Search

CN-121636784-B - Dynamic information position determining method

CN121636784BCN 121636784 BCN121636784 BCN 121636784BCN-121636784-B

Abstract

The application discloses a dynamic information position determining method which comprises the steps of obtaining a document object model and visual layout characteristic information of a target interaction area, recording an interaction behavior sequence of a user on the target interaction area, carrying out multi-mode fusion analysis on the document object model, the visual layout characteristic information and the interaction behavior sequence to obtain a user intention classification result, selecting target dynamic information from a preset dynamic information base according to the user intention classification result, carrying out visual saliency analysis on the visual layout characteristic information to obtain a visual center area, predicting a target display position based on the visual center area and the interaction behavior sequence, dynamically adjusting the document object model according to the target display position, and generating a dynamic information container. According to the technical scheme, the personalized matching and dynamic optimization of the dynamic information display content and the display position are realized by fusion analysis of the characteristics of the user interaction behavior and the target interaction region, and the dynamic information touch rate is improved.

Inventors

  • YU SHAOYUN

Assignees

  • 广州易尊网络科技股份有限公司

Dates

Publication Date
20260508
Application Date
20260205

Claims (8)

  1. 1. A method for dynamic information location determination, the method comprising: Acquiring a document object model and visual layout characteristic information of a target interaction area, and recording an interaction behavior sequence of a user on the target interaction area; performing multi-mode fusion analysis on the document object model, the visual layout characteristic information and the interactive behavior sequence to obtain a user intention classification result, and selecting target dynamic information from a preset dynamic information base according to the user intention classification result; The method comprises the steps of carrying out visual saliency analysis on visual layout characteristic information to obtain a visual center area, dividing the target interaction area into a plurality of display areas, mapping interaction positions corresponding to interaction behaviors in the visual center area and the interaction behavior sequence into corresponding display areas to obtain a display area sequence, determining assigned weights corresponding to transfer path information according to interaction behavior attribute information corresponding to transfer path information between adjacent display areas in the display area sequence, calculating weighted transfer times of the display areas to other display areas according to the transfer path information and the assigned weights for each display area, determining state transfer probability of the display areas to other display areas according to the weighted transfer times of the display areas to other display areas to obtain a state transfer probability matrix, predicting target display positions based on the state transfer probability matrix, dynamically adjusting a document object model according to the target display positions and generating a dynamic information container, and displaying the target dynamic information in the dynamic information container.
  2. 2. The method for determining a location of dynamic information according to claim 1, wherein the performing a multi-modal fusion analysis on the document object model, the visual layout feature information and the interactive behavior sequence to obtain a user intention classification result includes: Respectively extracting the characteristics of the document object model, the visual layout characteristic information and the interactive behavior sequence to obtain document structural characteristics, visual layout characteristics and interactive behavior characteristics; Carrying out semantic alignment on the document structural features, the visual layout features and the interactive behavior features to obtain a cross-mode unified semantic feature vector; Inputting the cross-modal unified semantic feature vector to a pre-trained intention classification model to obtain a user intention classification result output by the intention classification model, wherein the intention classification model carries out joint optimization training based on intention classification loss, cross-modal feature alignment loss and dynamic information conversion association loss.
  3. 3. The method for determining a location of dynamic information according to claim 2, wherein the performing semantic alignment on the document structural feature, the visual layout feature and the interactive behavior feature to obtain a cross-modal unified semantic feature vector comprises: mapping the document structural features, the visual layout features and the interactive behavior features to a shared lingering space respectively; calculating the attention weights of the visual layout features and the interactive behavior features relative to the document structural features in the shared latent semantic space; Respectively carrying out feature enhancement on the visual layout features and the interactive behavior features according to the attention weights to obtain enhanced visual features and enhanced behavior features; And splicing the document structural features, the enhanced visual features and the enhanced behavior features, and performing dimension reduction and integration on the spliced results to obtain the cross-mode unified semantic feature vector.
  4. 4. The method of claim 1, wherein predicting the target presentation location based on the state transition probability matrix comprises: Acquiring a display area corresponding to the current interaction behavior as a first initial display area, and taking the display area corresponding to the visual center area as a second initial display area; Respectively counting the accumulated arrival probability of each display area in the preset transition times by taking the first initial display area and the second initial display area as starting points according to the state transition probability matrix; Carrying out weighted summation calculation on the two accumulated arrival probabilities of each display area to obtain target accumulated arrival probabilities of each display area; And determining the target display position based on the display area with the highest target accumulated arrival probability.
  5. 5. The method of claim 4, wherein determining the target display position based on the display area with the highest target cumulative arrival probability comprises: selecting at least one blank subarea in a display area with the highest target accumulated arrival probability according to the visual layout characteristic information and a preset dynamic information container size; Calculating distance information between the blank subregion and the visual center region, and determining peripheral visual interference coefficients of the blank subregion according to the visual layout characteristic information; carrying out weighted summation calculation on the distance information and the peripheral vision interference coefficient to obtain a dynamic information display comprehensive score of the blank subarea; and determining the blank subregion with the highest dynamic information display comprehensive score as a target display position.
  6. 6. A dynamic information position determining apparatus, the apparatus comprising: the feature acquisition module is used for acquiring a document object model and visual layout feature information of a target interaction area and recording an interaction behavior sequence of a user on the target interaction area; The information selection module is used for carrying out multi-mode fusion analysis on the document object model, the visual layout characteristic information and the interactive behavior sequence to obtain a user intention classification result, and selecting target dynamic information from a preset dynamic information base according to the user intention classification result; The position determining module is used for carrying out visual saliency analysis on the visual layout characteristic information to obtain a visual center region, dividing the target interaction region into a plurality of display regions, mapping interaction positions corresponding to all interaction behaviors in the visual center region and the interaction behavior sequence into corresponding display regions to obtain a display region sequence, determining assignment weights corresponding to the transfer path information according to interaction behavior attribute information corresponding to transfer path information between adjacent display regions in the display region sequence, calculating weighted transfer times of the display regions to other display regions according to the transfer path information and the assignment weights for each display region, determining state transfer probability of the display region to other display regions according to the weighted transfer times of the display region to other display regions to obtain a state transfer probability matrix, predicting a target display position based on the state transfer probability matrix, and dynamically adjusting the document object model according to the target display position to generate a dynamic information container, and displaying the target dynamic information in the dynamic information container.
  7. 7. An electronic device comprising a processor, a memory and a program or instruction stored on the memory and executable on the processor, which when executed by the processor implements the dynamic information position determining method of any of claims 1-5.
  8. 8. A readable storage medium, characterized in that the readable storage medium has stored thereon a program or instructions which, when executed by a processor, implements the dynamic information position determining method according to any of claims 1-5.

Description

Dynamic information position determining method Technical Field The application belongs to the technical field of data processing, and particularly relates to a dynamic information position determining method. Background Dynamic information is used as an important carrier for digital content transmission, and the touch efficiency and conversion effect of the dynamic information directly influence the overall value of information transmission. In the digital content display carrier such as the web page and the application interface browsed by the user, the dynamic information is embedded into the digital content display carrier and presented to the user, so that the effective touch and conversion of the dynamic information can be realized. The existing dynamic information display mode generally performs static display at a fixed position of a target interaction area, namely, a display area and a display period of dynamic information are preset in an implementation code of the target interaction area, and the same dynamic information is pushed to all users accessing the target interaction area. The method is easy to cause low touch rate of the dynamic information due to misplacement of the dynamic information position and the visual focus of the user, and the same dynamic information cannot be matched with interests and requirements of different users, so that the touch rate and conversion rate of the dynamic information can be further reduced. Disclosure of Invention The embodiment of the application provides a dynamic information position determining method, which aims to realize personalized matching and dynamic optimization of dynamic information display content and display positions and improve the dynamic information touch rate by fusion analysis of user interaction behavior and target interaction region characteristics. In a first aspect, an embodiment of the present application provides a method for determining a location of dynamic information, where the method includes: Acquiring a document object model and visual layout characteristic information of a target interaction area, and recording an interaction behavior sequence of a user on the target interaction area; performing multi-mode fusion analysis on the document object model, the visual layout characteristic information and the interactive behavior sequence to obtain a user intention classification result, and selecting target dynamic information from a preset dynamic information base according to the user intention classification result; And performing visual saliency analysis on the visual layout characteristic information to obtain a visual center region, predicting a target display position based on the visual center region and the interactive behavior sequence, and dynamically adjusting the document object model according to the target display position to generate a dynamic information container, and rendering and displaying the target dynamic information in the dynamic information container. Optionally, the performing multi-mode fusion analysis on the document object model, the visual layout feature information and the interactive behavior sequence to obtain a user intention classification result includes: Respectively extracting the characteristics of the document object model, the visual layout characteristic information and the interactive behavior sequence to obtain document structural characteristics, visual layout characteristics and interactive behavior characteristics; Carrying out semantic alignment on the document structural features, the visual layout features and the interactive behavior features to obtain a cross-mode unified semantic feature vector; Inputting the cross-modal unified semantic feature vector to a pre-trained intention classification model to obtain a user intention classification result output by the intention classification model, wherein the intention classification model carries out joint optimization training based on intention classification loss, cross-modal feature alignment loss and dynamic information conversion association loss. Optionally, the performing semantic alignment on the document structural feature, the visual layout feature and the interactive behavior feature to obtain a cross-modal unified semantic feature vector includes: mapping the document structural features, the visual layout features and the interactive behavior features to a shared lingering space respectively; calculating the attention weights of the visual layout features and the interactive behavior features relative to the document structural features in the shared latent semantic space; Respectively carrying out feature enhancement on the visual layout features and the interactive behavior features according to the attention weights to obtain enhanced visual features and enhanced behavior features; And splicing the document structural features, the enhanced visual features and the enhanced behavior features, and performing d