlight media content item, from the collection of media content items, to represent the collection of media content items based on a set of scores associated with the highlight media content item.
With respect to searching, the method may further comprise: indexing the plurality of collections of media content items; searching the plurality of collections of media content items based on a search request from a client device to identify a set of collections (e.g., collections of media content items) of interest; and providing search response to the client device based on the searching, the search response identifying the stored collection of media content items as part of the set of collections of interest. The search response may provide access to the stored collection of media content items through at the client device.
The set of attributes may include at least one of a caption associated with the collection of media content items, a place name associated with the collection of media content items, and a category associated with the collection of media content items. The set of annotations may comprise at least one annotation relating to a time, a geographic location, a place, a topic, a visual feature, a caption, an event, a media quality of the particular media content item, or a user statistic relating to the particular media content item.
In the example architecture of
The operating system 4602 may manage hardware resources and provide common services. The operating system 4602 may include, for example, a kernel 4622, services 4624 and drivers 4626. The kernel 4622 may act as an abstraction layer between the hardware and the other software layers. For example, the kernel 4622 may be responsible for memory management, processor management (e.g., scheduling), component management, networking, security settings, and so on. The services 4624 may provide other common services for the other software layers. The drivers 4626 are responsible for controlling or interfacing with the underlying hardware. For instance, the drivers 4626 include display drivers, camera drivers, Bluetooth® drivers, flash memory drivers, serial communication drivers (e.g., Universal Serial Bus (USB) drivers), Wi-Fi® drivers, audio drivers, power management drivers, and so forth depending on the hardware configuration.
The libraries 4620 provide a common infrastructure that is used by the applications 4616 and/or other components and/or layers. The libraries 4620 provide functionality that allows other software components to perform tasks in an easier fashion than to interface directly with the underlying operating system 4602 functionality (e.g., kernel 4622, services 4624 and/or drivers 4626). The libraries 4620 may include system libraries 4644 (e.g., C standard library) that may provide functions such as memory allocation functions, string manipulation functions, mathematical functions, and the like. In addition, the libraries 4620 may include API libraries 4646 such as media libraries (e.g., libraries to support presentation and manipulation of various media format such as MPREG4, H.264, MP3, AAC, AMR, JPG, PNG), graphics libraries (e.g., an OpenGL framework that may be used to render 2D and 3D in a graphic content on a display), database libraries (e.g., SQLite that may provide various relational database functions), web libraries (e.g., WebKit that may provide web browsing functionality), and the like. The libraries 4620 may also include a wide variety of other libraries 4648 to provide many other APIs to the applications 4616 and other software components/modules.
The frameworks/middleware 4618 (also sometimes referred to as middleware) provide a higher-level common infrastructure that may be used by the applications 4616 and/or other software components/modules. For example, the frameworks/middleware 4618 may provide various graphic user interface (GUI) functions, high-level resource management, high-level location services, and so forth. The frameworks/middleware 4618 may provide a broad spectrum of other APIs that may be utilized by the applications 4616 and/or other software components/modules, some of which may be specific to a particular operating system 4602 or platform.
The applications 4616 include built-in applications 4638 and/or third-party applications 4640. Examples of representative built-in applications 4638 may include, but are not limited to, a contacts application, a browser application, a book reader application, a location application, a media application, a messaging application, and/or a game application. Third-party applications 4640 may include an application developed using the ANDROID™ or IOS™ software development kit (SDK) by an entity other than the vendor of the particular platform, and may be mobile software running on a mobile operating system such as IOS™, ANDROID™, WINDOWS® Phone, or other mobile operating systems. The third-party applications 4640 may invoke the API calls 4608 provided by the mobile operating system (such as operating system 4602) to facilitate functionality described in this document.
The applications 4616 may use built-in operating system functions (e.g., kernel 4622, services 4624 and/or drivers 4626), libraries 4620, and frameworks/middleware 4618 to create user interfaces to interact with users of the system. Alternatively, or additionally, in some systems, interactions with a user may occur through a presentation layer, such as presentation layer 4614. In these systems, the application/component “logic” can be separated from the aspects of the application/component that interact with a user.
The machine 4700 may include processors 4704, memory memory/storage 4706, and I/O components 4718, which may be configured to communicate with each other such as via a bus 4702. Processors 4704 may comprise of a single processor or, as shown, comprise of multiple processors (e.g., processors 4708 to 4712). The memory/storage 4706 may include a memory 4714, such as a main memory, or other memory storage, and a storage unit 4716, both accessible to the processors 4704 such as via the bus 4702. The storage unit 4716 and memory 4714 store the instructions 4710 embodying any one or more of the methodologies or functions described in this document. The instructions 4710 may also reside, completely or partially, within the memory 4714, within the storage unit 4716, within at least one of the processors 4704 (e.g., within the processor 4708's cache memory), or any suitable combination thereof, during execution thereof by the machine 4700. Accordingly, the memory 4714, the storage unit 4716, and the memory of processors 4704 are examples of machine-readable media.
The I/O components 4718 may include a wide variety of components to receive input, provide output, produce output, transmit information, exchange information, capture measurements, and so on. The specific I/O components 4718 that are included in a particular machine 4700 will depend on the type of machine. For example, portable machines such as mobile phones will likely include a touch input device or other such input mechanisms, while a headless server machine will likely not include such a touch input device. It will be appreciated that the I/O components 4718 may include many other components that are not shown in
In further example embodiments, the I/O components 4718 may include biometric components 4730, motion components 4734, environment components 4736, or position components 4738 among a wide array of other components. For example, the biometric components 4730 may include components to detect expressions (e.g., hand expressions, facial expressions, vocal expressions, body gestures, or eye tracking), measure biosignals (e.g., blood pressure, heart rate, body temperature, perspiration, or brain waves), identify a person (e.g., voice identification, retinal identification, facial identification, fingerprint identification, or electroencephalogram based identification), and the like. The motion components 4734 may include acceleration sensor components (e.g., accelerometer), gravitation sensor components, rotation sensor components (e.g., gyroscope), and so forth. The environment components 4736 may include, for example, illumination sensor components (e.g., photometer), temperature sensor components (e.g., one or more thermometer that detect ambient temperature), humidity sensor components, pressure sensor components (e.g., barometer), acoustic sensor components (e.g., one or more microphones that detect background noise), proximity sensor components (e.g., infrared sensors that detect nearby objects), gas sensors (e.g., gas detection sensors to detection concentrations of hazardous gases for safety or to measure pollutants in the atmosphere), or other components that may provide indications, measurements, or signals corresponding to a surrounding physical environment. The position components 4738 may include location sensor components (e.g., a Global Position System (GPS) receiver component), altitude sensor components (e.g., altimeters or barometers that detect air pressure from which altitude may be derived), orientation sensor components (e.g., magnetometers), and the like.
Communication may be implemented using a wide variety of technologies. The I/O components 4718 may include communication components 4740 operable to couple the machine 4700 to a network 4732 or devices 4720 via coupling 4722 and coupling 4724 respectively. For example, the communication components 4740 may include a network interface component or other suitable device to interface with the network 4732. In further examples, communication components 4740 may include wired communication components, wireless communication components, cellular communication components, Near Field Communication (NFC) components, Bluetooth® components (e.g., Bluetooth® Low Energy), Wi-Fi® components, and other communication components to provide communication via other modalities. The devices 4720 may be another machine or any of a wide variety of peripheral devices (e.g., a peripheral device coupled via a Universal Serial Bus (USB)).
Moreover, the communication components 4740 may detect identifiers or include components operable to detect identifiers. For example, the communication components 4740 may include Radio Frequency Identification (RFID) tag reader components, NFC smart tag detection components, optical reader components (e.g., an optical sensor to detect one-dimensional bar codes such as Universal Product Code (UPC) bar code, multi-dimensional bar codes such as Quick Response (QR) code, Aztec code, Data Matrix, Dataglyph, MaxiCode, PDF417, Ultra Code, UCC RSS-2D bar code, and other optical codes), or acoustic detection components (e.g., microphones to identify tagged audio signals). In addition, a variety of information may be derived via the communication components 4740, such as, location via Internet Protocol (IP) geo-location, location via Wi-Fi® signal triangulation, location via detecting a NFC beacon signal that may indicate a particular location, and so forth.
It will be understood that various components used in this context (e.g., system components) refers to a device, physical entity or logic having boundaries defined by function or subroutine calls, branch points, application program interfaces (APIs), or other technologies that provide for the partitioning or modularization of particular processing or control functions. Components may be combined via their interfaces with other components to carry out a machine process. A component may be a packaged functional hardware unit designed for use with other components and a part of a program that usually performs a particular function of related functions. Components may constitute either software components (e.g., code embodied on a machine-readable medium) or hardware components. A hardware component is a tangible unit capable of performing certain operations and may be configured or arranged in a certain physical manner. In various example embodiments, one or more computer systems (e.g., a standalone computer system, a client computer system, or a server computer system) or one or more hardware components of a computer system (e.g., a processor 4708 or a group of processors 4704) may be configured by software (e.g., an application 4616 or application portion) as a hardware component that operates to perform certain operations as described in this document. A hardware component may also be implemented mechanically, electronically, or any suitable combination thereof. For example, a hardware component may include dedicated circuitry or logic that is permanently configured to perform certain operations. A hardware component may be a special-purpose processor, such as a field-programmable gate array (FPGA) or an application specific integrated circuit (ASIC). A hardware component may also include programmable logic or circuitry that is temporarily configured by software to perform certain operations. For example, a hardware component may include software executed by a general-purpose processor 4708 or other programmable processor 4708. Once configured by such software, hardware components become specific machines (or specific components of a machine 4700) uniquely tailored to perform the configured functions and are no longer general-purpose processors 4704. It will be appreciated that the decision to implement a hardware component mechanically, in dedicated and permanently configured circuitry, or in temporarily configured circuitry (e.g., configured by software) may be driven by cost and time considerations. Accordingly, the phrase “hardware component” (or “hardware-implemented component”) should be understood to encompass a tangible entity, be that an entity that is physically constructed, permanently configured (e.g., hardwired), or temporarily configured (e.g., programmed) to operate in a certain manner or to perform certain operations described in this document. Considering embodiments in which hardware components are temporarily configured (e.g., programmed), each of the hardware components need not be configured or instantiated at any one instance in time. For example, where a hardware component comprises a general-purpose processor 4708 configured by software to become a special-purpose processor, the general-purpose processor 4708 may be configured as respectively different special-purpose processors (e.g., comprising different hardware components) at different times. Software accordingly configures a particular processor 4708 or processors 4704, for example, to constitute a particular hardware component at one instance of time and to constitute a different hardware component at a different instance of time. Hardware components can provide information to, and receive information from, other hardware components. Accordingly, the described hardware components may be regarded as being communicatively coupled. Where multiple hardware components exist contemporaneously, communications may be achieved through signal transmission (e.g., over appropriate circuits and buses) between or among two or more of the hardware components. In embodiments in which multiple hardware components are configured or instantiated at different times, communications between such hardware components may be achieved, for example, through the storage and retrieval of information in memory structures to which the multiple hardware components have access. For example, one hardware component may perform an operation and store the output of that operation in a memory device to which it is communicatively coupled. A further hardware component may then, at a later time, access the memory device to retrieve and process the stored output. Hardware components may also initiate communications with input or output devices, and can operate on a resource (e.g., a collection of information). The various operations of example methods described in this document may be performed, at least partially, by one or more processors 4704 that are temporarily configured (e.g., by software) or permanently configured to perform the relevant operations. Whether temporarily or permanently configured, such processors 4704 may constitute processor-implemented components that operate to perform one or more operations or functions described in this document. As used in this document, “processor-implemented component” refers to a hardware component implemented using one or more processors 4704. Similarly, the methods described in this document may be at least partially processor-implemented, with a particular processor 4708 or processors 4704 being an example of hardware. For example, at least some of the operations of a method may be performed by one or more processors 4704 or processor-implemented components. Moreover, the one or more processors 4704 may also operate to support performance of the relevant operations in a “cloud computing” environment or as a “software as a service” (SaaS). For example, at least some of the operations may be performed by a group of computers (as examples of machines 4700 including processors 4704), with these operations being accessible via a network 4732 (e.g., the Internet) and via one or more appropriate interfaces (e.g., an application program interface (API)). The performance of certain of the operations may be distributed among the processors 4704, not only residing within a single machine 4700, but deployed across a number of machines 4700. In some example embodiments, the processors 4704 or processor-implemented components may be located in a single geographic location (e.g., within a home environment, an office environment, or a server farm). In other example embodiments, the processors 4704 or processor-implemented components may be distributed across a number of geographic locations.
“CLIENT DEVICE” in this context refers to any machine that interfaces to a communications network to obtain resources from one or more server systems or other client devices. A client device may be, but is not limited to, a mobile phone, desktop computer, laptop, portable digital assistants (PDAs), smart phones, tablets, ultra books, netbooks, laptops, multi-processor systems, microprocessor-based or programmable consumer electronics, game consoles, set-top boxes, or any other communication device that a user may use to access a network.
“COMMUNICATIONS NETWORK” in this context refers to one or more portions of a network that may be an ad hoc network, an intranet, an extranet, a virtual private network (VPN), a local area network (LAN), a wireless LAN (WLAN), a wide area network (WAN), a wireless WAN (WWAN), a metropolitan area network (MAN), the Internet, a portion of the Internet, a portion of the Public Switched Telephone Network (PSTN), a plain old telephone service (POTS) network, a cellular telephone network, a wireless network, a Wi-Fi® network, another type of network, or a combination of two or more such networks. For example, a network or a portion of a network may include a wireless or cellular network and the coupling may be a Code Division Multiple Access (CDMA) connection, a Global System for Mobile communications (GSM) connection, or other type of cellular or wireless coupling. In this example, the coupling may implement any of a variety of types of data transfer technology, such as Single Carrier Radio Transmission Technology (1×RTT), Evolution-Data Optimized (EVDO) technology, General Packet Radio Service (GPRS) technology, Enhanced Data rates for GSM Evolution (EDGE) technology, third Generation Partnership Project (3GPP) including 3G, fourth generation wireless (4G) networks, Universal Mobile Telecommunications System (UMTS), High Speed Packet Access (HSPA), Worldwide Interoperability for Microwave Access (WiMAX), Long Term Evolution (LTE) standard, others defined by various standard setting organizations, other long range protocols, or other data transfer technology.
“EMPHEMERAL MESSAGE” in this context refers to a message that is accessible for a time-limited duration. An ephemeral message may be a text, an image, a video and the like. The access time for the ephemeral message may be set by the message sender. Alternatively, the access time may be a default setting or a setting specified by the recipient. Regardless of the setting technique, the message is transitory.
“MACHINE-READABLE MEDIUM” in this context refers to a component, device or other tangible media able to store instructions and data temporarily or permanently and may include, but is not be limited to, random-access memory (RAM), read-only memory (ROM), buffer memory, flash memory, optical media, magnetic media, cache memory, other types of storage (e.g., Erasable Programmable Read-Only Memory (EEPROM)) and/or any suitable combination thereof. The term “machine-readable medium” should be taken to include a single medium or multiple media (e.g., a centralized or distributed database, or associated caches and servers) able to store instructions. The term “machine-readable medium” shall also be taken to include any medium, or combination of multiple media, that is capable of storing instructions (e.g., code) for execution by a machine, such that the instructions, when executed by one or more processors of the machine, cause the machine to perform any one or more of the methodologies described in this document. Accordingly, a “machine-readable medium” refers to a single storage apparatus or device, as well as “cloud-based” storage systems or storage networks that include multiple storage apparatus or devices. The term “machine-readable medium” excludes signals per se.
“COMPONENT” in this context refers to a device, physical entity or logic having boundaries defined by function or subroutine calls, branch points, application program interfaces (APIs), or other technologies that provide for the partitioning or modularization of particular processing or control functions. Components may be combined via their interfaces with other components to carry out a machine process. A component may be a packaged functional hardware unit designed for use with other components and a part of a program that usually performs a particular function of related functions. Components may constitute either software components (e.g., code embodied on a machine-readable medium) or hardware components. A “hardware component” is a tangible unit capable of performing certain operations and may be configured or arranged in a certain physical manner. In various example embodiments, one or more computer systems (e.g., a standalone computer system, a client computer system, or a server computer system) or one or more hardware components of a computer system (e.g., a processor or a group of processors) may be configured by software (e.g., an application or application portion) as a hardware component that operates to perform certain operations as described in this document. A hardware component may also be implemented mechanically, electronically, or any suitable combination thereof. For example, a hardware component may include dedicated circuitry or logic that is permanently configured to perform certain operations. A hardware component may be a special-purpose processor, such as a Field-Programmable Gate Array (FPGA) or an Application Specific Integrated Circuit (ASIC). A hardware component may also include programmable logic or circuitry that is temporarily configured by software to perform certain operations. For example, a hardware component may include software executed by a general-purpose processor or other programmable processor. Once configured by such software, hardware components become specific machines (or specific components of a machine) uniquely tailored to perform the configured functions and are no longer general-purpose processors. It will be appreciated that the decision to implement a hardware component mechanically, in dedicated and permanently configured circuitry, or in temporarily configured circuitry (e.g., configured by software) may be driven by cost and time considerations. Accordingly, the phrase “hardware component” (or “hardware-implemented component”) should be understood to encompass a tangible entity, be that an entity that is physically constructed, permanently configured (e.g., hardwired), or temporarily configured (e.g., programmed) to operate in a certain manner or to perform certain operations described in this document. Considering embodiments in which hardware components are temporarily configured (e.g., programmed), each of the hardware components need not be configured or instantiated at any one instance in time. For example, where a hardware component comprises a general-purpose processor configured by software to become a special-purpose processor, the general-purpose processor may be configured as respectively different special-purpose processors (e.g., comprising different hardware components) at different times. Software accordingly configures a particular processor or processors, for example, to constitute a particular hardware component at one instance of time and to constitute a different hardware component at a different instance of time. Hardware components can provide information to, and receive information from, other hardware components. Accordingly, the described hardware components may be regarded as being communicatively coupled. Where multiple hardware components exist contemporaneously, communications may be achieved through signal transmission (e.g., over appropriate circuits and buses) between or among two or more of the hardware components. In embodiments in which multiple hardware components are configured or instantiated at different times, communications between such hardware components may be achieved, for example, through the storage and retrieval of information in memory structures to which the multiple hardware components have access. For example, one hardware component may perform an operation and store the output of that operation in a memory device to which it is communicatively coupled. A further hardware component may then, at a later time, access the memory device to retrieve and process the stored output. Hardware components may also initiate communications with input or output devices, and can operate on a resource (e.g., a collection of information). The various operations of example methods described in this document may be performed, at least partially, by one or more processors that are temporarily configured (e.g., by software) or permanently configured to perform the relevant operations. Whether temporarily or permanently configured, such processors may constitute processor-implemented components that operate to perform one or more operations or functions described in this document. As used in this document, “processor-implemented component” refers to a hardware component implemented using one or more processors. Similarly, the methods described in this document may be at least partially processor-implemented, with a particular processor or processors being an example of hardware. For example, at least some of the operations of a method may be performed by one or more processors or processor-implemented components. Moreover, the one or more processors may also operate to support performance of the relevant operations in a “cloud computing” environment or as a “software as a service” (SaaS). For example, at least some of the operations may be performed by a group of computers (as examples of machines including processors), with these operations being accessible via a network (e.g., the Internet) and via one or more appropriate interfaces (e.g., an Application Program Interface (API)). The performance of certain of the operations may be distributed among the processors, not only residing within a single machine, but deployed across a number of machines. In some example embodiments, the processors or processor-implemented components may be located in a single geographic location (e.g., within a home environment, an office environment, or a server farm). In other example embodiments, the processors or processor-implemented components may be distributed across a number of geographic locations.
“PROCESSOR” in this context refers to any circuit or virtual circuit (a physical circuit emulated by logic executing on an actual processor) that manipulates data values according to control signals (e.g., “commands”, “op codes”, “machine code”, etc.) and which produces corresponding output signals that are applied to operate a machine. A processor may, for example, be a Central Processing Unit (CPU), a Reduced Instruction Set Computing (RISC) processor, a Complex Instruction Set Computing (CISC) processor, a Graphics Processing Unit (GPU), a Digital Signal Processor (DSP), an Application Specific Integrated Circuit (ASIC), a Radio-Frequency Integrated Circuit (RFIC) or any combination thereof. A processor may further be a multi-core processor having two or more independent processors (sometimes referred to as “cores”) that may execute instructions contemporaneously.
“TIMESTAMP” in this context refers to a sequence of characters or encoded information identifying when a certain event occurred, for example giving date and time of day, sometimes accurate to a small fraction of a second.
Where a phrase similar to “at least one of A, B, or C,” “at least one of A, B, and C,” “one or more A, B, or C,” or “one or more of A, B, and C” is used, it is intended that the phrase be interpreted to mean that A alone may be present in an embodiment, B alone may be present in an embodiment, C alone may be present in an embodiment, or that any combination of the elements A, B and C may be present in a single embodiment; for example, A and B, A and C, B and C, or A and B and C.
As used in this document, the term “or” may be construed in either an inclusive or exclusive sense. Moreover, plural instances may be provided for resources, operations, or structures described herein as a single instance. Additionally, boundaries between various resources, operations, modules, engines, and data stores are somewhat arbitrary, and particular operations are illustrated in a context of specific illustrative configurations. Other allocations of functionality are envisioned and may fall within a scope of various embodiments of the present disclosure. In general, structures and functionality presented as separate resources in the example configurations may be implemented as a combined structure or resource. Similarly, structures and functionality presented as a single resource may be implemented as separate resources. These and other variations, modifications, additions, and improvements fall within a scope of embodiments of the present disclosure as represented by the appended claims. The specification and drawings are, accordingly, to be regarded in an illustrative rather than a restrictive sense.
It will be understood that changes and modifications may be made to the disclosed embodiments without departing from the scope of the present disclosure. These and other changes or modifications are intended to be included within the scope of the present disclosure.
This application is a continuation of U.S. patent application Ser. No. 18/341,542, filed Jun. 26, 2023, which application is a continuation of U.S. patent application Ser. No. 17/559,166, filed Dec. 22, 2021, which application is a continuation of U.S. patent application Ser. No. 15/900,187, filed on Feb. 20, 2018, now issued as U.S. Pat. No. 11,250,075, which claims the benefit of priority under 35 USC 119(e) of U.S. Provisional Patent Application Ser. No. 62/460,549, filed on Feb. 17, 2017, and U.S. Provisional Patent Application Ser. No. 62/460,583 filed on Feb. 17, 2017, the contents of each being incorporated herein by reference in their entireties.
| Number | Date | Country | |
|---|---|---|---|
| 62460549 | Feb 2017 | US | |
| 62460583 | Feb 2017 | US |
| Number | Date | Country | |
|---|---|---|---|
| Parent | 18341542 | Jun 2023 | US |
| Child | 18653380 | US | |
| Parent | 17559166 | Dec 2021 | US |
| Child | 18341542 | US | |
| Parent | 15900187 | Feb 2018 | US |
| Child | 17559166 | US |
