Patent Application
20180109574
People within a working group, especially busy professionals, may contend with a deluge of information shared among collaborators and it may be overwhelming. The information may come in one or more forms, from one or more sources, and may include such things as, e.g., SMS messages, documents, webpages, email alerts, email, blogs, news feeds, social media messages/feeds, and many others. Collaboration tools today generally only work better if everyone used them, yet it is often difficult to get everyone in a large organization to adopt the same collaboration tool and it becomes even more difficult when multiple organizations are involved. For example, if a company needs to do a project with a partner, client, or vendor, they often do not have a shared collaboration tool.
In one example implementation, a method, performed by one or more computing devices, may include but is not limited to acquiring, by a computing device, data representing a plurality of collaboration items, each collaboration item being associated with one of a communication and a collaboration among a subset of one or more users. Using a machine learning procedure, one of at least one latent variable and at least one action variable in a model of the data representing the plurality of collaboration items may be determined. At least one of a representation of the collaboration items may be presented to one or more users based upon, at least in part, the at least one latent variable, and potential collaboration actions may be presented to the one or more users based upon, at least in part, the at least one action variable.
One or more of the following example features may be included. The model of the data may be a model of at least one of human collaboration and relationships. One or more of the at least one action variable the at least one latent variable in the model of the data may include information identifying at least one of what task users of the one or more users are working on, what users of the one or more users are working on the task together, when the task is being worked on, why the task is being worked on, and how the one or more users participate in the collaboration. Why the task is being worked on may include a relationship of the project being worked on relative to one or more other projects within the collaboration. The machine learning procedure may infer one or more of the at least one action variable the at least one latent variable about at least one of what task users of the one or more users are working on, what users of the one or more users are working on the task together, when the task is being worked on, why the task is being worked on, and how the one or more users participate in the collaboration. The machine learning process may include a second probabilistic model generated by modifying a first probabilistic model, the modification based upon, at least in part, inferences of one or more of the at least one action variable the at least one latent variable. One or more of the at least one latent variable and the at least one action variable determined using the machine learning procedure may be based upon, at least in part, user feedback received from the one or more users.
In another example implementation, a computing system may include one or more processors and one or more memories configured to perform operations that may include but are not limited to acquiring, by a computing device, data representing a plurality of collaboration items, each collaboration item being associated with one of a communication and a collaboration among a subset of one or more users. Using a machine learning procedure, one of at least one latent variable and at least one action variable in a model of the data representing the plurality of collaboration items may be determined. At least one of a representation of the collaboration items may be presented to one or more users based upon, at least in part, the at least one latent variable, and potential collaboration actions may be presented to the one or more users based upon, at least in part, the at least one action variable.
One or more of the following example features may be included. The model of the data may be a model of at least one of human collaboration and relationships. One or more of the at least one action variable the at least one latent variable in the model of the data may include information identifying at least one of what task users of the one or more users are working on, what users of the one or more users are working on the task together, when the task is being worked on, why the task is being worked on, and how the one or more users participate in the collaboration. Why the task is being worked on may include a relationship of the project being worked on relative to one or more other projects within the collaboration. The machine learning procedure may infer one or more of the at least one action variable the at least one latent variable about at least one of what task users of the one or more users are working on, what users of the one or more users are working on the task together, when the task is being worked on, why the task is being worked on, and how the one or more users participate in the collaboration. The machine learning process may include a second probabilistic model generated by modifying a first probabilistic model, the modification based upon, at least in part, inferences of one or more of the at least one action variable the at least one latent variable. One or more of the at least one latent variable and the at least one action variable determined using the machine learning procedure may be based upon, at least in part, user feedback received from the one or more users.
In another example implementation, a computer program product may reside on a computer readable storage medium having a plurality of instructions stored thereon which, when executed across one or more processors, may cause at least a portion of the one or more processors to perform operations that may include but are not limited to acquiring, by a computing device, data representing a plurality of collaboration items, each collaboration item being associated with one of a communication and a collaboration among a subset of one or more users. Using a machine learning procedure, one of at least one latent variable and at least one action variable in a model of the data representing the plurality of collaboration items may be determined. At least one of a representation of the collaboration items may be presented to one or more users based upon, at least in part, the at least one latent variable, and potential collaboration actions may be presented to the one or more users based upon, at least in part, the at least one action variable.
One or more of the following example features may be included. The model of the data may be a model of at least one of human collaboration and relationships. One or more of the at least one action variable the at least one latent variable in the model of the data may include information identifying at least one of what task users of the one or more users are working on, what users of the one or more users are working on the task together, when the task is being worked on, why the task is being worked on, and how the one or more users participate in the collaboration. Why the task is being worked on may include a relationship of the project being worked on relative to one or more other projects within the collaboration. The machine learning procedure may infer one or more of the at least one action variable the at least one latent variable about at least one of what task users of the one or more users are working on, what users of the one or more users are working on the task together, when the task is being worked on, why the task is being worked on, and how the one or more users participate in the collaboration. The machine learning process may include a second probabilistic model generated by modifying a first probabilistic model, the modification based upon, at least in part, inferences of one or more of the at least one action variable the at least one latent variable. One or more of the at least one latent variable and the at least one action variable determined using the machine learning procedure may be based upon, at least in part, user feedback received from the one or more users.
The details of one or more example implementations are set forth in the accompanying drawings and the description below. Other possible example features and/or possible example advantages will become apparent from the description, the drawings, and the claims. Some implementations may not have those possible example features and/or possible example advantages, and such possible example features and/or possible example advantages may not necessarily be required of some implementations.
Like reference symbols in the various drawings indicate like elements.
People within a working group, especially busy professionals, may contend with a deluge of information and it may be overwhelming. The information may come in one or more forms, from one or more sources, and may include such things as, e.g., SMS messages, documents, webpages, email alerts, email, blogs, news feeds, social media messages/feeds, and many others. Collaboration tools today generally only work better if everyone is using them, yet it is often difficult to get everyone in a large organization to adopt the same collaboration tool and it becomes even more difficult when multiple organizations are involved. For example, if a company needs to do a project with a partner, client, or vendor, they often do not have a shared collaboration tool.
For instance, using email as an example, one's email may come from many different sources. One person may be sent an email to do something for a project being worked on with another. Later, one may be sent an unrelated email asking if the group should buy a new printer. Generally, one's email browser may show these emails in chronological order, one right after the other, regardless of the fact that moving from one to the next may require an entire context switch for your brain. Documents from multiple unrelated projects and collaborations may be interleaved and shuffled together with no rhyme or reason.
The situation may be exacerbated on one's phone or tablet or other smaller screen mobile computing device. For those devices, to get something done that requires multiple documents, one may have to constantly switch between different apps, and then scroll or move to the relevant document within each app.
As such, the present disclosure may deal with these example issues that each source of documents within a collaboration that one may consume has no way of knowing about the documents coming from other sources, and therefore all of one's sources may be merged together in a disorganized way regardless of the project or collaboration they are related to. As will be discussed in greater detail below, the present disclosure may take all of these information feeds within a collaboration, and automatically re-collate and organize them in a manner that may be beneficial for how and/or when one plans to use or consume the information.
In some implementations, the present disclosure may be embodied as a method, system, or computer program product. Accordingly, in some implementations, the present disclosure may take the form of an entirely hardware implementation, an entirely software implementation (including firmware, resident software, micro-code, etc.) or an implementation combining software and hardware aspects that may all generally be referred to herein as a “circuit,” “module” or “system.” Furthermore, in some implementations, the present disclosure may take the form of a computer program product on a computer-usable storage medium having computer-usable program code embodied in the medium.
In some implementations, any suitable computer usable or computer readable medium (or media) may be utilized. The computer readable medium may be a computer readable signal medium or a computer readable storage medium. The computer-usable, or computer-readable, storage medium (including a storage device associated with a computing device or client electronic device) may be, for example, but is not limited to, an electronic, magnetic, optical, electromagnetic, infrared, or semiconductor system, apparatus, device, or any suitable combination of the foregoing. More specific examples (a non-exhaustive list) of the computer-readable medium may include the following: an electrical connection having one or more wires, a portable computer diskette, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or Flash memory), an optical fiber, a portable compact disc read-only memory (CD-ROM), an optical storage device, a digital versatile disk (DVD), a static random access memory (SRAM), a memory stick, a floppy disk, a mechanically encoded device such as punch-cards or raised structures in a groove having instructions recorded thereon, a media such as those supporting the internet or an intranet, or a magnetic storage device. Note that the computer-usable or computer-readable medium could even be a suitable medium upon which the program is stored, scanned, compiled, interpreted, or otherwise processed in a suitable manner, if necessary, and then stored in a computer memory. In the context of the present disclosure, a computer-usable or computer-readable, storage medium may be any tangible medium that can contain or store a program for use by or in connection with the instruction execution system, apparatus, or device.
In some implementations, a computer readable signal medium may include a propagated data signal with computer readable program code embodied therein, for example, in baseband or as part of a carrier wave. In some implementations, such a propagated signal may take any of a variety of forms, including, but not limited to, electro-magnetic, optical, or any suitable combination thereof. In some implementations, the computer readable program code may be transmitted using any appropriate medium, including but not limited to the internet, wireline, optical fiber cable, RF, etc. In some implementations, a computer readable signal medium may be any computer readable medium that is not a computer readable storage medium and that can communicate, propagate, or transport a program for use by or in connection with an instruction execution system, apparatus, or device.
In some implementations, computer program code for carrying out operations of the present disclosure may be assembler instructions, instruction-set-architecture (ISA) instructions, machine instructions, machine dependent instructions, microcode, firmware instructions, state-setting data, or either source code or object code written in any combination of one or more programming languages, including an object oriented programming language such as Java®, Smalltalk, C++ or the like. Java® and all Java-based trademarks and logos are trademarks or registered trademarks of Oracle and/or its affiliates. However, the computer program code for carrying out operations of the present disclosure may also be written in conventional procedural programming languages, such as the “C” programming language, PASCAL, or similar programming languages, as well as in scripting languages such as Javascript, PERL, or Python. The program code may execute entirely on the user's computer, partly on the user's computer, as a stand-alone software package, partly on the user's computer and partly on a remote computer or entirely on the remote computer or server. In the latter scenario, the remote computer may be connected to the user's computer through a local area network (LAN) or a wide area network (WAN), or the connection may be made to an external computer (for example, through the internet using an Internet Service Provider). In some implementations, electronic circuitry including, for example, programmable logic circuitry, field-programmable gate arrays (FPGAs) or other hardware accelerators, micro-controller units (MCUs), or programmable logic arrays (PLAs) may execute the computer readable program instructions/code by utilizing state information of the computer readable program instructions to personalize the electronic circuitry, in order to perform aspects of the present disclosure.
In some implementations, the flowchart and block diagrams in the figures illustrate the architecture, functionality, and operation of possible implementations of apparatus (systems), methods and computer program products according to various implementations of the present disclosure. Each block in the flowchart and/or block diagrams, and combinations of blocks in the flowchart and/or block diagrams, may represent a module, segment, or portion of code, which comprises one or more executable computer program instructions for implementing the specified logical function(s)/act(s). These computer program instructions may be provided to a processor of a general purpose computer, special purpose computer, or other programmable data processing apparatus to produce a machine, such that the computer program instructions, which may execute via the processor of the computer or other programmable data processing apparatus, create the ability to implement one or more of the functions/acts specified in the flowchart and/or block diagram block or blocks or combinations thereof. It should be noted that, in some implementations, the functions noted in the block(s) may occur out of the order noted in the figures (or combined or omitted). For example, two blocks shown in succession may, in fact, be executed substantially concurrently, or the blocks may sometimes be executed in the reverse order, depending upon the functionality involved.
In some implementations, these computer program instructions may also be stored in a computer-readable memory that can direct a computer or other programmable data processing apparatus to function in a particular manner, such that the instructions stored in the computer-readable memory produce an article of manufacture including instruction means which implement the function/act specified in the flowchart and/or block diagram block or blocks or combinations thereof
In some implementations, the computer program instructions may also be loaded onto a computer or other programmable data processing apparatus to cause a series of operational steps to be performed (not necessarily in a particular order) on the computer or other programmable apparatus to produce a computer implemented process such that the instructions which execute on the computer or other programmable apparatus provide steps for implementing the functions/acts (not necessarily in a particular order) specified in the flowchart and/or block diagram block or blocks or combinations thereof.
Referring now to the example implementation of
In some implementations, as will be discussed below in greater detail, an organization process (OP), such as OP 10 of
In some implementations, the instruction sets and subroutines of OP 10, which may be stored on storage device, such as storage device 16, coupled to computer 12, may be executed by one or more processors and one or more memory architectures included within computer 12. In some implementations, storage device 16 may include but is not limited to: a hard disk drive; a flash drive, a tape drive; an optical drive; a RAID array (or other array); a random access memory (RAM); and a read-only memory (ROM).
In some implementations, network 14 may be connected to one or more secondary networks (e.g., network 18), examples of which may include but are not limited to: a local area network; a wide area network; or an intranet, for example.
In some implementations, computer 12 may include a data store, such as a database (e.g., relational database, object-oriented database, triplestore database, etc.) and may be located within any suitable memory location, such as storage device 16 coupled to computer 12. In some implementations, data, metadata, information, etc. described throughout the present disclosure may be stored in the data store. In some implementations, computer 12 may utilize any known database management system such as, but not limited to, DB2, in order to provide multi-user access to one or more databases, such as the above noted relational database. In some implementations, the data store may also be a custom database, such as, for example, a flat file database or an XML database. In some implementations, any other form(s) of a data storage structure and/or organization may also be used. In some implementations, OP 10 may be a component of the data store, a standalone application that interfaces with the above noted data store and/or an applet/application that is accessed via client applications 22, 24, 26, 28. In some implementations, the above noted data store may be, in whole or in part, distributed in a cloud computing topology. In this way, computer 12 and storage device 16 may refer to multiple devices, which may also be distributed throughout the network.
In some implementations, computer 12 may execute a collaboration application (e.g., collaboration application 20), examples of which may include, but are not limited to, e.g., a web conferencing application, a video conferencing application, a voice-over-IP application, a video-over-IP application, an Instant Messaging (IM)/“chat” application, a short messaging service (SMS)/multimedia messaging service (MMS) application, an email application, a social media application, a website application, or other application that allows for virtual meeting and/or remote collaboration. In some implementations, OP 10 and/or collaboration application 20 may be accessed via one or more of client applications 22, 24, 26, 28. In some implementations, OP 10 may be a standalone application, or may be an applet/application/script/extension that may interact with and/or be executed within collaboration application 20, a component of collaboration application 20, and/or one or more of client applications 22, 24, 26, 28. In some implementations, collaboration application 20 may be a standalone application, or may be an applet/application/script/extension that may interact with and/or be executed within OP 10, a component of OP 10, and/or one or more of client applications 22, 24, 26, 28. In some implementations, one or more of client applications 22, 24, 26, 28 may be a standalone application, or may be an applet/application/script/extension that may interact with and/or be executed within and/or be a component of OP 10 and/or collaboration application 20. Examples of client applications 22, 24, 26, 28 may include, but are not limited to, e.g., a web conferencing application, a video conferencing application, a voice-over-IP application, a video-over-IP application, an Instant Messaging (IM)/“chat” application, a short messaging service (SMS)/multimedia messaging service (MMS) application, an email application, a social media application, a website application, or other application that allows for virtual meeting and/or remote collaboration, a standard and/or mobile web browser, an email application (e.g., an email client application), a textual and/or a graphical user interface, a customized web browser, a plugin, an Application Programming Interface (API), or a custom application. The instruction sets and subroutines of client applications 22, 24, 26, 28, which may be stored on storage devices 30, 32, 34, 36, coupled to client electronic devices 38, 40, 42, 44, may be executed by one or more processors and one or more memory architectures incorporated into client electronic devices 38, 40, 42, 44.
In some implementations, one or more of storage devices 30, 32, 34, 36, may include but are not limited to: hard disk drives; flash drives, tape drives; optical drives; RAID arrays; random access memories (RAM); and read-only memories (ROM). Examples of client electronic devices 38, 40, 42, 44 (and/or computer 12) may include, but are not limited to, a personal computer (e.g., client electronic device 38), a laptop computer (e.g., client electronic device 40), a smart/data-enabled, cellular phone (e.g., client electronic device 42), a notebook computer (e.g., client electronic device 44), a tablet, a server, a television, a smart television, a media (e.g., video, photo, etc.) capturing device, and a dedicated network device. Client electronic devices 38, 40, 42, 44 may each execute an operating system, examples of which may include but are not limited to, Android™, Apple® iOS®, Mac® OS X®; Red Hat® Linux®, Windows® Mobile, Chrome OS, Blackberry OS, Fire OS, or a custom operating system.
In some implementations, one or more of client applications 22, 24, 26, 28 may be configured to effectuate some or all of the functionality of OP 10 (and vice versa). Accordingly, in some implementations, OP 10 may be a purely server-side application, a purely client-side application, or a hybrid server-side/client-side application that is cooperatively executed by one or more of client applications 22, 24, 26, 28 and/or OP 10.
In some implementations, one or more of client applications 22, 24, 26, 28 may be configured to effectuate some or all of the functionality of collaboration application 20 (and vice versa). Accordingly, in some implementations, collaboration application 20 may be a purely server-side application, a purely client-side application, or a hybrid server-side/client-side application that is cooperatively executed by one or more of client applications 22, 24, 26, 28 and/or collaboration application 20. As one or more of client applications 22, 24, 26, 28, OP 10, and collaboration application 20, taken singly or in any combination, may effectuate some or all of the same functionality, any description of effectuating such functionality via one or more of client applications 22, 24, 26, 28, OP 10, collaboration application 20, or combination thereof, and any described interaction(s) between one or more of client applications 22, 24, 26, 28, OP 10, collaboration application 20, or combination thereof to effectuate such functionality, should be taken as an example only and not to limit the scope of the disclosure.
In some implementations, one or more of users 46, 48, 50, 52 may access computer 12 and OP 10 (e.g., using one or more of client electronic devices 38, 40, 42, 44) directly through network 14 or through secondary network 18. Further, computer 12 may be connected to network 14 through secondary network 18, as illustrated with phantom link line 54. OP 10 may include one or more user interfaces, such as browsers and textual or graphical user interfaces, through which users 46, 48, 50, 52 may access OP 10.
In some implementations, the various client electronic devices may be directly or indirectly coupled to network 14 (or network 18). For example, client electronic device 38 is shown directly coupled to network 14 via a hardwired network connection. Further, client electronic device 44 is shown directly coupled to network 18 via a hardwired network connection. Client electronic device 40 is shown wirelessly coupled to network 14 via wireless communication channel 56 established between client electronic device 40 and wireless access point (i.e., WAP) 58, which is shown directly coupled to network 14. WAP 58 may be, for example, an IEEE 802.11a, 802.11b, 802.11g, 802.11n, 802.11ac, Wi-Fi®, RFID, and/or Bluetooth™ (including Bluetooth™ Low Energy) device that is capable of establishing wireless communication channel 56 between client electronic device 40 and WAP 58. Client electronic device 42 is shown wirelessly coupled to network 14 via wireless communication channel 60 established between client electronic device 42 and cellular network/bridge 62, which is shown by example directly coupled to network 14.
In some implementations, some or all of the IEEE 802.11x specifications may use Ethernet protocol and carrier sense multiple access with collision avoidance (i.e., CSMA/CA) for path sharing. The various 802.11x specifications may use phase-shift keying (i.e., PSK) modulation or complementary code keying (i.e., CCK) modulation, for example. Bluetooth™ (including Bluetooth™ Low Energy) is a telecommunications industry specification that allows, e.g., mobile phones, computers, smart phones, and other electronic devices to be interconnected using a short-range wireless connection. Other forms of interconnection (e.g., Near Field Communication (NFC)) may also be used.
Referring also to the example implementation of
In some implementations, client electronic device 38 may include a processor (e.g., microprocessor 200) configured to, e.g., process data and execute the above-noted code/instruction sets and subroutines. Microprocessor 200 may be coupled via a storage adaptor to the above-noted storage device(s) (e.g., storage device 30). An I/O controller (e.g., I/O controller 202) may be configured to couple microprocessor 200 with various devices (e.g., via wired or wireless connection), such as keyboard 206, pointing/selecting device (e.g., touchpad, touchscreen, mouse 208, etc.), custom device (e.g., device 215), USB ports, and printer ports. A display adaptor (e.g., display adaptor 210) may be configured to couple display 212 (e.g., touchscreen monitor(s), plasma, CRT, or LCD monitor(s), etc.) with microprocessor 200, while network controller/adaptor 214 (e.g., an Ethernet adaptor) may be configured to couple microprocessor 200 to the above-noted network 14 (e.g., the Internet or a local area network).
As will be discussed below, OP 10 may at least help, e.g., improve existing feed organizational technologies necessarily rooted in machine learning computer technology, in order to overcome an example and non-limiting problem specifically arising in the realm of computer networks, and improve existing technological processes associated with, e.g., culminating, organizing, and/or providing multiple content from feeds using machine learning technology.
As discussed above and referring also at least to the example implementations of
As also discussed above, and referring also at least to the example implementations of
As will be discussed below, in some implementations, OP 10 may (e.g., via machine learning procedures) support group collaboration and relationships.
In some implementations, OP process may acquire 316, by a computing device, data representing a plurality of collaboration items, each collaboration item being associated with one of a communication and a collaboration among a subset of one or more users. For instance, in some implementations, organization process OP 10 may identify 300, by a computing device (e.g., computer 12), a plurality of content from at least one source. For instance, assume for example purposes only that that an email server is the source. In the example, one or more emails received by a user (e.g., user 50) may be the content identified 300 by OP 10. It will be appreciated that the source may include other sources. For example, the source of content may include an entire network of entities communicating with each other. An entire network of entities could be an entire governmental organization such as the Internal Revenue Service, where communications from the Internal Revenue Service are considered to be from the entire network of people that make up that organization rather than any one individual in that organization. As another example, a user may include a “virtual user”, where an example of a virtual user may include an automated computer-implemented system that sends (e.g., via OP 10) messages and/or reminders, such as a computer-implemented personal assistant, or computer-implemented system that keeps track of meeting schedules within an organization. As yet another example, user 50 may receive content/communication in the form of, e.g., tweets, instant messages, text messages, or any other form of digital communication including those noted above. As such, the use of emails as “content” and/or receiving content from any particular source(s) should be taken as example only and not to otherwise limit the scope of the disclosure.
In some implementations, and referring at least to the example implementation of
In some implementations, OP 10 may categorize 302 a first portion of the plurality of content in a first feed category based on a first probabilistic model and OP 10 may categorize 304 a second portion of the plurality of content in a second feed category based on the first probabilistic model. For instance, and referring at least to the example implementation of
In some implementations, the machine learning procedure may infer one or more of the at least one action variable the at least one latent variable about at least one of what task users of the one or more users are working on, what users of the one or more users are working on the task together, when the task is being worked on, why the task is being worked on, and how the one or more users participate in the collaboration. For example, consider a group of office workers preparing a budget estimate for a company. The machine learning procedure of OP 10 may infer that the list of users are working on the budget estimation project. Furthermore, the machine learning procedure of OP 10 may infer that the name of the project is ‘budget proposal’. Furthermore, the machine learning procedure of OP 10 may infer when the project is being worked on in the form of start and end dates as well as milestones within the project. Furthermore, the machine learning procedure of OP 10 may infer why the task is being worked on in the form of the relationship between this project and other projects, such as the ‘budget proposal’ is a sub project of the larger project ‘presentation to advisory board’. Furthermore, the machine learning procedure of OP 10 may infer how the team members communicate with one another in the form of a list of the types of communications used by the team, such as email, SMS text, or other messaging methods.
In some implementations, OP 10 may use machine learning system to categorize 302/304 the content by, e.g., inferring latent variables from structured and unstructured communication arising in a collaboration, such as: which projects are active, which members are working on which projects and how much, predicted project timelines and progress, which emails are related to which project, keywords or semantic groups of words that are related to particular projects, and the temporal state of a project or relationship. In some implementations, OP 10 may use machine learning to generate titles automatically for the channels by, e.g., looking for statistically salient N grams in the content within that channel, and/or user 50 may generate his/her own titles. For instance, an example title may be “Patents.” In the example, statistically, “Patents” may be a word or words that is used frequently in a particular feed category, or a word or words that is used frequently and early on to “kick off” a conversation, or a word or words that is used by a person who is a hub in the conversation (e.g., a person who statistically people tend to email and the person tends to respond to everyone), frequently and early on in the conversation.
In some implementations, the model of the data may be a model of at least one of human collaboration and relationships. For example the model (e.g., via OP 10) may include explicit relationships between humans such as the hierarchical relationship between managers and staff members. The model (e.g., via OP 10) may assume that managers within the group will act as ‘hubs’ of communication and be more likely to receive and send more communications than staff members. This assumption may help identify managers and staff members from the communications even in the case where managers and staff members are not explicitly identified. The model (e.g., via OP 10) may further include the human relationship notion of ‘friends’ as individuals who communicate about non-work related topics in addition to work related topics. For example, the model (via OP 10) may identify phrases within a message such as ‘what are you doing this weekend’ and ‘how is your family’ as non-work related topics and identify individuals sending and receiving these messages as ‘friends’.
In some implementations, OP 10 may determine 318, using a machine learning procedure, one of at least one latent variable and at least one action variable in a model of the data representing the plurality of collaboration items. For instance, as will be discussed below, OP 10 may use a machine learning system that may infer latent variables and/or action variables using observed data arising in a collaboration and/or relationship.
In some implementations, why the task is being worked on may include a relationship of the project being worked on relative to one or more other projects within the collaboration. For example, the model (e.g., via OP 10) may contain a variable that may store the fact that a project may be being worked on because, e.g., it is a sub-project of a larger project, such as the completion of a ‘budget report’ project as a component of a ‘presentation to advisory board’ project. In another example, the model (e.g., via OP 10) may contain a variable that may store the fact that the project ‘budget report’ must begin after the project ‘quarterly tax estimation’. It will be appreciated that other variables may be used without departing from the scope of the disclosure.
Generally, latent variables may be described as values that the machine learning system (e.g., via OP 10) cannot directly observe, but may be estimated or inferred. Examples of latent variables may include, but are not limited to, which projects are active, which members are working on which projects and how much, predicted project timelines and progress, which emails are related to which project, keywords or semantic groups of words that are related to particular projects, the temporal state of a project or relationship (e.g., is it just starting, is it nearing a particular milestone, or is it nearing completion), a field in a CRM or WM database that is not directly observable from data, among many others. In some implementations, the variables may include observable data, such as cells that are directly observable from data without inference, and may include, e.g., days since last activity or date of last contact, message time stamps, message contents, message meta data, meeting dates and times, git commits, documents created and/or edited, cloud storage sync's, data from wearable devices, among many others.
In some implementations, OP 10 may use combinations of subject matter, people involved in a collaboration and/or relationship, temporal relationships in collaborations and/or relationships, or with other events in order to make more advanced and accurate classifications of messages, recommend documents, understand the latent structure of collaborations, and provide a wide range of enhancements and support for collaborations and/or relationships. For example, imagine a software development working group which is communicating about a project they are working on. Further imagine that an important milestone has been identified by the engineers within the workgroup as being impossible to complete on time. OP 10 can identify the ‘milestone will not be completed on time’ status of the collaboration by looking for words and phrases in the communications like ‘milestone’, ‘problem’, ‘not enough time’, ‘missed deadline’, and further note that the manager of the working group is not receiving these communications. OP 10 may then recommend an enhancement of the collaboration by suggesting to the work group that they forward some of their messages to the manager to keep her up to date with project planning. In some examples, OP 10 may parse messages and/or other unstructured information in order to extract information about tasks, including for example who is doing what with whom, when, and/or why.
In some implementations, the probabilistic model used to categorize 302/304 the content into its appropriate channel may include “feeding” the content into an unsupervised topic model. In other examples, the topic model may be supervised or semi-supervised. An example of a topic model may be Latent Dirichlet Allocation (LDA). Inference on the topic model, conditioned on the content, may result in topic vectors. A topic vector for content may be a collection of probabilities (or at least magnitudes) representing the degree to which each topic is present within the given content. A topic, itself, may be a collection of probabilities (or at least magnitudes) representing the degree to which each of many possible words is present within a topic.
In some implementations, categorization 302/304 via probabilistic modeling may go beyond topic clustering. For example, OP 10 may sample a probabilistic process that may generate Gantt charts from a prior probability distributions over Gantt charts. A Gantt chart may include in this example context, projects, sub-projects, sub-sub-projects, and so forth. It may also include names or other IDs of people who may work on one or more project spreading their total effort across one or more projects. It may include timelines for projects, where projects all tend to be, e.g., 4 years long, and where projects tend to be, e.g., 4× longer than their sub-projects which may average, e.g., a year in length, and so forth with sub-sub-projects being on average, e.g., 3 months in length. In some implementations, content authored within a project may have a time stamp that may fit the duration of the project (or sub-project or sub-sub-project, etc.). Each project, sub-project, or sub-sub-project may have a distribution over the occurrences or co-occurrences of words in that topic. Content authored within a project may have word statistics similar to other content within the same project (or sub-project or sub-sub-project, etc.). The authors and any recipients/readers of an element of content may fit the overall distribution over people participating within a project (or sub-project or sub-sub-project, etc.).
It will be appreciated that any technique capable of categorizing content may be used without departing from the scope of the disclosure. For instance, the amount of content N may be chosen dynamically by a relevance algorithm. In another example, all the content within some radius of the exemplar content may be found and categorized into the appropriate channel. In another example, spheres may be defined around the ends of the topic vectors of each of the exemplar contents, and all of the contents within the region of intersection of these spheres may be found and categorized into the appropriate channel. As another example, the topic vectors may be clustered by a clustering module. For instance, the clustering module may be K-Means. In another example, OP 10 may order the display of content via date, or according to some other property, like distance within topic vector space (e.g., the measure of nearness may be a Euclidean distance in a vector space of the topic vectors). As yet another example, OP 10 may include, e.g., neural networks, support vector machines, probabilistic graphical models, probabilistic programs, probabilistic context free grammars, natural language parsers, and/or other machine learning components. Additionally, it will be appreciated that probabilistic programs are only one such representation of probabilistic models, and that the generative models may be, but are not necessarily, represented as probabilistic programs. As such, the use of any particular categorizing algorithm used for machine learning with the probabilistic models should be taken as example only and not to otherwise limit the scope of the disclosure.
In some implementations, once a collection of content is found using one or more example contents, the contents may be given a label to mark them as all belonging to a cluster or class of contents within a content ontology, which may be output by OP 10 for use in displaying channels/feeds or file ontologies to user 50. For example, each content element (e.g., email) may have associated data that records the probability that it participates in each feed, sub-feed, sub-sub-feed, etc. In some implementations, this record may be truncated by OP 10 to only store the top categorization of the content, or the k most likely categorizations of the content. For example:
In some implementations, the second feed category may be a sub-feed of the first feed category. For instance, the second feed category may be a subcategory (or sub-subcategory, sub-feed, child feed, etc.) of the first feed category. For example, the first feed category may be labeled “Project 1” and the second feed category may be labeled a subcategory “Project 1.” In some implementations, the second feed category may be a non-compartmentalized category (e.g., parent feed). For instance, the second feed category may be entirely unrelated and therefore separate from the first feed category. For example, the first feed category may be labeled “Project 1” and the second feed category may be labeled “Project 2.”
In some implementations, as noted above, OP 10 may employ a labeling and/or tagging scheme that enables some or all of the identified 300 content to be organized by an ontology or ontologies, which may generally be referred to as labels. In some implementations, the ontology may use multiple labels per content. As noted above, the ontology does not need to be hierarchical or strictly hierarchical, although it may be if desired. In some implementations, OP 10 may enable the user, if desired, to manually adjust the tag hierarchy, if such a hierarchy exists. In some examples, the ontology may have a single content (e.g., email) appear in multiple categories if multiple labels are assigned to it. In other words, the ontology may be an “over-lapping” ontology. Thus, when user 50 views a collection of content that each has the same label, new incoming documents to which OP 10 applies the same label may automatically appear in the view of the associated feed category with the label. For example, OP 10 may enable a user to manually apply any set of tags to any email. For instance, an email from one's spouse requesting that one be home from work early may be tagged with both a tag “work schedule” and another tag “home schedule” simultaneously. The “home schedule” tag may be a sub-tag of the general “home” tag/category. The “work schedule” tag may be a sub-tag of the general “work” tag/category. In the example, when one brings up the “home” feed, this email may be visible. When one brings up the “home schedule” feed, this email may be visible. When one brings up the “work” feed, this email may be visible. When one brings up the “work schedule” feed, this email may be visible. In some implementations, a machine learning portion of OP 10 may apply or suggest these tags automatically, and may become more and more accurate at applying or suggesting these tags after receiving 306 feedback from the user as discussed below. Thus, it will be appreciated that particular ontology may be used (singly or in combination) without departing from the scope of the disclosure.
It will be appreciated that the concepts of ontology and channels may be used interchangeably with the present disclosure. That is, a feed category may be any type of organizational/categorizable technique of related content (e.g., one, two, or three dimensions using groupings such as spatial clusters, containers, graphs, folders, or other visual elements). In some implementations, when using “channels,” each content may have none, one, or more labels indicating that it belongs to no channels, one or many channels, respectively. In some implementations, each label (or some of the labels) may also include a probability indicating how likely that label is to be correctly assigned to specific content.
In some implementations, the above-noted labels may be produced by the machine learning portion of OP 10, and may be used to display the content in the above-noted folders (e.g., channels/feed category) and/or in a folder hierarchy, or other technique. In some implementations, OP 10 may enable a view such that content (e.g., emails) in the user's inbox are removed and immediately show up within folders in their email client application. In some implementations, these folders may be labeled as “channels” and there may be no folder hierarchy. In some implementations, the folders may be labeled according to a particular profession. For instance, assume for example purposes only that user 50 is an attorney. In the example, the folders may be labeled by, e.g., client (or other attorney related subject) (e.g., via OP 10), and each client folder may have sub-folders for each legal matter, project, or other activity pertaining to that client/project.
In some implementations, OP 10 may automatically remove the emails from the inbox to the appropriate feed category, or may not remove the emails from the inbox to the appropriate feed category until the user has an opportunity to view them and they are marked “read.” In some implementations, OP 10 may wait to “file” an email until the user has had a chance to respond or react to the email in some way. In some implementations, OP 10 may recognize when the user has read an email, but is not truly ready to file in the appropriate channel/feed category. For example, OP 10 may be able to recognize states of an email, such as when the user has responded to an email by replying something like, “let me look into this and get back to you tomorrow.” In some implementations, OP 10 may determine when the user is truly done with an email (or other type of message) using a “done estimator” via a naive Bayes model of word or n-gram frequencies. In some implementations, other states/modes of a typical work flow may be used. For instance, non-limiting examples may include, (1) notification/alert that one or more content(s) is highly relevant right now, (2) reading those one or more content(s), (3) drafting/creating a response regarding one or more content(s), (4) awaiting feedback on the draft from other people, (5) task completed, decision made, etc.
In some implementations, OP 10 may receive 306 user feedback to change the categorization of a first content of the first portion of the plurality of content in the first feed category to the second feed category. For instance, assume for example purposes only that user 50 does not fully like the categorization ontology OP 10 produced. In the example, further assume that an email was originally categorized/labeled into the first feed category using the first probabilistic model, but that user 50 has decided that the email would be better categorized/labeled into the second feed category. In some implementations, the user feedback may be received 306 via a user interface of a second computing device (e.g., client electronic device 50) and sent to computer 12. In the example, user 50 (e.g., via OP 10) may provide user feedback (e.g., feedback 17) to change the categorization of that email from the first feed category to the second feed category (e.g., by changing the label of the email, physically “drag and drop” the email into the second feed category, different folder(s), container(s), cluster(s), etc.), which may be received 306 via OP 10 (e.g., via network 14, client application 26, etc.).
In some implementations, one or more of the at least one latent variable and the at least one action variable determined using the machine learning procedure may be based upon, at least in part, user feedback received from the one or more users. For example, imagine that the machine learning procedure is implemented in the form of computer software running on a server that monitors emails and instant messenger messages among a team of software engineers. Suppose that the team is working on two projects called ‘website prototype’ and ‘final website’. The server (or other computing device) implementing the machine learning procedure may (via OP 10) monitor communications and determine that the ‘final website’ should be completed before the ‘website prototype’, however, members of the team of software engineers know that ‘website prototype’ should be completed before ‘final website’. One or more of the teams of engineers may provide feedback to the machine learning procedure in the form of a graphical user interface to show that the ‘website prototype’ should in fact be completed before ‘final website’. This may change the latent variables of ‘when’ each project should be completed .
In some implementations, the user feedback received 306 via the user interface may include a gesture. For instance, and referring at least to the example implementation of
It will be appreciated that any other types of gestures or user actions may be used for the user feedback without departing from the scope of the disclosure. For instance, swiping in any direction, “tapping” or “clicking” on a particular spot on the user interface, shaking, etc. may also be used as user feedback. As such, the specific example of swiping should be taken as example only and not to otherwise limit the scope of the disclosure.
In some implementations, the user interface of OP 10 may include a slider with, e.g., three example settings, “good label,” “neutral label,” and “bad label” to help user 50 provide the user feedback. In one example, these sliders may be displayed at the top of each content in the feed category. In some implementations, each of these sliders may start in the neutral position, and user 50 may (via OP 10) move it to “good” or “bad” states. In some implementations, there may be a neutral setting (e.g., in the middle of such a slider), and all of the contents may be neutral except for the top content, which may be initially on (with the slider having a dark grey background), or two or more such contents, if the channel was seeded by two or more contents. In the example, if the user slides any lower email from neutral to on, OP 10 may turn the content dark grey and may snap to the top under the other dark ones. In the example, if the user slides the slider to off, the content may fade away, the other contents may snap up to fill in its space, and (eventually) a small undo message may appear at the bottom of the user interface.
In some implementations, the machine learning process may include a second probabilistic model generated by modifying a first probabilistic model, the modification based upon, at least in part, inferences of one or more of the at least one action variable the at least one latent variable. For example, imagine a group of people designing a building. Furthermore, imagine that the machine learning procedure is a computer program running an email server that monitors email communications among members of the groups. Within the larger group there are a sub-group of engineers designing the plumbing and a sub-group of engineers designing the electrical wiring. The machine learning procedure may suggest all communications sent between engineers in the plumbing sub-group also be sent to the electrical sub-group. One or more engineers in the electrical sub-group could determine that these communications should not be sent to them and they give feedback to computer software on the email server that the computer software should only suggest that emails be sent to members within their respective sub-group. The machine learning procedure (via OP 10) may then infer that the original model was imperfect and may produce a modified version of itself that may be much less likely to suggest that emails sent between members of a sub-group be sent to members of another subgroup.
In some implementations, OP 10 may generate 308 a second probabilistic model based upon, at least in part, the user feedback. For instance, and continuing with the above example where user 50 (e.g., via OP 10) provides user feedback 17 to change the categorization of an email previously categorized 302 by OP 10 in the first feed category, e.g., to the second feed category, (e.g., by changing the label of the email, physically “drag and drop” the email into the second feed category, different folder(s), container(s), cluster(s), etc.), which may be received 306 via OP 10 (e.g., via network 14). In the example, OP 10 may generate 308 a second probabilistic model using user feedback 17, where (as noted above) at least one of the first probabilistic model and the second probabilistic model may be generated 308 via machine learning. In some implementations, OP 10 may use Bayesian probabilistic models, as described in the Gantt chart description above. Generally, the received 306 user feedback may be used to “condition” any variable or parameter for this probabilistic model. In some implementations, OP 10 may leverage user feedback 17 in order to improve the estimate of what content ontologies/channels/feeds are desired by user 50 and/or to improve an existing channel/feed category.
In some implementations, OP 10 may reorganize 310 the categorization of a second content of the first portion of the plurality of content in the first feed category based upon, at least in part, the second probabilistic model. For instance, and continuing with the above example where user 50 (e.g., via OP 10) provides user feedback 17 to change the categorization of an email previously categorized 302 by OP 10 in the first feed category to the second feed category (e.g., by changing the label of the email, physically “drag and drop” the email into the second feed category, different folder(s), container(s), cluster(s), etc.), which may be received 306 via OP 10 (e.g., via network 14). In the example, OP 10 may review some or all of the plurality of content that was originally categorized 302 according to the first probabilistic model, and may reorganize 310 (e.g., recategorize) that content according to the second probabilistic model generated based upon user feedback 17. For instance, assume for example purposes only that a different email (e.g., email “A”) was labeled/tagged and categorized 302 to be placed in the first feed category according to the first probabilistic model. In the example, based upon user feedback 17, OP 10 may use the second probabilistic model to determine that email “A” should now be categorized to be placed in the second feed category, since it may now have a new/updated label assigned by OP 10 based upon the second probabilistic model. Thus, in the example, user feedback 17 may be used by OP 10 to further refine its labeling/categorization for future content according to the second probabilistic model, but may also be used to reorganize 310 the content previously categorized according to the first probabilistic model (e.g., before the user feedback was received to generate the second probabilistic model). It will be appreciated that user feedback need not always move a content from a first feed category to a second feed category. For instance, in some implementations, reorganizing 310 the categorization of the second content of the first portion of the plurality of content in the first feed category may include removing 314 the second content from the first feed category. In the example, OP 10 may remove it from a feed (e.g., the first feed category) and OP 10 may then determine what to do with it (e.g., using one or more future probabilistic models).
In some implementations, OP 10 may generate/refine a new probabilistic model each time user feedback is received, making the above-noted reorganization an iterative process. In some implementations, OP 10 may generate a message (e.g., a pop-up message) asking if the user wants to generate/refine the probabilistic model and/or have the new probabilistic model applied to the existing and/or new content.
In some implementations, receiving 306 the user feedback may include receiving 312 user feedback from a plurality of users. For instance, assume for example purposes only that two users (e.g., user 50 and user 38) both have access to the same feed categories or content. In the example, further assume that it is user 38, and not user 50, that provides the user feedback to change the categorization of that email from the first feed category to the second feed category. In the example, OP 10 may similarly change the categorization of that email for the feed categories of user 50 as well as user 38. In some implementations, shared access may not be required to receive 312 user feedback from a plurality of users. For example, assume that user 50 and user 48 do not know each other nor do they share information with one another. Further assume they both follow the Red Sox and Celtics. Further assume that user 48 makes one channel for updates about the Red Sox and another about the Celtics (or puts these two teams in two separate sub-feeds under their “sports” feed). In the example, OP 10 may notice that user 50 also follows both teams. OP 10 may automatically suggest to user 50 that he too may like to have separate channels for the two sports teams under a main “sports” feed. User 50 may manually reorganize this if he likes, providing additional new feedback to OP 10. As such, the example of only a single user providing user feedback to generate the second probabilistic model should be taken as example only and not to otherwise limit the scope of the disclosure.
In some implementations, the above-noted labels assigned to content according to the probabilistic model may be delivered, served, shared, or otherwise made available so that they may be used by other applications. For example, OP 10 may enable the display of emails according to tags as noted above. OP 10 may apply the tags it learns within the email client, so that if a user logs into the email client user interface to look at only their email, the same tags may be present. As another example, users in a project may share to other users assigned to the same project their tags/ontology for how they organize their sub-projects and sub-sub-projects and so forth.
In some implementations, the labels may be made available, not to just anyone, but may be made available with restricted access. For instance, the learned Gantt chart organization may be exported to a project planning tool so that the team may view visually the emergent machine inferred organization of their project. However, assume for example purposes only that one does not want people outside one's company understanding the overall project with this kind of global view, or one only wants a few managers of the project to have this overall view. In the example, OP 10 may enable the user to only share the tags and overall inferred project organization to those who are authorized to have access to them (e.g., using known authorization techniques). As another example, assume that Human Resources is using the labels/tags to understand the emergent behavior and communications patterns in the company. For example, they may want to know that a certain person improves the probability of success of any given project with which they interact, and therefore they deserve a salary raise. This may be confidential information for human resources.
In some implementations, OP 10 may include a user interface for selecting one or more contents to be converted into a task item, in which case OP 10 may use one of the above-noted learning algorithms to extract information from the content and populate structured fields for the task, such as, e.g., task name, priority, requester, followers, owner, due date, duration, effort level, task type, links to other relevant documents, etc. In some implementations, OP 10 may enable the structured data with the above information to be output as a .CSV or other file format, which may then be ingested or displayed in a spreadsheet, CRM, workflow or task management tool, or other tabular or database system.
It will be appreciated that OP 10 may be used for other purposes without departing from the scope of the disclosure. For instance, OP 10 may be used for auto time carding. For example, OP 10 may analyze content (e.g., emails, documents, and edit logs), and may use this information to determine when/how long a particular user was working on various content and/or the various projects associated with the content. OP 10 may also infer from this information an estimate of how many words per minute a person produces when working on a document or other content. Given that OP 10 may cluster content, OP 10 may therefore infer how much time and which times the user spent on each activity. For instance, the activities may be legal matters and the user(s) may be lawyer(s), or the activities may be engineering projects and the user(s) may be engineer(s), etc. In some implementations, OP 10 may use the inferences to auto-populate or auto-suggest a time-card. For instance, a user may work on a given project with 60% of their effort on Monday and 20% of their effort on Tuesday, and no further effort on the other days of the week. In the example, OP 10 may recognize this and pre-populate their time-card with that information. In some implementations, this auto- populated time-card may then be reviewed and possibly edited by the user (or administrator) as a detailed report or prefilled entries for final entry into a billing system or time tracking system.
In some implementations, OP 10 may present 320b potential collaboration actions to the one or more users based upon, at least in part, the at least one action variable. For example, in some implementations, OP 10 may perform inference to learn rules for filing content. For instance, as noted above, these rules may be probabilistic in nature, which may be referred to as automated induction of probabilistic programs. An example of a probabilistic filing rule may be, e.g., when a document is from <xx>, then 60% of the time that document receives the <yy> label. In some implementations, OP 10 may choose the most relevant and/or impactful time to display or send particular information to the user(s). For example, OP 10 may decide to turn on the throb that calls someone's attention to a particular feed at a particular time. In the example, assume they are working on a provisional patent application for Company X in one feed, and they have a second feed for a second Company X patent application. Further assume that these two patent applications are both under an omnibus feed for Company X. If the person is actively reading feed 1, and a new message comes in on feed 2, then OP 10 may make the feed 2 indicator throb because it knows the user is actively thinking about Company X patent applications in feed 1, and this is likely not a distraction and may be important and relevant to what the user is are doing.
In some implementations, OP 10 may (e.g., via the machine learning procedure) perform action(s) and user(s) may also perform action(s). Generally, action variables are about actuation. Examples of action variables may include, but are not limited to, whether or not the system should send a message, when it should send a message, whether and when it should push a document to user(s), whether and when the user(s) should follow up with a particular client, which tasks should be prioritized, among many others. The action variables are storage of whether or not an action should be performed. For example, the machine learning procedure (via OP 10) may store the probability that the action of recommending a user follow up with a particular client. The machine learning system, based on the probability it should recommend a follow up, may (via OP 10) then perform the action of sending a message to recommend the follow up.
For example, OP 10 may automatically learn that some members of a working group prefer to get email messages, e.g., during the morning and may delay sending them, e.g., during the evening. This may be accomplished by recording when users are most likely to open work related emails. One member of the working group may not open emails during the evening, thus OP 10 may then conclude that receiving messages in the evening is not this user's preference. A different member of the working group may work productively in the late evening and thus OP 10 would not delay evening messages intended for this different member of the working group.
In some implementations, OP 10 may automatically identify particularly important messages by pushing them to the team members and labelling them as urgent (or other appropriate label), while learning that other messages are of less importance and labelling them as low importance (or other appropriate label). For example, imagine that a group of workers is working on the drafting of a patent application that is due very soon (based upon one or more of the above-noted variables). OP 10 may conclude that all messages sent between members of the working group be labelled as urgent.
In some implementations, OP 10 may suggest recipients to add to messages based on previous communication patterns. This may be done by OP 10 watching the streams of information with or without directional involvement of users. For example, imagine a group of software engineers working on a ‘website redesign’ project and that one member of the group has left to join a different project. Further imagine that this group primarily uses email to communicate. OP 10 may recognize that the group member who left is no longer sending emails to or from the ‘website redesign’ group and thus no longer suggests to the ‘website redesign’ group that this group member be included on future emails.
Unlike existing technological processes, such as those provided by enterprise search approaches that include keyword search query capability for information stored across an organization, OP 10 may also incorporate machine learning to take the initiative to recommend information to the user(s) at any time, not only when the user(s) asks for it via the keyword search query.
Furthermore, unlike known keyword matching technological processes, OP 10 may calculate the relevance of a document (or other content) to a user(s), at least in part, by what collaborations and/or relationships the user(s) are working on and/or by a combination that may include inference about who is working on what with whom, when, a statistical model of collaborations and/or relationships, inference of latent and/or action variables in a relationship or collaboration, temporal dynamics of collaborations or relationships, inference and modeling of semantic language structures relating to ideas that are being shared in a collaboration or relationship, among many others.
In some implementations, OP 10 may present 320a at least one of a representation of the collaboration items to one or more users based upon, at least in part, the at least one latent variable. Thus, by inferring latent and/or action variables of collaboration, OP 10 may choose which information to display to the user at a particular time and/or place in order to maximize relevance and/or optimize for a particular user interface, such as a smaller screen. For example, the collaboration items to be presented to the user could determine that three users were collaboration on a project and suggest that emails sent between any two users in the collaboration be sent to the third user as well.
In some implementations, the machine learning system (via OP 10) may suggest improvements for a collaboration and/or relationship. For example, the machine learning system (via OP 10) may suggest to a manager that they send messages to check in with a staff member if they have not communicated for a long time.
In some implementations, the machine learning system (via OP 10) may suggest other people to include in a collaboration and/or relationship. For example, imagine a very large organization with two website related projects titled ‘website browse page redesign’ and ‘website purchase page redesign’. Further, imagine that there are two working groups for these projects. The machine learning system (via OP 10) may identify that these projects are very related in their technical content by observing the words used in email communications among the groups separately. The machine learning system (via OP 10) may then suggest that communications among one group be routed to the other because the members of the two groups are solving the same technical challenges and can learn from each other to better solve their separate projects.
In some implementations, the machine learning system (via OP 10) may suggest scheduling a meeting as part of a collaboration and/or relationship. For example, image a salesperson with a list of sales prospects. The machine learning system (via OP 10) may maintain a list of the sales prospects and further store the fact that each of these sales prospects should be contacted on a regular basis. The machine learning system (via OP 10) may keep track of the emails sent by the salesperson and recommend sales contacts to be emailed if the salesperson has not communicated with them recently.
In some implementations, the machine learning system (via OP 10) may suggest scheduling a meeting as part of a collaboration and/or relationship when it notices that messaging within a group are increasing in frequency, and may infer that a real-time synchronous meeting (e.g., in-person, over video and/or audio chat, or using instant messaging, etc.) could be useful. For example, imagine a salesman and sales prospect communicating very frequently over email about a possible sale. The machine learning procedure (via OP 10) may then recommend an in-person meeting or video chat to help the salesperson and sales prospect communicate more information more quickly.
In some implementations, the machine learning system (via OP 10) may suggest sending a message as part of a collaboration and/or relationship. For example, imagine a group of coworkers working on different components of the same project. The machine learning system may monitor the frequency of communication between coworkers and may remind coworkers to occasionally check in on the progress of each other by suggesting a quick message such as ‘How is the project going’ so as to keep the team synchronized.
In some implementations, the machine learning system (via OP 10) may suggest sending a thank you message as part of a collaboration and/or relationship. For example, imagine a salesperson who has just had an in-person meeting with a sales prospect. The machine learning system (via OP 10) may monitor the communications of the sales person and the sales prospect and determine that words representative of a thank you message like ‘Thank you’ and ‘Great to see you in person’ were not mentioned in the communication following the meeting and, thus, suggest thank you message be sent from the salesperson to the sales prospect.
In some implementations, the machine learning system (via OP 10) may choose the most relevant and/or impactful information to display or send to the user(s). For example, imagine an email sent with dozens of attached files sent to a member of a working group. Further imagine that this email is sent repeatedly and contains the same or similar type of email attachments. Further imagine that this working group member only downloads the images attached to the email and none of the other types of attachment. The machine learning system (via OP 10) may record that images are the only relevant email attachments for this user and emphasize the images by placing them at the top of the list of the possible downloads.
In some implementations, OP 10 may be used across multiple companies or user teams to analyze multilateral projects and tasks. For example, imagine a project between two construction companies for building a road. One company digs the ditches for the road and the other company pours the asphalt for the road. The overall project of building the road is shared across the companies and the machine learning system (via OP 10) may perform as if the project was contained entirely within one company.
In some implementations, OP 10 may suggest communication and/or connect with relevant domain experts. For example, imagine a large company that may contain a biologist and a team of salespeople without biology training. Further imagine that the salespeople are trying to sell their product to biologists at a biotech company. The machine learning system (via OP 10) may recognize that terms like ‘biotech’ and ‘biology’ are mentioned in the emails between the salespeople and the biologists within the biotech company. The machine learning system (via OP 10) may recommend that the sales people contact the biologist within their own company to help them understand the types of problems the biologists within the biotech company are trying to solve.
In some implementations, OP 10 may suggest to domain expert user(s) conversations and/or communications to join. For example, imagine a large company that contains a biologist and a team of salespeople without biology training. Further imagine that the salespeople are trying to sell their product to biologists at a biotech company. The machine learning system (via OP 10) may recognize that terms like ‘biotech’ and ‘biology’ are mentioned in the emails between the salespeople and the biologists within the biotech company. The machine learning system (via OP 10) may recommend to the biologist that they begin communications with the salespeople.
In some implementations, the machine learning system (via OP 10) may incorporate a model of tasks and/or steps that may occur during collaboration(s) and/or relationship(s). For example, imagine a company that has a sales pipeline that consist of pre-sales, post-sales, and long term relationship management. The machine learning system (via OP 10) may contain a model that explicitly expects all messages between salespeople and sales prospects being related to one or more of the pre-sales, post-sales, and long term relationship management categories.
In some implementations, the machine learning system (via OP 10) may infer the state of variables about the user(s)'s actions, behaviors, activities, beliefs or plans. For example, imagine a member of a workgroup planning on leaving a collaboration to start a new collaboration. The user may send fewer emails to their original collaborators and more emails to their new collaborators. The machine learning system (via OP 10) may infer from the frequency of emails and their recipients that the workgroup member plans on leaving the collaboration to start a new one.
In some implementations, the machine learning system (via OP 10) may model the connection between latent states and natural language utterances. For example, the machine learning system (via OP 10) may associate the latent state of a project as being very important by looking for words such as ‘important project’, ‘needs to get done’, and ‘critical’ in communications relevant to the project. The machine learning model (via OP 10) may further look at statistics such as how frequently pairs or other combinations of words occur and their order to determine latent states of the project.
In some implementations, the machine learning system may include a model of ways that user(s) may behave within a collaboration(s) and/or relationship(s). For example, the model may explicitly contain the concept of manager and concept of employee and note that managers are usually hubs of communications. For instance, a workgroup that tends to send the most communications to a particular individual, and that individual frequently responds to communications, may then be indicative to OP 10 that the individual is possibly the manager of the project.
In some implementations, OP 10 (e.g., via the machine learning) may leverage its ability to infer latent and/or action variables in a collaboration and/or relationship in order to better translate a natural language query into a database query. Thus, unlike known technological processes that merely convert natural language queries into database queries, OP 10 may leverage domain expertise about projects, collaborations, and/or relationships. For example, by knowing which relationship(s) and/or project(s) the user(s) is actively working on, OP 10 may query only (or mostly) for tasks within those projects and/or relationships. As another example, by knowing which other people the user(s) is relating to or collaborating with, OP 10 may return information relating only (or mostly) to the tasks of those people. As yet another example, by knowing which deadlines or next steps are in the critical path or are coming due soon, OP 10 may return information from the database relating only (or mostly) to tasks relating to those deadlines or next steps.
In some implementations, OP 10 may be used to share links within a group. For instance, users in a group of users may browse the web and find useful web pages. They may share these links in some way, for example, by sharing their entire browser history with OP 10, by sharing their browser bookmarks with OP 10, by indicating to OP 10 that they want to share a page, etc. OP 10 may then analyze the text and metadata of these web sites similarly to how OP 10 may analyze emails or other documents and apply labels to each of the web pages. In some implementations, these web pages or links to these web pages may become documents viewable in channels or other ontologies, or within the user's email browser or other browser.
In some implementations, the content labels from OP 10 may be used to better understand which projects are active, what kinds of communication is happening in them, who is working on which ones, and when is activity happening within each project. For example, at the top of a feed, the user interface may have icons for each of the people who are presently reading or working within that feed. As another example, the feed information may be exported to a visualize tool along the lines of a project management tool, where a manager may see who was working on which projects at what times and for how many hours per day. In some implementations, the projects may be a sales process, and OP 10 may provide information for enhancing management of the sales process and/or collaboration and/or communication in the sales process. In other examples, the process may be an engineering development or other kind of business process.
It will be appreciated that there may be various ways to view/display content, access content, provide user feedback for content, etc. according to the above disclosure. The various example and non-limiting views, access controls, displays, colors, layout, etc. are shown via example implementations
The terminology used herein is for the purpose of describing particular implementations only and is not intended to be limiting of the disclosure. As used herein, the singular forms “a”, “an” and “the” are intended to include the plural forms as well, unless the context clearly indicates otherwise. As used herein, the language “at least one of A, B, and C” (and the like) should be interpreted as covering only A, only B, only C, or any combination of the three, unless the context clearly indicates otherwise. It will be further understood that the terms “comprises” and/or “comprising,” when used in this specification, specify the presence of stated features, integers, steps (not necessarily in a particular order), operations, elements, and/or components, but do not preclude the presence or addition of one or more other features, integers, steps (not necessarily in a particular order), operations, elements, components, and/or groups thereof
The corresponding structures, materials, acts, and equivalents (e.g., of all means or step plus function elements) that may be in the claims below are intended to include any structure, material, or act for performing the function in combination with other claimed elements as specifically claimed. The description of the present disclosure has been presented for purposes of illustration and description, but is not intended to be exhaustive or limited to the disclosure in the form disclosed. Many modifications, variations, substitutions, and any combinations thereof will be apparent to those of ordinary skill in the art without departing from the scope and spirit of the disclosure. The implementation(s) were chosen and described in order to explain the principles of the disclosure and the practical application, and to enable others of ordinary skill in the art to understand the disclosure for various implementation(s) with various modifications and/or any combinations of implementation(s) as are suited to the particular use contemplated.
Having thus described the disclosure of the present application in detail and by reference to implementation(s) thereof, it will be apparent that modifications, variations, and any combinations of implementation(s) (including any modifications, variations, substitutions, and combinations thereof) are possible without departing from the scope of the disclosure defined in the appended claims.
This application is a continuation-in-part of U.S. application Ser. No. 15/062,688, filed Mar. 7, 2016, which claims the benefit of U.S. Provisional Application No. 62/128,671 filed Mar. 5, 2015, the contents of which are incorporated herein by reference. This application is also a continuation-in-part of U.S. application No. 15/624,012, filed on Jun. 15, 2017, which claims the benefit of U.S. Provisional Application No. 62/350,440, filed on Jun. 15, 2016, the contents of which are all incorporated by reference.
| Number | Date | Country | |
|---|---|---|---|
| 62128671 | Mar 2015 | US | |
| 62350440 | Jun 2016 | US |
| Number | Date | Country | |
|---|---|---|---|
| Parent | 15062688 | Mar 2016 | US |
| Child | 15843904 | US | |
| Parent | 15624012 | Jun 2017 | US |
| Child | 15062688 | US |
