RU2495485C1 - Method of monitoring user preference information based on analysis of content perception of said users (versions) - Google Patents

Abstract

FIELD: information technology.

SUBSTANCE: user authorisation is performed on a server. The server provides the user with a list of content available for use. The user selects content from the list while sending information on the selected content to the server, and also uses the content while recording evaluation of emotional experience of the user by entering one of proposed evaluations on a recording device. The recording device sends the evaluations to the server, which processes the evaluation information received from the users, generates information models of user preferences as well as information time models of content. When the user uses the content, the emotional state of the user is automatically monitored using an emotion monitoring means installed on the recording device and evaluations of the emotional experience of the user is determined from content usage.

EFFECT: high efficiency of monitoring user preference information based on analysis of content perception of the user.

15 cl

Description

The group of inventions relates to information technology and can be used to monitor information about user preferences based on an analysis of their perception of content. There are two options that can be divided into two groups: monitoring information about user preferences using the user's independent assessment of their emotional experiences and monitoring information about user preferences with automatic hardware-software assessment of emotional experiences of the user.

Content is any content perceived by a person. Content can be distributed through a specific medium, such as the Internet or television. The proposed method does not depend on the specific type of content delivery and can be used with arbitrary content.

There are known methods for obtaining information about user preferences based on a generalized assessment of Internet content using the “Like / Dislike” vote or various ratings (see the “Like” widget in the Contact page,

http://vk.com/developers.php?p=Like ;

Facebook Like Button, http://developers.facebook.com/docs/reference/plugins/like/ ;

Youtube API v2.0 - Ratings,

https://developers.qooqle.com/voutube/2.0/developers_guide_protocol ratings ). In these methods, all consumers of content are encouraged to evaluate it. Further analysis of the data obtained allows us to determine the popularity of the content and draw conclusions about user preferences. However, in these methods there is no reference to the time of the content, that is, the user himself generates an average value judgment in relation to the entire content. Moreover, a lot of evaluated content is limited to content accessible via the Internet.

In addition, thanks to the development of mobile communications, it became possible to use televoting (including SMS voting) in various interactive programs, surveys, quizzes, contests, as well as for determining ratings (see Televoting, http://en.wikipedia.org / wiki / Televoting ). To express your preference and vote for a particular candidate, just call or send a CMC message to the corresponding phone number. Based on all sent messages, the winner is determined. CMC voting is not tied to a specific type of content and can be used to evaluate events that occur in real life, but, as in the case of Internet voting, the user forms an average value judgment in relation to all content.

There is also a known method of monitoring information about the preferences of a TV audience using specialized equipment, the so-called People meters, which are installed on televisions of a certain representative sample of people and record their viewing of television programs (see Rating Technique, http: //www.grp. com / effectiv / ; People Meter, http://en.wikipedia.org/wiki/People_meter ). The received data is accumulated and transmitted to the information processing center. Based on these data, the preferences of various target audiences are determined and the ratings of programs and channels are compiled. The method requires the installation of additional equipment and is applicable only to TV channels. In addition, when using it, only quantitative characteristics of consumption are recorded and then analyzed — who watched which channels at certain points in time.

Thus, the known methods are not universal. They use the average value judgment of the content consumer in relation to the entire content as a whole. Such an average estimate is rather rough and seriously limits the ability to obtain accurate information about user preferences. In addition, some methods require the installation of additional special equipment or are applicable only to a certain type of content.

The objective of the present invention is to expand the arsenal of tools, namely the creation of a new method (service) for monitoring information about user preferences based on the analysis of their perception of content that eliminates the above disadvantages.

To solve the problem, according to the first embodiment of the present invention, user registration devices and a server are used. The server stores information about users, including identification data, and informational temporary content models. Server software allows you to create information models of user preferences based on processing information about users' ratings of their emotional experiences while consuming content. At the first stage, the user is authorized on the server; providing the server with this user a list of content available for consumption and user selection of content from the list with sending information about the selected content to the server. After that, the user consumes the content with fixing the ratings of his emotional experiences by entering one of the proposed ratings on the recording device. Recorded scores are sent through the recording device to the server. The server aggregates information about ratings received from users and generates information models of user preferences based on processing information about users, information time models of content and aggregated information about recorded ratings.

The standard implementation of the recording device is an electronic device on which a consumer agent is installed, containing a user interface for entering estimates of emotional experiences of the user.

Such a recording device may be a mobile phone on which a consumer agent is installed as an application for the telephone operating system.

The recording device may also be a personal computer. A personal computer can be both stationary and mobile (laptop, tablet network computer). Also, personal computers include a smartphone, personal digital assistant (PDA) or an e-book. A consumer agent may be an application for the operating system of this device. In addition, other implementations of a consumer agent for a personal computer are possible.

If the personal computer contains a web browser, the consumer agent can be implemented as a web application. The user accesses the consumer agent through a web browser.

If the personal computer contains a software streaming video player that can be expanded using plug-ins, the consumer agent is implemented as the corresponding plug-in.

If the personal computer contains a BD-Live software Blu-ray player, then the consumer agent is implemented as the corresponding BD-J application.

It is possible to apply this method to a digital television receiver or digital set-top box with support for Internet services and widgets. In this case, the consumer agent is implemented as a corresponding widget, and the user interface uses a remote control to enter estimates of emotional experiences.

Also, a Blu-ray player with BD-Live support can be a recording device. In this case, the consumer agent is implemented as the corresponding BD-J application.

In all of the aforementioned consumer agent implementations, the user’s emotional experience scores are captured using manual data entry. But there are also other possible ways of fixing estimates of the emotional experiences of the user. For example, if the electronic device is a device that supports voice recognition, then the user interface uses voice commands to enter estimates of emotional experiences. Similarly, in the case of a device with support for recognizing facial expressions and / or body movements, the user interface uses gesture commands to enter estimates of emotional experiences.

Thus, the first embodiment of the present invention describes a method for monitoring information about user preferences with the active input of user ratings of their emotional experiences using a consumer agent.

The second embodiment of the invention basically repeats the first, it also uses user registration devices and a server. The server stores information about users, including identification data, and informational temporary content models. Server software allows you to create information models of user preferences based on the processing of automatically obtained information about ratings by content users. At the first stage, the user is authorized on the server, the server provides this user with information about the content available for consumption, and the user selects the content with the information about the selected content sent to the server. In contrast to the first embodiment of the invention, in the second embodiment, during a user consumes content, they automatically monitor their emotional state using emotional monitoring tools installed on a recording device and automatically record estimates of the user's emotional experiences from consuming content. Recorded scores are sent through the recording device to the server. The server aggregates the received information and creates information models of user preferences based on processing information about users, information time models of content, and aggregated information about recorded ratings.

The emotional monitoring tool may be a tool capable of recognizing and analyzing facial expressions and / or body movements of a user.

Also, the emotional monitoring tool can be a tool capable of recognizing and analyzing the words spoken by the user and emotionally colored sounds (for example, “Oh!”, “Ah!” Or “Mmm ...”).

The informational temporary content model describes the content and contains the following data structures:

1. Identification data;

2. Metadata.

Identification data includes: the name of the content, the type of content and the method of its delivery to the consumer, the duration of the content. In addition, the identity contains a pointer to the “location” of the content. For example, for television content, the index contains the number of the television channel and the planned start time for broadcasting this content on this channel. For Internet content, the index contains the URL (single index of resources on the Internet) of the content.

Metadata consists of time zones and meta-information associated with these zones, which is of interest to the provider or producer of the content. An example of such information may be semantic meta-information, which gives a semantic description of the corresponding piece of content. Time zones can have both zero and positive duration. In case of zero duration, the time zone is considered as a momentary event within the content. The number of time zones is not limited - they can intersect or overlap each other. Information associated with time zones may be of a different nature; for example, it can be a set of keywords. For example, in the case of a video, the supplier may mark out fragments of the film where hidden advertising is shown. Thus, metadata can be represented as a set of matches:

Time Zone ета Meta Information

The user's attitude to the content is determined by the assessment of the emotional experiences that he experiences during his consumption. Different sets of emotional experience ratings may be used. The choice of a specific set of acceptable estimates depends on various factors, including the type of content consumed, the simplicity and accuracy of comparing the assessment of emotional experiences with user emotions, etc. For example, a set of acceptable assessments of emotional experiences may include: “pleasure”, “desire”, “suffering”, “fear”.

User emotions excited at certain points in the content are related to the time of the content. For example, the user's emotional experiences can be rated as “fear” at the 10th second of the video or from the 15th minute to the 16th minute. Thus, the user’s attitude to the content can be represented as a correspondence:

Content time ↔ Assessment of user emotional experiences

When users consume content, they fix estimates of their emotional experiences or there is an automatic hardware-software fixation of estimates of the user's emotional experiences from consuming content and sending the recorded ratings to the server. The user may not be limited in the frequency of entering estimates of emotional experiences that can be recorded when consuming one content.

Aggregated ratings (all ratings received and stored on the server) of the user's emotional experiences are tied to time and content. Thanks to this, meta-information of the content and data reflecting the estimates of the emotional experiences of the user have a common reference to the time of the content. This circumstance allows the server to unambiguously for each consumer of content compare his assessment of emotional experiences with meta-information:

Assessment of emotional experiences of the user ↔

Meta information

Thus, for each content and the user consuming it, there are many pairs (assessment of the emotional experiences of the user, meta-information of the content). The combination of such pairs for the entire set of registered content and many users forms an information model of user preferences.

In order to better demonstrate the distinguishing features of the invention, as examples without any limiting nature, the preferred embodiments of this method are described below. For the examples described below:

- a database is created on the server for storing information about users, including identification data, for example, user name and password, information time models of content, aggregated estimates of emotional experiences of the user and information models of user preferences;

- user data and informational temporary models of content are entered into the database.

Example 1

As a recording device, a mobile phone or smartphone is used. The user installs a consumer agent on it in the form of an application for the device’s operating system.

When the consumer agent starts, the user is authorized. It consists in sending the corresponding user information to the server. For example, this information may consist of a username and password. The server checks the sent data and sends the verification result to the recording device.

In case of successful authorization of the user, the server determines the list of content available for consumption by this user and sends it to the user's registration device. This list of content is formed on the basis of various information, including the current time, restrictions (for example, age prohibitions - “for children under 16”) and user preferences if there is a history of calls, etc. For example, a list of content may include a repertoire of films shown in a particular movie theater.

The list of content available for consumption is displayed as a list on the screen of a mobile phone or smartphone and the user selects content for consumption from it. After selecting the content, the consumer agent displays the content consumption interface on the device screen. This interface allows the user to do the following:

1. Mark the moment of the start of consumption of content;

2. Mark the end of the consumption of content;

3. Enter estimates of their emotional experiences.

If this device contains a built-in keyboard, then it can be used in the content consumption interface to select one of the above actions. For example, the set of proposed evaluations of emotional experiences may consist of the following ratings: “pleasure”, “desire”, “suffering”, “fear”. In this case, the pressed “1” button may correspond to the input of the “pleasure” rating, etc.

If the device contains a touch screen, then it can also be used in the content consumption interface to select one of the above actions.

This consumer agent supports any type of content and is independent of its distribution channel. He is a universal agent.

At the moment the content consumption starts, the user selects the appropriate action (“Mark the moment the content consumption starts”) using the content consumption interface. According to this action, the user agent sends information about previously selected content to the server, which records the start time of consumption of this content by this user.

During the consumption of content, the user enters one of the proposed ratings on the mobile phone or smartphone using the appropriate action (“Enter estimates of their emotional experiences”) of the content consumption interface. The consumer agent records the entered estimates of the user's emotional experiences and sends them to the server. It is worth noting that the consumer agent can either send each rating received to the server immediately after user input, or accumulate several ratings and then send a group of accumulated ratings to the server.

At the end (or pause) of content consumption, the user selects the appropriate action ("Mark the moment the content is consumed") using the content consumption interface. According to this action, the user agent sends the corresponding information to the server, which records the end time of consumption of this content by this user.

For example, the user selected some content and introduced the following assessments of their emotional experiences when consuming this content: “suffering” at 14 seconds of content, “desire” at 62 seconds.

Moreover, the metadata of the information time model of this content contains the following markup: the time interval (15 s, 20 s) corresponds to the keyword (meta-information) “smoking”, and the time interval (53 s, 67 s) corresponds to the keyword “Eiffel Tower”.

As a result of the analysis of the data presented (corresponding overlay on the timeline of aggregated information about the recorded ratings and metadata of the information time model of this content), the server receives the following elements of the information model of user preferences related to this user: (meta-information “smoking” → assessment of emotional experiences “suffering” ), (meta-information “Eiffel Tower” → assessment of emotional experiences “desire”).

Example 2

The example as a whole repeats the previous one. The difference lies in the use of another recording device. A digital television receiver or digital television set-top box with support for Internet services and widgets is used as a recording device. This device is also used to play content.

In this case, the corresponding TV widget will be the consumer agent. It is worth mentioning that such a consumer agent is integrated. In contrast to the universal agent from Example 1, it integrates with the content reproduction tool and therefore depends on the content and its distribution channel. Thanks to closer integration with content when using such an agent, there is no need to select content - it will be selected automatically. On the other hand, for the same reason, the agent is not applicable to another type of content. Other examples of an integrated agent are a plug-in for a streaming video player, a BD-J application, etc.

A digital television receiver remote control is used to control the user agent and enter data. It is used similarly to the built-in keyboard of the device from Example 1 (button → assessment of emotional experiences): pressing a button on the remote control corresponds to entering from the remote control the assessment of emotional experiences corresponding to it.

In this case, there is no need to specifically mark the start or end of the consumption of content. Turning on the digital television receiver or switching it to a certain channel means the beginning of consumption of this channel, and vice versa - turning off the digital television receiver or switching it to another channel means the end of consumption of this channel.

Example 3

An example illustrates a second embodiment of the invention (with automatic monitoring of information about user preferences using emotional monitoring tools).

Example 3 repeats basically Example 1. As a device capable of recognizing the movements of a user's body and its facial expressions, a Microsoft Xbox 360 game console with a connected Microsoft Kinect touch device, which is equipped with depth sensors, a digital video camera and a microphone, is used (see Xbox 360 + Kinect, http: //www.xbox.com/kinect ).

The emotional monitoring tool is implemented as an application for the Microsoft Xbox 360. Monitoring the emotional state of the user is carried out on the basis of data received from the Microsoft Kinect touch device, aimed at the consumer of the content. For example, a user's smile, fixed and recognized as an emotion, can mean his pleasure in consuming content. As a result, estimates of emotional experiences of the user are automatically received and recorded, which are sent to the server for further construction (similar to example 1) of the information model of user preferences.

Since the Microsoft Kinect touch device contains a microphone, an emotional monitoring tool that can recognize and analyze user-spoken words and emotionally colored sounds is likewise implemented.

Also, the Microsoft Xbox 360 game console with a connected Microsoft Kinect touch device can be used as a recording device with support for voice or gesture commands for entering assessments of emotional experiences.

Similar examples can be described with respect to a Blu-ray player with support for BD-Live, a personal computer, including a desktop computer, or a laptop, or a tablet network computer, or a smartphone, or a personal digital assistant (PDA), or an e-book.

Information models of user preferences can be used to obtain various information, for example, information on the attitude of a particular user to smoking or information on the attitude of a certain group of users (for example, men with higher education, 35-50 years old) to products of any brand.

Thus, the proposed method, both in the first and in the second embodiment, performs a deeper analysis of user preferences based on detailed multi-level assessments of users' perception of content. In addition, it is not tied to a specific type of content and can be used with arbitrary content. The method does not require specially designed equipment and can be used with existing everyday user devices.

Claims (15)

1. A method for monitoring information about user preferences based on an analysis of their perception of content, characterized in that it uses user registration devices and a server that stores information about users, including identification data, is programmed to create information models of user preferences based on processing information about assessments by users of their emotional experiences while consuming content, the method includes the preliminary creation of information a temporary content model containing the content identification data and metadata, consisting of time zones and meta-information associated with these zones, and storing it on the server, the method also includes authorizing the user on the server; providing the server with this user a list of content available for consumption; user selecting content from the list with sending information about the selected content to the server; user consumption of content with fixing estimates of his emotional experiences by entering one of the proposed ratings on the recording device; sending through the recording device these estimates to the server; server aggregation of information about ratings received from users; formation by the server of information models of user preferences based on processing information about users, information time models of content and aggregated information about recorded ratings. 2. The method of monitoring information according to claim 1, characterized in that the said recording device is an electronic device on which a consumer agent is installed, containing a user interface for entering estimates of emotional experiences of the user. 3. The method of monitoring information according to claim 2, characterized in that said electronic device is a mobile phone, and the consumer agent is an application for the operating system of this phone. 4. The method of monitoring information according to claim 2, characterized in that the electronic device is a digital television receiver or digital television set-top box with support for Internet services and widgets, and the consumer agent is the corresponding TV widget. 5. The method of monitoring information according to claim 2, characterized in that said electronic device is a Blu-ray player with BD-Live support, and the consumer agent is a BD-J application. 6. The method of monitoring information according to claim 2, characterized in that said electronic device is a device with support for voice recognition, and the user interface uses voice commands to enter estimates of emotional experiences. 7. The method of monitoring information according to claim 2, characterized in that said electronic device is a device that supports recognition of facial expressions and / or body movements, and the user interface uses gesture commands to enter estimates of emotional experiences. 8. The method of monitoring information according to claim 2, characterized in that said electronic device is a personal computer, including a stationary computer, or laptop, or tablet network computer, or smartphone, or personal digital assistant (PDA), or an electronic book . 9. The method of monitoring information according to claim 8, characterized in that the consumer agent is an application for the operating system of a personal computer. 10. The method of monitoring information according to claim 8, characterized in that said personal computer contains a web browser, and the consumer agent is a web application. 11. The method of monitoring information according to claim 8, characterized in that said personal computer contains a streaming video player that is expandable using plug-ins, and the consumer agent is the corresponding plug-in for this player. 12. The method of monitoring information according to claim 8, characterized in that said personal computer contains a software Blu-ray player with BD-Live support, and the consumer agent is a BD-J application. 13. A method for monitoring information about user preferences based on an analysis of their perception of content, characterized in that it uses user registration devices and a server that stores information about users, including identification data, and is programmed to create information models of user preferences based on information processing about users evaluating their emotional experiences while consuming content, the method includes the preliminary creation of information th temporal model of content identification data comprising the content and metadata composed of time zones, and these zones associated with the meta information, and storing it on a server, the method also includes the user authentication on the server; providing the server with this user a list of content available for consumption; user selecting content from the list with sending information about the selected content to the server; user consumption of content; automatic monitoring of the emotional state of the user with the help of emotional monitoring installed on the recording device; automatic fixing of estimates of emotional experiences of the user; sending through the recording device these estimates to the server; Aggregation by the server of the received assessment information; formation by the server of information models of user preferences based on processing information about users, information time models of content and aggregated information about the said ratings. 14. The method of monitoring information according to item 13, characterized in that the said means of emotional monitoring is a tool capable of recognizing and analyzing facial expressions and / or body movements of the user. 15. The method of monitoring information according to item 13, wherein the said means of emotional monitoring is a means that can recognize and analyze words uttered by the user and emotionally colored sounds.