CN-121982124-A - Thermodynamic diagram calibration method, device and equipment

CN121982124ACN 121982124 ACN121982124 ACN 121982124ACN-121982124-A

Abstract

The embodiment of the specification discloses a calibration method, a device and equipment of a thermodynamic diagram, wherein the method comprises the steps of obtaining a current behavior track and a complete behavior track which are positioned in the same page session, wherein the current behavior track is a sequence formed by user behavior data which is up to the reporting time of a preset embedded point, determining adjacent subsequent behaviors of the current behavior track based on the complete behavior track of a user terminal, determining state probability of each page state through the terminal behaviors of the current behavior track and the adjacent subsequent behaviors, calibrating the page state corresponding to the current behavior track, and generating the page thermodynamic diagram according to the calibrated page state and the user behavior data.

Inventors

LI XUEPENG

Assignees

支付宝(杭州)数字服务技术有限公司

Dates

Publication Date: 20260505
Application Date: 20260116

Claims (20)

1. A method of calibrating a thermodynamic diagram, the method comprising: acquiring a current behavior track and a complete behavior track of a same page session, wherein the current behavior track is a sequence formed by user behavior data from a preset embedded point reporting moment; Based on the complete behavior track of the user side, determining adjacent subsequent behaviors of the current behavior track; Determining the state probability of each page state through the terminal behavior of the current behavior track and the adjacent subsequent behaviors, and calibrating the page state corresponding to the current behavior track; And generating a page thermodynamic diagram according to the calibrated page state and the user behavior data.
2. The method of claim 1, wherein determining the state probability of each page state through the end behavior of the current behavior trace and the adjacent subsequent behavior, and calibrating the page state corresponding to the current behavior trace specifically comprises: Forming a behavior pair of the current behavior track based on the terminal behavior and the adjacent subsequent behavior; Matching a state transition path corresponding to the behavior pair in a preset page state machine; determining the state probability of each state transition path according to the occurrence probability corresponding to each state transition path and the conditional probability of the behavior pair; and calibrating the page state corresponding to the current behavior track based on the state probability of the state transition path.
3. The method of claim 2, based on the state probability of the state transition path, calibrating the page state corresponding to the current behavior trace, specifically comprising: acquiring the initial page state corresponding to each state transition path; Determining whether the current behavior track is an abnormal track or not based on the state probability of the state transition path; If the current behavior track is not the abnormal track, screening a corresponding initial page state based on the state probability, and taking the initial page state as the page state after the current behavior track is calibrated.
4. A method as claimed in claim 3, the method further comprising: And if the current behavior track is an abnormal track, marking the current behavior track as invalid data.
5. A method as claimed in claim 3, the method further comprising: if the current behavior track is an abnormal track, comparing the state probability of the abnormal track with a preset low confidence threshold; marking the state probabilities which are smaller than the preset low confidence threshold or the current behavior track with the empty state transition path as invalid data; Acquiring fuzzy behavior tracks except the invalid data in the abnormal tracks, and determining the weight of each initial page state based on the state probability of each state transition path corresponding to the fuzzy behavior tracks; And acquiring the composite page state after the current behavior track calibration based on the weight of each initial page state.
6. The method of claim 1, the method further comprising: the page session is uniquely identified based on a page sequence identifier; monitoring whether a target page corresponding to the page session has a page session reset event or not, wherein the page session reset event comprises background reactivation and page refreshing; if so, starting a next page session and generating a page sequence identifier of the next page session.
7. The method of claim 1, generating a page thermodynamic diagram based on the calibrated page status and the user behavior data, comprising: Dividing the user behavior data into page state levels based on calibrated page states corresponding to the user behavior data; converting the coordinates of the user behavior data based on the display window coordinates corresponding to each page state level to obtain local coordinates; According to the local coordinates corresponding to the user behavior data, a thermodynamic density matrix of each page state level is obtained to generate thermodynamic diagram layers corresponding to each page state level; and generating a page thermodynamic diagram based on the thermodynamic diagram layer corresponding to each page state level.
8. The method of claim 7, wherein generating a page thermodynamic diagram based on thermodynamic diagrams corresponding to the page state levels, specifically comprises: Ordering the thermodynamic diagram layers based on the cascading style data corresponding to each page state level, and sequentially rendering each thermodynamic diagram layer to obtain a page thermodynamic diagram; wherein, rendering each thermodynamic diagram layer to obtain a page thermodynamic diagram comprises: Determining the interaction time interval of each page state level in the page session; determining the attention attenuation coefficient of each thermodynamic layer based on the relation between the interaction time interval and the current rendering time; and weighting the corresponding thermodynamic diagram layer through the attention attenuation coefficient, and performing superposition rendering to obtain the page thermodynamic diagram.
9. The method of claim 1, wherein the step of obtaining the current behavior trace and the complete behavior trace of the user side in the same page session specifically comprises: Based on preset buried point events, obtaining user behavior data sequences which are cut off at the reporting time of each preset buried point to form a current behavior track; And forming a complete behavior track according to the user behavior data sequence corresponding to the page session.
10. The method of claim 1, wherein determining the next-to-next behavior of the current behavior trace based on the complete behavior trace of the user terminal specifically comprises: Searching a behavior track segment corresponding to the terminal behavior of the current behavior track in the complete behavior track; And extracting user behavior data adjacent to the terminal behavior in the behavior track section to serve as adjacent subsequent behaviors of the current behavior track.
11. The method of claim 5, generating a page thermodynamic diagram based on the calibrated page status and the user behavior data, comprising: Distributing the thermodynamic value corresponding to the user behavior data according to the calibrated composite page state to obtain a state thermodynamic matrix of each initial page state; dividing a target page based on a document object model of the target page corresponding to the page session to obtain a multi-page area; acquiring an association relationship between the initial page state and the page area; based on the association relation, distributing each state thermodynamic matrix to the coordinates of a page area to obtain a thermodynamic density matrix corresponding to the initial page state; and rendering the thermodynamic density matrix to generate a page thermodynamic diagram, wherein the page thermodynamic diagram distinguishes and displays thermodynamic distribution under different initial page states.
12. The method of claim 11, further comprising: based on the page thermodynamic diagram and a preset thermodynamic threshold, identifying a user interaction dense region and a user interaction sparse region in different page states; Determining the difference characteristics of the user interaction dense region and the user interaction sparse region according to the association relation between the initial page state and the page region, wherein the difference characteristics comprise the difference of theoretical association strength and actual association strength; Based on the difference features, an optimization suggestion is generated.
13. A thermodynamic diagram calibration device, the device comprising: The system comprises an acquisition module, a storage module and a processing module, wherein the acquisition module is used for acquiring a current behavior track and a complete behavior track of a same page session, wherein the current behavior track is a sequence formed by user behavior data from a preset embedded point reporting moment; the determining module is used for determining adjacent subsequent behaviors of the current behavior track based on the complete behavior track of the user side; The calibration module is used for determining the state probability of each page state through the tail end behavior of the current behavior track and the adjacent subsequent behaviors and calibrating the page state corresponding to the current behavior track; and the generation module is used for generating a page thermodynamic diagram according to the calibrated page state and the user behavior data.
14. The apparatus of claim 13, the calibration module comprising in particular: Forming a behavior pair of the current behavior track based on the terminal behavior and the adjacent subsequent behavior; Matching a state transition path corresponding to the behavior pair in a preset page state machine; determining the state probability of each state transition path according to the occurrence probability corresponding to each state transition path and the conditional probability of the behavior pair; and calibrating the page state corresponding to the current behavior track based on the state probability of the state transition path.
15. The apparatus of claim 14, the calibration module comprising in particular: acquiring the initial page state corresponding to each state transition path; Determining whether the current behavior track is an abnormal track or not based on the state probability of the state transition path; If the current behavior track is not the abnormal track, screening a corresponding initial page state based on the state probability, and taking the initial page state as the page state after the current behavior track is calibrated.
16. The apparatus of claim 15, the apparatus further comprising: and the marking module is used for marking the current behavior track as invalid data if the current behavior track is an abnormal track.
17. The apparatus of claim 15, further comprising a composite page status calibration module; the composite page state calibration module is used for comparing the state probability of the abnormal track with a preset low confidence threshold value if the current behavior track is the abnormal track; marking the state probabilities which are smaller than the preset low confidence threshold or the current behavior track with the empty state transition path as invalid data; Acquiring fuzzy behavior tracks except the invalid data in the abnormal tracks, and determining the weight of each initial page state based on the state probability of each state transition path corresponding to the fuzzy behavior tracks; And acquiring the composite page state after the current behavior track calibration based on the weight of each initial page state.
18. The apparatus of claim 13, the apparatus further comprising: The monitoring module is used for carrying out unique identification on the page session based on the page sequence identifier; monitoring whether a target page corresponding to the page session has a page session reset event or not, wherein the page session reset event comprises background reactivation and page refreshing; if so, starting a next page session and generating a page sequence identifier of the next page session.
19. The apparatus of claim 13, the generating module specifically comprises: Dividing the user behavior data into page state levels based on calibrated page states corresponding to the user behavior data; converting the coordinates of the user behavior data based on the display window coordinates corresponding to each page state level to obtain local coordinates; According to the local coordinates corresponding to the user behavior data, a thermodynamic density matrix of each page state level is obtained to generate thermodynamic diagram layers corresponding to each page state level; and generating a page thermodynamic diagram based on the thermodynamic diagram layer corresponding to each page state level.
20. The apparatus of claim 19, the generating module specifically comprises: Ordering the thermodynamic diagram layers based on the cascading style data corresponding to each page state level, and sequentially rendering each thermodynamic diagram layer to obtain a page thermodynamic diagram; wherein, rendering each thermodynamic diagram layer to obtain a page thermodynamic diagram comprises: Determining the interaction time interval of each page state level in the page session; determining the attention attenuation coefficient of each thermodynamic layer based on the relation between the interaction time interval and the current rendering time; and weighting the corresponding thermodynamic diagram layer through the attention attenuation coefficient, and performing superposition rendering to obtain the page thermodynamic diagram.

Description

Thermodynamic diagram calibration method, device and equipment Technical Field The present disclosure relates to the field of data visualization technologies, and in particular, to a thermodynamic diagram calibration method, device, and apparatus. Background Currently, in order to count the operation behaviors of a user on a page in the webpage, thermodynamic diagram drawing is usually performed on the user behaviors of the user on the webpage, so that the operated condition of the webpage is intuitively displayed for the user. However, with the improvement of the interface interaction complexity, in modern complex pages including multi-modal interaction components such as popup windows and drawers, the same screen coordinates bear completely different functions and user intentions in different states, and if the page states are not identified, thermodynamic diagrams are difficult to accurately represent the user intentions. In the traditional scheme, all operations from entering the page to the current moment of the user are recorded in sequence to form a behavior sequence. This sequence is then compared to a predefined page state machine describing the state switching of the page to identify the state in which the page is located. Along with the improvement of interface interaction complexity, the state of the page in the terminal is complex, two or more mode interaction components basically exist, the construction difficulty of a state machine is multiplied by the complex state, the situation that the expected exception exists can lead to the fact that the state of the page can not be accurately identified only by combining the current behavior sequence with the page state machine, and a certain random error exists in the generated thermodynamic diagram. Based on this, a way to be able to calibrate page states to improve thermodynamic diagram accuracy is needed. Disclosure of Invention One or more embodiments of the present disclosure provide a thermodynamic diagram calibration method, apparatus, and device, which are used for solving the technical problem that a way capable of calibrating a page state to improve thermodynamic diagram accuracy is needed. To solve the above technical problems, one or more embodiments of the present specification are implemented as follows: One or more embodiments of the present disclosure provide a thermodynamic diagram calibration method, including: acquiring a current behavior track and a complete behavior track of a same page session, wherein the current behavior track is a sequence formed by user behavior data from a preset embedded point reporting moment; Based on the complete behavior track of the user side, determining adjacent subsequent behaviors of the current behavior track; Determining the state probability of each page state through the terminal behavior of the current behavior track and the adjacent subsequent behaviors, and calibrating the page state corresponding to the current behavior track; And generating a page thermodynamic diagram according to the calibrated page state and the user behavior data. One or more embodiments of the present disclosure provide a thermodynamic diagram calibration device, the device comprising: The system comprises an acquisition module, a storage module and a processing module, wherein the acquisition module is used for acquiring a current behavior track and a complete behavior track of a same page session, wherein the current behavior track is a sequence formed by user behavior data from a preset embedded point reporting moment; the determining module is used for determining adjacent subsequent behaviors of the current behavior track based on the complete behavior track of the user side; The calibration module is used for determining the state probability of each page state through the tail end behavior of the current behavior track and the adjacent subsequent behaviors and calibrating the page state corresponding to the current behavior track; and the generation module is used for generating a page thermodynamic diagram according to the calibrated page state and the user behavior data. One or more embodiments of the present disclosure provide a thermodynamic diagram calibration apparatus, comprising: at least one processor, and A memory communicatively coupled to the at least one processor, wherein, The memory stores instructions executable by the at least one processor, the instructions are executable by the at least one processor to enable the at least one processor to: acquiring a current behavior track and a complete behavior track of a same page session, wherein the current behavior track is a sequence formed by user behavior data from a preset embedded point reporting moment; Based on the complete behavior track of the user side, determining adjacent subsequent behaviors of the current behavior track; Determining the state probability of each page state through the terminal behavior of the current behavior track and