Patent Application
20150234901
The invention relates to a data collector arranged to collect data regarding application usage in an end user device.
Recommender systems are known in the art and are used for various purposes, such as recommending movies, music, pictures etc. Such recommender systems are for instance used by Amazon, Last.fm, etc. Recommender systems assist a user in finding interesting items without the user having to explicitly state what he or she wants.
A commonly used recommender method is collaborative filtering which produces recommendations by computing the similarity between users and/or items based on consumption history. Another well-known recommender method is content based recommendation. In essence, content based recommendations are based on a description, such as metadata, associated with the content. From a user's item consumption, preferences of the user in terms of item attributes are derived, to thereby find similar items.
One area where recommendations are applicable is applications (also known as apps) on smartphones, TVs, tablets and similar devices, that support the download and installation of applications by the end user. The number of applications available in different app stores is exploding, making it very hard for users to find applications that are relevant for them. By introducing a recommender system, this process is made easier for the end-users.
There are recommender systems known in the art. For example, Apple has provided a recommender system called Genius for applications on their iOS platform. However, such systems are limited in the data for basing the recommendations on.
It is an object to provide more relevant data regarding application usage for use by recommender systems.
According to a first aspect, it is presented a data collector arranged to collect data regarding application usage in an end user device. The data collector is arranged to be located in a mobile communication network between the end user device and a server. The data collector comprises: a processor; and a computer program product storing instructions that, when executed by the processor, causes the data collector to: obtain data sent between the end user device and the server; match the data against a list of patterns; and when a matching pattern is found in the list of patterns, store application activity associated with the matching pattern in a database for application usage.
Using a data collector located in a mobile communication network between the end user device and the server provides an opportunity to extract greatly relevant data. The pattern matching makes such data extraction relevant, even when a large amount of data is sent between the end user device and the server.
The instructions to match may comprise instructions to match a uniform resource identifier, URI, comprised in the data with the list of patterns. The URI, or other metadata, can be an efficient way to gain knowledge of a variety of application actions performed by the user.
The instructions to match may comprise instructions to match payload data of the data with the list of patterns. By matching patterns in payload data, it is possible to catch additional user actions regarding an application.
The instructions to store may comprise instructions to store a timestamp in the database. The timestamp can e.g. be used for analysis of when application usage, such as time of day, weekend, etc.
The computer program product may further comprise instructions to obtain additional data associated with the end user device; and the instructions to store may comprise instructions to store the additional data. Additional data can e.g. include location data. Optionally, the additional data can include subscriber data available through subscriber management systems of the mobile communication network.
The instructions to store may refrain from storing a user identifier. This improves privacy for the end user.
The computer program product may further comprise instructions to: reduce accuracy of any geographical location data associated with the end user device; and the instructions to store may comprise instructions to store the geographical location data with reduced accuracy. By not storing the exact location of the end user, privacy of the end user is improved.
The application activity may be an application activity in the list consisting of application installation, application uninstallation, application usage, application search in an application directory, and application usage duration. These are all application activities which can be used by a recommender system to provide application recommendations.
The network node may be arranged to be provided in a core network node of the mobile communication network.
According to a second aspect, it is presented a method for collecting data regarding application usage in an end user device, the method being performed in a data collector located between the end user device and a server. The method comprises the steps of: obtaining data sent between the end user device and the server; matching the data against a list of patterns; and when a matching pattern is found in the list of patterns, storing application activity associated with the matching pattern in a database for application usage.
The step of matching may comprise matching a uniform resource identifier comprised in the data with the list of patterns.
The step of matching may comprise matching payload data of the data with the list of patterns.
The step of storing may comprise storing a timestamp in the database.
The method may further comprise the step of: obtaining additional data associated with the end user device; and the step of storing may comprise storing the additional data.
The step of storing may refrain from storing a user identifier.
The method may further comprise the step of: reducing accuracy of any geographical location data associated with the end user device; and the step of storing may comprise storing the geographical location data with reduced accuracy.
In the step of storing, the application activity may be an application activity in the list consisting of application installation, application uninstallation, application usage, application search in an application directory, and application usage duration.
According to a third aspect, it is presented a computer program for collecting data regarding application usage in an end user device. The computer program comprises computer program code which, when run on a data collector located between an end user device and a server, causes the data collector to: obtain data sent between the end user device and the server; match the data against a list of patterns; and when a matching pattern is found in the list of patterns, store application activity associated with the matching pattern in a database for application usage.
According to a fourth aspect, it is presented a computer program product comprising a computer program according to the third aspect and a computer readable means on which the computer program is stored.
It is to be noted that any feature of the first, second, third or fourth aspects can, where applicable, form part of any other aspect.
Generally, all terms used in the claims are to be interpreted according to their ordinary meaning in the technical field, unless explicitly defined otherwise herein. All references to “a/an/the element, apparatus, component, means, step, etc.” are to be interpreted openly as referring to at least one instance of the element, apparatus, component, means, step, etc., unless explicitly stated otherwise. The steps of any method disclosed herein do not have to be performed in the exact order disclosed, unless explicitly stated.
The invention is now described, by way of example, with reference to the accompanying drawings, in which:
The invention will now be described more fully hereinafter with reference to the accompanying drawings, in which certain embodiments of the invention are shown. This invention may, however, be embodied in many different forms and should not be construed as limited to the embodiments set forth herein; rather, these embodiments are provided by way of example so that this disclosure will be thorough and complete, and will fully convey the scope of the invention to those skilled in the art. Like numbers refer to like elements throughout the description.
The mobile communication system 9 can e.g. comply with any one or a combination of LTE (Long Term Evolution), W-CDMA (Wideband Code Division Multiplex), EDGE (Enhanced Data Rates for GSM Evolution, GPRS (General Packet Radio Service)), CDMA2000 (Code Division Multiple Access 2000), etc., or any other existing or future mobile communication system standard, as long as the principles described hereinafter are applicable.
The network nodes 3 are also optionally connected to a radio network controller (RNC) 6 which is overall responsible for coordinating radio communication. The RNC is further connected via a Serving GPRS (General Packet Radio Service) Support Node (SGSN) 10, a gateway GPRS Support Node (GGSN) 11, and a proxy 12 to a wide area network 13.
The SGSN 10 routes data uplink and downlink and is responsible for management of the end user devices, such as attach/detach and location data. The GGSN 11 is the traditional gateway between the mobile communication network and external data networks such as the wide area network 13. The wide area network 13 can e.g. be the Internet.
Optionally, there is a proxy provided between the GGSN 11 and the wide area network 13.
There are furthermore one or more servers 15 connected to the wide area network 13. In this way, the end user device can communicate in either direction with the server 15.
One server 15 can house an application directory. The application directory allows a user of the end user device 2 to browse lists of applications which can be installed on the end user device 2. Furthermore, the application directory provides detailed data about the applications and links to install and/or uninstall the applications in the end user device 2. Furthermore, another server 15 can house the server end of an application installed in the end user device 2.
Regardless of the purpose of the server 15, the communication to and from the server 15 passes through the RNC 6, the SGSN 10, the GGSN 11 and the proxy 12, allowing the data sent between the end user device 2 and the server 15 in order to be analysed to extract data, e.g. regarding application usage in the end user device 2.
In
Optionally, different data collectors 1 or different parts of the data collector 1 can be housed in multiple devices.
The computer program product 54 can be a memory being any combination of read and write memory (RAM) and read only memory (ROM). The memory also comprises persistent storage, which, for example, can be any single one or combination of magnetic memory, optical memory, solid state memory or even remotely mounted memory. The processor 50 controls the general operation of the data collector 1.
The data collector 1 further comprises a data memory 59, being a read and write memory. The data memory 59 may also comprise persistent storage, which, for example, can be any single one or combination of magnetic memory, optical memory, solid state memory or even remotely mounted memory. Optionally, the computer program product 54 and the data memory 59 can form part of the same memory.
The data collector 1 further comprises an I/O interface 57 for communicating with other devices. Other components of the data collector 1 are omitted in order not to obscure the concepts presented herein.
In an obtain data step 60, data sent between the end user device and the server is obtained. This step can e.g. be part of the routing functions of the host device.
In a match data step 62, the data is matched against a list of patterns. The list of patterns can e.g. be stored in a table, where each pattern is associated with an application activity. The application activity can e.g. be application installation, application uninstallation, application usage, application search in an application directory. Other possible application activities are: rating of an application, liking an application, unliking an application. Also, the installation duration, defined as time between install and install of an application can be obtained. The patterns can for example be based on regular expressions (regexp).
The matching can for instance use a uniform resource identifier (URI) comprised in the data and match the URI with the list of patterns. Often, the URI contains commands for application installation or uninstallation, or commands for showing more data of an application in an application directory, etc. Alternatively or additionally other metadata can be used, such as header fields, etc.
Alternatively or additionally, the matching can use payload data of the data and match this with the list of patterns, e.g. using deep packet inspection (DPI).
In the conditional match step 65, the method is routed to a store application activity step 66 when a matching pattern is found in the list of patterns in the match data step 62. Otherwise, the method ends.
In the store application activity step 66, the application activity associated with the matching pattern is stored as application activity data in a database for application usage. Optionally, a timestamp is stored in the application activity data in the database for application usage. The timestamp can be the current time of the host device, a time extracted from the data or any other suitable time which is at least comparable to timestamps in other records in the database for application usage. This can be used for evaluation of time-of-day of an event, weekday/weekend, etc.
Optionally, the store application activity step 66 refrains from storing a user identifier, even if such an identifier has been obtained from the data. This may e.g. be due to privacy reasons as configured by the operator of the mobile communication network or by the end user.
By removing the user identifier, individual users are effectively decoupled from the application activity data to prevent, e.g. the ability to read that user X used application Y at location Z.
Preserving end user privacy can be done in several different ways. This can be done during data collection, by removing user identifier (key) from the data or during subsequent popularity calculations
During data collection, the user identity could be refrained from being stored at all, i.e. consumption data for a certain application does not need any user identity correlated to it.
By removing the key from the data, it is possible to keep statistical data without risking the privacy of the end user. This process can periodically generate random identifiers for the data (e.g. every hour), which makes it possible to keep track of users at a certain location, but not follow them over time. Of course, the actual data with the real key can still be forwarded to services that the user signed up to, which makes it possible for e.g. recommender systems to build collaborative filtering models on the usage data, without any need for background services on the actual clients.
During popularity calculations for a certain application all relevant consumption data for that application will be used. The amount of data can in this case vary from only a single entry to several thousands of entries. In the case of only one data entry, the exact location can be obfuscated by introducing an error to it. In the case there is a lot of data, all data can be used to calculate an average location for this application.
Optionally, as part of the store application activity step 66, a duration of application usage is calculated and stored in the database for application usage. This duration can e.g. be calculated as a time difference between the first active message for a particular application and the last active message for a particular application. The application usage duration can be reset after a certain period of inactivity, e.g. after 30 minutes of activity.
Optionally, the collected application activity data is aggregated prior to storing. Once the application activity data is stored in the database, the application activity data can be used by recommender systems, such as an application recommender system. The recommender system can then use the application activity data to calculate what application to recommend to different end users.
The recommendation of applications can, for example, be calculated based on:
In one example, the process comprises the following steps:
The method shown in
Here, the method is routed, in the conditional match step 65, to an obtain additional data step 63 when there is a match. Otherwise, the method returns to the obtain data step 60 to repeat the method.
In the optional obtain additional data step 63, additional data associated with the end user device is obtained. For example, geographical location can be obtained, either from a node in the mobile communication system of from the end user device.
The additional data could also be based on collecting other types of information, such as information from the Internet, weather channels, street maps, calendars etc. For example, the data collector can derive that there is a football game going on in the area or that bad weather is to occur in the region. Street maps and calendars may also add additional information such as public holidays and other metadata about the area.
In the optional reduce accuracy step 64, accuracy of any geographical location data associated with the end user device is reduced. For example, the geographical location could be stored as GPS coordinates with a reduced number of significant digits or with reference to a city or a region, even though greater accuracy is available.
In the store application activity step 66, the additional data is stored in the database for application data along with the application activity associated with the matching pattern. Also, when the reduce accuracy step 64 is performed, the geographical location data is stored with reduced accuracy, as explained above, to improve user privacy.
After the store application activity step 66, the method returns to the obtain data step 60.
The invention has mainly been described above with reference to a few embodiments. However, as is readily appreciated by a person skilled in the art, other embodiments than the ones disclosed above are equally possible within the scope of the invention, as defined by the appended patent claims.
| Filing Document | Filing Date | Country | Kind | 371c Date |
|---|---|---|---|---|
| PCT/SE2012/050922 | 8/30/2012 | WO | 00 | 3/26/2015 |
