Patent Application
20130179441
Present invention relates to the field of mobile equipment applications, more specifically to the field of solutions assessing the relevancy of digital contents (mainly text information but also photos, video, sound, text synthesized into speech, multimedia) presented to the user via mobile equipment and identifying the user's personal interests and, based on that, developing content recommendations and ranking particular content.
Several positioning-based software applications are known from prior art for conveying information on sights of interest, food and entertainment sites and other objects via mobile phones and smart phones. Widely known applications include positioning, map application and database system, based on the satellite communication, integrated into the mobile communication device, to which various service providers have added information about them. There are several well-known solutions of the kind. For example, United States patent application US2009036145 describes a system and method for providing location aware digital content to a tourist. Described solution includes a portable communications device and positioning device, by which the location of the point of interest is identified and information on the object is delivered to the user. Examples of providing location aware digital content to the user include solutions described by international patent application WO2009083744 and German patent application DE10132714, which include solution comprising a mobile phone equipped with a user location positioning feature or electronic travel guide for communicating digital tourism information to the user. International patent application WO2007134508 describes an ontology-based tourism information system, including mobile device, location positioning instrument and information server.
The limitation of described solutions is that these (a) provide no feedback on whether received information did interest the user or not, (b) do no allow the user to receive personalized information according to interests in the further.
Interest mining is essential for profiling the user mainly for (a) providing targeted advertising, (b) monitoring the feedback of users (viewers, readers). Web server log analysis is a well-known method for observing the internet users' preferences. E-Commerce applications is one of the examples (N. Hoebel, R. V. Zicari, “Creating User Profiles of Web Visitors Using Zones, Weights and Actions”, 2008 10th IEEE Conference on E-Commerce Technology and the Fifth IEEE Conference on Enterprise Computing, E-Commerce and E-Services, pp. 190-197). In interest mining the occurrence of keywords in data packets is monitored in the electronic communication (US2010131335, US20090276377). Patent application US2011072448 describes the implicit interest mining of the mobile user in case of media channels by measuring time from the beginning of media stream to stopping the stream (“stop”, “new page/channel”) by the user. It provides the possibility to monitor the mobile device sensors (location, movement) in order to identify also the user context, e.g. training situation. Existing interest mining methods based on the Access time (time when certain content, text or web page was presented on the screen for viewing) do not function well in case of a mobile user, as the content monitoring time is fragmented and the user attention/concentration level is not adequately assessed.
Various micromechanical and other sensors, e.g. camera, are used for controlling the mobile device in addition to keyboard, touch screen and voice commands. For example, by using the tilt sensor the screen view is changed according to whether the user holds the device in his hands horizontally or vertically. Also, various solutions are known that use the accelerometer, gyroscope, located in the mobile device for monitoring people's movement, e.g. for counting walking steps, identifying physical activity level (Zhou, H. and Hu, H. 2004. A Survey—Human Movement Tracking and Stroke Rehabilitation, TECHNICAL REPORT: CSM-420, University of Essex, ISSN 1744-8050) or using for some applications, e.g. for playing, in the mobile phone. It is known from prior art that the accelerometer has so far been used as part of the user interface for controlling the mobile device (EP1271288). Accelerometer and other micromechanical sensors have been employed for determining user orientation and movement in the room to measure distance to the certain point of interest e.g exhibition artefact (US20100332324). Camera has been used for tracking the movement trajectory of eyes in order to identify interesting areas of screen and actual viewing of the screen. The solution of user interest monitoring based on the camera is complicated and energy-consuming. There exist no solutions based on micromechanical sensors of mobile devices, which aim to monitor the user's digital content preference and attention.
The object of present invention is to provide a method for continuous assessing of the personal interest of the mobile device user regarding the read or viewed digital content, which allows receiving feedback on user preferences. For achieving the object of the invention a sensor integrated into the mobile device is used to continuously assess the user movement, mobile device position; temporal order of user's physical activity and device position change and, based on the sensor information, which changes in time, also user's behaviour pattern and, through that, interest towards digital content provided at given time is assessed. Method according to the invention is targeted for example at tourists acquiring information from the Internet via mobile device, but also at other users for a) assessing their interest towards specific digital content, as expressed by text, images and multimedia for the purpose of user pleasantness feedback; b) allowing to prepare user's personalised interest profile on the basis of preferred content.
Mobile devices used according to the method include for example mobile phones, smart phones, tablet PC-s, and other portable electronic devices. For example, accelerometer, magnetometer, electrostatic field sensor, tilt sensor or their combination is used as the detector identifying human movement, position or location. User's attention rate is identified either by sensor readings for the moment (mobile device position, intensity of user movement) or by temporal order of the sensor signal (device position change, order of changes in user movement intensity in time). Location information from satellite positioning systems, wireless communications transmitters, RFID tags may be employed as additional information. Web pages with descriptions of cultural heritage objects, information on entertainment and dining places, wikis and other service providers, or recorded digital textual or audiovisual information, for example, are used as digital media sources. By monitoring user preferences one can create user's personal interests profiles, which are stored either in a mobile device or in one or several servers.
Information on user preferences that is gathered by mobile device sensors can be combined with user location, with information from public web pages and portals; user calendar and social networks information or combination of these sources can be employed. In selecting the best information for the user, e.g. during the Internet search engine query, the listing is sorted according to the existing user interests profile.
The present method will now be further described with reference to the annexed drawings.
Method according to present invention for determining user preferences of digital content in a mobile device includes stages of transferring data flow from digital content source to the mobile device, monitoring user physical activity, calculating interest rate adjustment on the basis of movement information, delivering identified interest rate feedback to Content server and/or Preference server.
Based on consumer feedback, digital content providers e.g. website managers can enhance or replace their data; therefore user feedback is essential for them. Access log based monitoring methods are well known for web user interest monitoring, especially for travel and news industry. Server or host browser log monitoring used for ordinary desktop PC-s is insufficient for mobile user. Mobile user views screen information fragmentarily—walk, chats on a phone, while the web page connection stays still active. In these situations the assessment of feedback based on ordinary server logs would give a wrong judgment on user interests. With the method according to present invention the mobile device user interest in digital content is assessed and determined much more accurately. For example, it is possible to evaluate precisely what digital content was interesting during the walk for a museum visitor.
Based on the created user or user group profile the user is provided with suitable digital content and appropriate digital content presentation medium is determined for the user (e.g. text, text synthesis into sound, multimedia presentation). On the basis of received information the user is provided with suitable services (e.g. advertising, news, tourist information, information on entertainment, sports events and dining places, etc.) according to one's personal interests.
To get more appropriate content it is possible to create personal interest profiles to be stored on personal Internet Access device or remote Preference server. Profile data can be used for detailized/personalized Internet searches resulting in better matches. Content Access log-based profile building can be improved when physical activity information is taken into account. At first, effective content access time Teff can be measured. Additionally, based on experiments, certain common user movement patterns correctly indicate high interest level, which cannot be detected through the Access log-based measurements. Additionally, it is possible to record typical activity sensor patterns indicating interest range for a particular user.
For personalized content selection in a mobile device a user interest profile is created, which includes, for example, user interests, interests in digital content, preferences of the manner of presenting digital content. In one or several central profile servers a user interest profile is created, which includes, for example, user interests, interests in digital content and preferences of the manner of presenting digital content.
For monitoring user attention and interest in digital content a sensor (e.g. accelerometer, tilt sensor, magnetometer, location change, switching on and off of screen backlight, clock or any other quantifiable parameter related to the mobile device use, like applications operating in the mobile device, including phone calls) integrated into the mobile device is used, whereas at least one sensor is used simultaneously or, depending on the user's location and activities, various sensors are combined. Interest rate is assessed by a pattern of temporal changes of current values or sensor readings of one or several sensors.
In the preferred embodiment of current invention, for example, the mobile phone or smart phone or tablet PC is equipped with tilt sensor/accelerometer and/or magnetometer, electrostatic field change sensor. The sensor allows detecting whether user stands still or moves, and in which position the device is held by a user. According to test results, user prefers to view visual digital information, e.g. video or text information, without moving. Increased physical activity describes decreased interest and allows adjusting content ranking defined by logs. Real (effective) visual content access time Teff can be obtained by subtracting user's significant physical activity time Tmov from the time of displaying content on the screen Tlog, which is measured from the server or handheld device Access log. It is possible to use ‘content Access’ stability multiplier Tstab with a value between zero and one, which characterises how motionless, or, how attentively the user follows the content at given time. Physical activity level will be determined through the magnitude of movement sensor readings or external user positioning information. Larger magnitude of movement sensor readings correlate with low interest of the user. Device reading position will be determined by tilt sensing devices.
On the basis of information obtained by monitoring user attention and interest with regard to digital content typical movement/activity sensor patterns of the user are stored in the mobile device, reflecting typical user behaviour accessing content with different interest level. The classifying of typical patterns will be done using external information like questionnaires and behaviour learning methods. Different typical content access patterns for particular user, e.g. focused access period Ph1A (
Based on experiments certain human movement patterns indicate increased interest level of typical users. In
| Number | Date | Country | Kind |
|---|---|---|---|
| EP12150434 | Jan 2012 | EP | regional |
