New media art design auxiliary analysis method and system based on interaction behaviors

Abstract

The invention relates to the technical field of interactive design, in particular to a new media art design auxiliary analysis method and system based on interactive behaviors. The method comprises the following steps: acquiring corresponding interactive behavior data in real time and carrying out behavior action interactive frequency analysis to obtain interactive frequency data corresponding to each interactive behavior action; obtaining the corresponding interactive browsing duration of the user in the new media art design work through the background operation log, and performing work page jump evaluation to obtain the corresponding page interactive jump probability of the user in the new media art design work; and carrying out interaction hotspot plate determination and interaction mode identification analysis on the user in the new media art design work, obtaining each corresponding design element in the new media art design work, carrying out design auxiliary optimization analysis, and generating a design element interaction optimization suggestion scheme corresponding to the new media art work. The effect of the new media works in the actual interaction process can be evaluated, and improvement is facilitated.

Description

Technical Field

[0001] The present invention relates to the technical field of interactive design, and in particular to a new media art design auxiliary analysis method and system based on interactive behavior. Background Art

[0002] With the rapid development of digital technology, new media art, as an innovative form that integrates traditional art with modern technology, has become an important form of artistic creation and display. New media art emphasizes the interactive experience between the audience and the work. Artistic creation is no longer just a one-sided expression of the artist, but rather an interaction between the artist and the audience through technical means, stimulating the audience's sense of participation and creativity. Interactivity is one of the core characteristics of new media art. It enables the audience to directly influence the expression of the artwork through interactive feedback through touch, sound, vision, or other senses. However, although the interactivity of new media art has brought unprecedented freedom and expressive space to artistic creation, existing design methods and analysis tools have difficulty in efficiently capturing and analyzing the audience's interactive behavior. Currently, interactive behavior analysis in new media art design mainly relies on manual observation and traditional qualitative research methods. It is unable to systematically and efficiently process and deeply analyze large amounts of interactive data. As a result, there is a lack of effective methods to evaluate the effectiveness of new media art design works in the actual interactive process, which is not conducive to the improvement and optimization of the works. Summary of the Invention

[0003] Based on this, it is necessary for the present invention to provide a new media art design auxiliary analysis method and system based on interactive behavior to solve at least one of the above technical problems.

[0004] To achieve the above purpose, a new media art design auxiliary analysis method based on interactive behavior includes the following steps: Step S1: collecting the user's corresponding interactive behavior data in the new media art design work in real time, and performing behavioral action interaction frequency analysis based on the user's corresponding interactive behavior data in the new media art design work to obtain the interaction frequency data corresponding to each interactive behavior action of the user in the new media art design work; Step S2: Obtaining the user's corresponding interactive browsing time in the new media art design work through the background operation log, and evaluating the user's page jump in the new media art design work based on the interactive browsing time to obtain the user's corresponding page interactive jump probability in the new media art design work; Step S3: Based on the interaction frequency data corresponding to each interactive behavior action of the user in the new media art design work and in combination with the corresponding page interaction jump probability, the user's interactive hotspot section in the new media art design work is determined to obtain the user's new media work interactive hotspot section; by combining camera capture to obtain the corresponding user interaction action emotion data, and based on the user interaction action emotion data, interaction pattern recognition analysis is performed on the user's new media work interactive hotspot section to obtain the user's interactive behavior pattern in the corresponding hotspot section; Step S4: Obtain the corresponding design elements in the new media art design work, including graphics, colors, layout and sound effects, and conduct design-assisted optimization analysis on the new media art design work based on the corresponding design elements in the new media art design work and combined with the user's interactive behavior pattern in the corresponding hot section, and generate a design element interaction optimization suggestion plan corresponding to the new media art work.

[0005] Furthermore, step S1 includes the following steps: Step S11: collecting the user's corresponding interactive behavior data in the new media art design work in real time, including the interactive behavior information of the user's corresponding click, touch and slide actions with the new media art design work; Step S12: obtaining the number of interaction behaviors corresponding to each interaction behavior action of the user in the new media art design work through the corresponding interaction behavior data of the user in the new media art design work; Step S13: Based on the number of interaction behaviors corresponding to each interaction behavior action of the user in the new media art design work, the first interaction time point and the last interaction time point corresponding to each interaction behavior action are obtained during the user's interaction process with the new media art design work; Step S14: determining the interaction duration corresponding to each interactive action of the user in the new media art design work according to the first interaction time point and the last interaction time point corresponding to each interactive action; Step S15: Based on the behavioral interaction duration corresponding to each interactive behavior action of the user in the new media art design work, the behavioral interaction frequency analysis is performed on the number of interactive behaviors corresponding to each interactive behavior action of the user in the new media art design work to obtain the interaction frequency data corresponding to each interactive behavior action of the user in the new media art design work.

[0006] Furthermore, step S2 includes the following steps: Step S21: obtaining the user's corresponding interactive browsing time in the new media art design work through the background operation log; Step S22: determining the corresponding page stay jump time points on each page in the new media art design work based on the user's corresponding interactive browsing time mark in the new media art design work; Step S23: analyzing the page dwell time of the corresponding interactive browsing time based on the corresponding page dwell jump time points on each page in the new media art design work, so as to obtain the corresponding page browsing dwell time of the user on each page in the new media art design work; Step S24: Based on the corresponding page browsing dwell time of the user on each page in the new media art design work, the user's page jump evaluation in the new media art design work is performed to obtain the corresponding page interaction jump probability of the user in the new media art design work.

[0007] Furthermore, step S24 includes the following steps: Step S241: performing a user engagement depth analysis based on the page browsing dwell time of the user on each page of the new media art design work to obtain the page dwell engagement depth of the user on each page of the new media art design work; Step S242: tracking the user's jump behavior between various pages in the new media art design work, and determining the jump frequency and direction based on the user's jump behavior between various pages, so as to obtain the corresponding jump frequency and jump direction of the user on each page in the new media art design work; Step S243: performing jump frequency statistics based on the jump frequencies of the user on each page in the new media art design work, to obtain the jump frequencies of the user on each page in the new media art design work; Step S244: performing a page jump tendency evaluation based on the corresponding jump directions of the user on each page in the new media art design work, so as to obtain the corresponding inter-page jump tendency degree of the user on each page in the new media art design work; Step S245: Based on the user's corresponding page stay participation depth on each page in the new media art design work and the tendency to jump between pages and combined with the corresponding jump frequency, the user's jump probability in the new media art design work is evaluated and calculated to obtain the user's corresponding page interaction jump probability in the new media art design work.

[0008] Furthermore, step S3 includes the following steps: Step S31: determining the user's interaction hotspots in the new media art design work based on the interaction frequency data corresponding to each interactive behavior action of the user in the new media art design work and combining the corresponding page interaction jump probability to obtain the user's new media work interaction hotspots; Step S32: acquiring user interaction action emotion data corresponding to the corresponding interaction hotspot by combining camera capture; Step S33: quantifying the user interaction emotion according to the user interaction action emotion data to obtain a user interaction emotion score value; Step S34: performing interaction pattern recognition analysis on the corresponding user new media work interaction hotspot section based on the user interaction sentiment score value to obtain the user's interaction behavior pattern in the corresponding hotspot section.

[0009] Furthermore, step S31 includes the following steps: Based on the interaction frequency data corresponding to each user's interactive behavior in the new media art design work, the interaction heat of each user's interactive behavior in the new media art design work is analyzed in time and space, so as to accurately capture the interaction heat of each user's interactive behavior corresponding to each page area in the new media art design work according to the corresponding interaction frequency, and generate the interaction heat distribution corresponding to each interactive behavior in each page area; Based on the interaction heat distribution of each interactive action in each page area, perform page area cluster analysis according to the corresponding interaction heat to identify page area clusters corresponding to frequent user interaction behaviors and obtain a user interaction heat cluster area division map; By combining the corresponding page interaction jump probability, the interaction jump path density between each page area in the user interaction heat cluster area division diagram is determined, and the page area with the largest value is selected according to the interaction jump path density as the corresponding hot section in the new media art design works to obtain the user new media work interaction hot section.

[0010] Furthermore, step S33 includes the following steps: Extract facial expression and body movement entities from the user's interactive action emotion data to obtain the corresponding facial expression entities and body movement entities during the user's interaction process; The interactive emotional state analysis is performed based on the user's corresponding facial expression entities and body movement entities during the interaction process to obtain the user's corresponding interactive emotional state range during the interaction process, including happiness, frustration and anxiety; Determine the user's emotion fluctuation amplitude and the user's emotion duration corresponding to different emotion states during the interaction process based on the user's corresponding interaction emotion state interval during the interaction process; The user interaction emotion is quantified according to the user emotion fluctuation amplitude and user emotion duration corresponding to different emotional states during the interaction process to obtain the corresponding user interaction emotion score value.

[0011] Furthermore, the interaction pattern recognition analysis described in step S34 is specifically to identify the interaction behavior pattern of the hot spot section of the user's new media work interaction according to the scoring range standard corresponding to the user interaction emotion score value, so as to determine the interaction behavior pattern of the hot spot section corresponding to the user interaction emotion score value in the low range of -100 to -1 as the quick exit mode, determine the interaction behavior pattern of the hot spot section when the user interaction emotion score value is 0 as the browsing mode, and determine the interaction behavior pattern of the hot spot section corresponding to the user interaction emotion score value in the high range of 1-100 as the active participation mode.

[0012] Furthermore, step S4 includes the following steps: Step S41: Obtaining corresponding design elements in the new media art design work, including graphics, colors, layout, and sound effects; Step S42: performing correlation analysis between elements and interactive behaviors of users in corresponding hot spots based on corresponding design elements in the new media art design work, so as to obtain the degree of interactive correlation between each design element and each interactive behavior pattern in the hot spots; Step S43: Based on the degree of interactive correlation between each design element and each interactive behavior pattern under the hot section, a design-assisted optimization analysis is performed on the new media art design work, so as to make a comparative judgment between a preset correlation threshold and the degree of interactive correlation. If the degree of interactive correlation is greater than or equal to the preset correlation threshold, a positive feedback relationship exists between the corresponding design element and the corresponding interactive behavior pattern, and the design element corresponding to the positive feedback relationship is retained in the corresponding new media art design work; if the degree of interactive correlation is less than the preset correlation threshold, a negative feedback relationship exists between the corresponding design element and the corresponding interactive behavior pattern, and the design element corresponding to the negative feedback relationship is found in the corresponding new media art design work, and corresponding improvement and optimization suggestions are proposed for the problematic design element, including adjusting color matching and optimizing layout structure, to generate a design element interaction optimization suggestion scheme corresponding to the new media art work.

[0013] Furthermore, the present invention also provides a new media art design auxiliary analysis system based on interactive behavior, which is used to execute the new media art design auxiliary analysis method based on interactive behavior as described above. The new media art design auxiliary analysis system based on interactive behavior includes: The action interaction frequency analysis module is used to collect the user's corresponding interactive behavior data in the new media art design work in real time, and perform action interaction frequency analysis based on the user's corresponding interactive behavior data in the new media art design work to obtain the interaction frequency data corresponding to each interactive behavior action of the user in the new media art design work; A page jump evaluation module is used to obtain the user's corresponding interactive browsing time in the new media art design work through the background operation log, and evaluate the user's page jump in the new media art design work based on the interactive browsing time, thereby obtaining the user's corresponding page interactive jump probability in the new media art design work; The hotspot pattern recognition module is used to determine the user's interactive hotspots in the new media art design work based on the interaction frequency data corresponding to each interactive behavior action of the user in the new media art design work and the corresponding page interaction jump probability, so as to obtain the user's new media work interactive hotspots; by combining the camera to capture the corresponding user interaction action emotion data, and based on the user interaction action emotion data, perform interaction pattern recognition analysis on the user's new media work interactive hotspots, thereby obtaining the user's interactive behavior pattern in the corresponding hotspot; The work design element optimization module is used to obtain the corresponding design elements in new media art design works, including graphics, colors, layouts and sound effects, and conduct design-assisted optimization analysis on the new media art design works based on the corresponding design elements in the new media art design works and combined with the user's interactive behavior patterns in the corresponding hot sections, thereby generating corresponding design element interaction optimization suggestions for the new media art works.

[0014] Beneficial effects of the present invention: 1. The new media art design auxiliary analysis method based on interactive behavior proposed in the present invention has the beneficial effect of being able to understand the user's interactive mode and preferences in the work in detail by collecting the interactive behavior data of users in new media art design works in real time. The operation data of each user, including click, slide, drag, zoom, selection and other behaviors, will be recorded. These data not only reflect the user's basic operations, but also reflect their interest and focus of attention on certain design elements. By analyzing the frequency of these data, the user's common behavior patterns can be identified, helping designers to judge which elements or areas are most attractive to users and which interactive actions occur most frequently. This data-driven method makes art design no longer a simple creation, but more dependent on user interactive feedback, making the design more personalized and targeted, and then making targeted adjustments and optimizations. Secondly, the interactive browsing time of users in new media art design works is obtained through background operation logs, which provides quantitative data for understanding the user's interest in each page or module. The length of time users stay on each page can not only reflect the degree of attention to the content of that part, but also reveal which parts of the design trigger deeper interaction and which parts are ignored because they are not attractive enough to users. Based on this duration data, combined with user jump behavior, we can further evaluate the flow patterns of users between different pages, revealing which pages have high-frequency jumps and which pages are difficult to guide users to make effective conversions. Designers can use this data to optimize page structure and design layout, making user browsing behavior more natural and smooth, thereby improving the overall experience. This can also provide a clearer understanding of users' browsing paths, usage habits, and acceptance of different page content, and then adjust design elements to make the work more suitable for the needs of target users, enhance the user's interactive experience in the work, and better achieve efficient processing and in-depth analysis of large amounts of interactive data, thereby effectively capturing and analyzing the audience's interactive behavior. Then, by combining user interaction behavior data and page jump probability, it can help determine the user's interactive hot spots in new media art design works. Through these hot spots, designers can identify the areas where users interact most frequently and further analyze which design elements (such as color, graphics, layout) are most attractive to users. Through in-depth analysis of these hot spots, it is possible to discover the degree of user attention to different design elements, providing a very specific basis for subsequent optimization. At the same time, by combining user emotional data (such as facial expressions, body movements, etc.) obtained by the camera, it is possible to perform emotional analysis on the user's interactive behavior patterns in these hot sections. The addition of emotional data can help designers better understand the user's emotional reactions, thereby optimizing the interaction mode. Designers can improve the user experience by adjusting the usability or interaction method of the section.Finally, based on the interactive hot spots and user emotional response data obtained through analysis, a comprehensive optimization analysis is conducted in combination with various design elements in new media art design works (such as graphics, colors, layout and sound effects). The purpose of this stage is to enhance the user's interactive experience and the visual appeal of the work by optimizing the design elements. For example, through user emotional data, if a certain color combination makes most users express negative emotions, the designer can choose to adjust the color scheme and use a tone that is more in line with the user's emotional needs. The optimization of design elements is not limited to visual perception, but should also take into account the overall interactive experience. Through these data-driven analyses, designers can make optimization suggestions for each interactive element, so that the overall design is more in line with the user's actual needs, and can more accurately evaluate the effect of new media art design works in the actual interaction process, thereby enhancing the interactivity of new media art design works and the user's sense of participation, which is conducive to the improvement and optimization of new media art design works.

[0015] 2. The new media art design auxiliary analysis system based on interactive behavior proposed in the present invention is composed of an action interaction frequency analysis module, a page jump evaluation module, a hot section pattern recognition module and a work design element optimization module. It can realize any new media art design auxiliary analysis method based on interactive behavior described in the present invention, and is used to combine the operations between computer programs running on each module to realize the new media art design auxiliary analysis method based on interactive behavior. The internal structure of the system cooperates with each other, which can greatly reduce duplication of work and manpower investment, and can quickly and effectively provide a more accurate and efficient new media art design auxiliary analysis process based on interactive behavior, thereby simplifying the operation process of the new media art design auxiliary analysis system based on interactive behavior. BRIEF DESCRIPTION OF THE DRAWINGS

[0016] Other features, objects and advantages of the present invention will become more apparent upon reading the detailed description of non-limiting embodiments thereof made with reference to the following drawings: Figure 1 Schematic diagram of the steps of the new media art design auxiliary analysis method based on interactive behavior of the present invention; Figure 2 for Figure 1 Detailed step flow diagram of step S1; Figure 3 for Figure 1 Detailed step flow chart of step S2 in FIG. DETAILED DESCRIPTION

[0017] The following is a clear and complete description of the technical method of the present invention in conjunction with the accompanying drawings. It is obvious that the embodiments described are part of the embodiments of the present invention, but not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without making any creative efforts are within the scope of protection of the present invention.

[0018] In addition, the accompanying drawings are merely schematic illustrations of the present invention and are not necessarily drawn to scale. Identical reference numerals in the figures denote identical or similar parts, and thus repetitive descriptions thereof will be omitted. Some of the block diagrams shown in the accompanying drawings are functional entities that do not necessarily correspond to physically or logically separate entities. These functional entities may be implemented in software, in one or more hardware modules or integrated circuits, or in different network and / or processor and / or microcontroller approaches.

[0019] It should be understood that although the terms "first," "second," and the like may be used herein to describe various elements, these elements should not be limited by these terms. These terms are used solely to distinguish one element from another. For example, a first element may be referred to as a second element, and similarly, a second element may be referred to as a first element, without departing from the scope of the exemplary embodiments. The term "and / or" as used herein includes any and all combinations of one or more of the listed associated items.

[0020] To achieve this, please refer to Figures 1 to 3 The present invention provides a new media art design auxiliary analysis method based on interactive behavior, the method comprising the following steps: Step S1: collecting the user's corresponding interactive behavior data in the new media art design work in real time, and performing behavioral action interaction frequency analysis based on the user's corresponding interactive behavior data in the new media art design work to obtain the interaction frequency data corresponding to each interactive behavior action of the user in the new media art design work; Step S2: Obtaining the user's corresponding interactive browsing time in the new media art design work through the background operation log, and evaluating the user's page jump in the new media art design work based on the interactive browsing time to obtain the user's corresponding page interactive jump probability in the new media art design work; Step S3: Based on the interaction frequency data corresponding to each interactive behavior action of the user in the new media art design work and in combination with the corresponding page interaction jump probability, the user's interactive hotspot section in the new media art design work is determined to obtain the user's new media work interactive hotspot section; by combining camera capture to obtain the corresponding user interaction action emotion data, and based on the user interaction action emotion data, interaction pattern recognition analysis is performed on the user's new media work interactive hotspot section to obtain the user's interactive behavior pattern in the corresponding hotspot section; Step S4: Obtain the corresponding design elements in the new media art design work, including graphics, colors, layout and sound effects, and conduct design-assisted optimization analysis on the new media art design work based on the corresponding design elements in the new media art design work and combined with the user's interactive behavior pattern in the corresponding hot section, and generate a design element interaction optimization suggestion plan corresponding to the new media art work.

[0021] In the embodiment of the present invention, please refer to Figure 1 FIG. 1 is a flow chart of the steps of the interactive behavior-based new media art design auxiliary analysis method of the present invention. In this example, the interactive behavior-based new media art design auxiliary analysis method includes the following steps: Step S1: collecting the user's corresponding interactive behavior data in the new media art design work in real time, and performing behavioral action interaction frequency analysis based on the user's corresponding interactive behavior data in the new media art design work to obtain the interaction frequency data corresponding to each interactive behavior action of the user in the new media art design work; In an embodiment of the present invention, by deploying customized data collection modules for different terminal devices when collecting user interaction behavior data in real time, on the Web side, through the front-end JavaScript tracking technology, listen to events such as click, touchstart, touchmove, mousemove, and record the event type, occurrence time (accurate to milliseconds), screen coordinates (X / Y axis) and interactive element ID. For example, when a user clicks on a thumbnail of a work on the "Digital Gallery" page, the system immediately generates a log entry containing UserID, EventTime (such as 20231001143005), EventType (click), ElementID (P001-Thumbnail). On the mobile App, touch events are captured through Android's MotionEvent and iOS's UITouch interfaces, and pressure values are synchronously collected (if the device supports 3D Touch) and contact area, all interaction data are transmitted to the data center in real time via the HTTPS protocol and stored in a time series database (such as InfluxDB). When analyzing the frequency of behavioral interaction, the log data is first grouped by UserID and EventType (click, touch, slide), and the sliding window algorithm (window size 5 minutes, step size 1 minute) is used to calculate the number of interactions per unit time. For example, user "U001" generated 15 clicks and 8 slides between 14:00 and 14:05, with a click interaction frequency of 3 times / minute and a sliding frequency of 1.6 times / minute. For sliding behavior, the sliding distance (pixel level) and speed (pixels / second) are further calculated through coordinate changes as supplementary dimensions for frequency analysis. Finally, an interaction frequency data table is generated, which contains fields such as UserID, EventType, Frequency (times / minute), and interaction period, providing basic data for subsequent hot sector analysis.

[0022] Step S2: Obtaining the user's corresponding interactive browsing time in the new media art design work through the background operation log, and evaluating the user's page jump in the new media art design work based on the interactive browsing time to obtain the user's corresponding page interactive jump probability in the new media art design work; In the embodiment of the present invention, when the user interactive browsing time is obtained by parsing the background operation log, the log must contain fields such as UserID, PageID, EnterTime (page loading completion time), ExitTime (page unloading start time), SessionID, etc., and by using ETL tools (such as Apache NiFi) to clean the log, remove abnormal records with EnterTime ≥ ExitTime, and for sessions without ExitTime recorded, use 30 minutes of no operation timeout as the default exit time. For example, user "U002" in SessionID In "S20231001-001", the page enters the "Authorization Tool Page" (P002) at 15:00:10 and jumps to the "Material Library Page" (P003) at 15:05:30. The browsing time of P002 is 320 seconds. When evaluating page jumps based on interactive browsing time, we first build a page event sequence table to record the page access sequence in each session (such as P001→P002→P001→P003). By using a graph database (such as Neo4j) to store the jump relationship, the node is PageID, edge weight is jump times, and the formula used to calculate page interaction jump probability is: jump probability = target page jump times / total source page visits. For example, the "Homepage" (P001) has a total of 100 visits, of which 60 jumps to the "Work Details Page" (P002) have a jump probability of 60%; 30 jumps to the "User Center" (P003) have a probability of 30%; and 10 times have not jumped, with a probability of 10%. Finally, a jump probability matrix is generated to provide a jump frequency basis for hot section determination.

[0023] Step S3: Based on the interaction frequency data corresponding to each interactive behavior action of the user in the new media art design work and in combination with the corresponding page interaction jump probability, the user's interactive hotspot section in the new media art design work is determined to obtain the user's new media work interactive hotspot section; by combining camera capture to obtain the corresponding user interaction action emotion data, and based on the user interaction action emotion data, interaction pattern recognition analysis is performed on the user's new media work interactive hotspot section to obtain the user's interactive behavior pattern in the corresponding hotspot section; In an embodiment of the present invention, when determining interactive hotspots based on interaction frequency data and page interaction jump probability, a two-dimensional evaluation model is constructed: a page area with an interaction frequency ≥ 2 times / minute and a jump probability ≥ 50% is determined as a hotspot. First, the interaction frequency and jump probability are Z-score standardized, and the hotspot confidence formula is set: confidence = 0.6 × frequency Z value + 0.4 × jump probability Z value. Areas with confidence ≥ 1.2 trigger hotspot marking. For example, the click frequency of the "3D model editing page" (P004) is 3.5 times / minute (Z value = 1.8), and the probability of jumping to the "rendering settings page" (P005) is 65% (Z value = 1.5). The confidence = 0.6 × 1.8 + 0.4 × 1.5 = 1.68, which meets the hotspot conditions. Through an industrial-grade camera (such as Basler The acA2500-14gm (resolution 2592×1944, frame rate 14fps) captures the user interaction action emotion data of hot spots, combines the MTCNN algorithm to detect facial expressions, and the OpenPose model to recognize body movements. When quantifying the emotion data, the pre-trained Xception neural network model (based on the FER+ dataset training) is used to recognize 7 types of expressions and output probability values; body movements are divided into positive (amplitude > 20 pixels / frame), negative (amplitude < 5 pixels / frame), and neutral according to the movement amplitude of the joint points. The emotion score value is calculated as: expression probability × 0.7 + action polarity × 0.3 (positive = +1 to +100, negative = -100 to -1, neutral = 0). According to the preset score range standard, it is divided into three modes: quick exit mode (that is, score value -100 to -1), browsing mode (0), and active participation mode (1-100).

[0024] Step S4: Obtain the corresponding design elements in the new media art design work, including graphics, colors, layout and sound effects, and conduct design-assisted optimization analysis on the new media art design work based on the corresponding design elements in the new media art design work and combined with the user's interactive behavior pattern in the corresponding hot section, and generate a design element interaction optimization suggestion plan corresponding to the new media art work.

[0025] In an embodiment of the present invention, a multi-technology fusion parsing solution is adopted when obtaining design elements of new media art design works. The graphic elements are extracted with the SVG parser for path coordinates and rendering attributes. The color elements are extracted with the K-means clustering algorithm (K=5) for primary colors. The layout elements obtain element box model parameters (such as width, height, and margin) through the CSS calculation interface. The sound effect elements use the Librosa library to parse the MFCC features (Mel-frequency cepstral coefficients) and BPM values of the audio files. For example, the graphic elements of the "Interactive Story Page" (P006) include 12 clickable vector icons with the main color of #2B6CB0 (blue, accounting for 45%). The layout adopts the Flexbox elastic layout, the core buttons are concentrated in the bottom 30% area of the page, and the click sound effect is 440Hz. Pure sound (20ms duration) was used to perform auxiliary optimization analysis based on design elements and interaction patterns. Canonical correlation analysis (CCA) was used to calculate the correlation coefficient between element attributes and pattern characteristics. Taking the "active participation mode" as an example, if the correlation coefficient between the stroke width of the "brush tool icon" (4px) and the frequency of function usage was 0.85 (> the preset threshold of 0.5), it was determined to be positive feedback and recommended for retention. If the correlation coefficient between the "background color brightness" (value = 200) and the dwell time was -0.45 (< the threshold), it was determined to be negative feedback. The analysis found that the high-brightness background caused visual fatigue, and the following optimization solutions were proposed: ① Adjust the background color brightness to 180 (#E5E7EB) to reduce visual stimulation; ② Restructure the tool buttons using an F-shaped layout, fixing core functions (such as "Save" and "Undo") in the upper left corner of the screen with a spacing of ≤40px to improve operational efficiency. All optimization suggestions were accompanied by expected A / B testing indicators, such as a 15% increase in function usage frequency and a 2-minute extension in dwell time. Finally, a design element interaction optimization plan was generated, which included element parameters, correlation analysis, and improvements.

[0026] Furthermore, step S1 includes the following steps: Step S11: collecting the user's corresponding interactive behavior data in the new media art design work in real time, including the interactive behavior information of the user's corresponding click, touch and slide actions with the new media art design work; Step S12: obtaining the number of interaction behaviors corresponding to each interaction behavior action of the user in the new media art design work through the corresponding interaction behavior data of the user in the new media art design work; Step S13: Based on the number of interaction behaviors corresponding to each interaction behavior action of the user in the new media art design work, the first interaction time point and the last interaction time point corresponding to each interaction behavior action are obtained during the user's interaction process with the new media art design work; Step S14: determining the interaction duration corresponding to each interactive action of the user in the new media art design work according to the first interaction time point and the last interaction time point corresponding to each interactive action; Step S15: Based on the behavioral interaction duration corresponding to each interactive behavior action of the user in the new media art design work, the behavioral interaction frequency analysis is performed on the number of interactive behaviors corresponding to each interactive behavior action of the user in the new media art design work to obtain the interaction frequency data corresponding to each interactive behavior action of the user in the new media art design work.

[0027] As an embodiment of the present invention, refer to Figure 2 As shown, Figure 1 Detailed step flow diagram of step S1 in FIG. 1 , in this embodiment, step S1 includes the following steps: Step S11: collecting the user's corresponding interactive behavior data in the new media art design work in real time, including the interactive behavior information of the user's corresponding click, touch and slide actions with the new media art design work; In an embodiment of the present invention, when collecting user interaction behavior data in real time, corresponding data collection modules are deployed for different terminal devices. In touch-screen devices (such as smartphones and tablets), a system-level touch event monitoring interface is embedded to capture the coordinate location, pressure value (if supported by the device), and timestamp information of user clicks, touches, and slides in real time. For example, when a user clicks on the virtual canvas of a new media art design work, the system immediately records the screen coordinates (x, y), trigger time (accurate to the millisecond level), and interaction type (click) of the click event. For non-touch-screen devices (such as devices with somatosensory sensors), the inertial measurement unit (IMU) is used to collect user body movement data, and computer vision algorithms are used to identify interactive behaviors such as sliding and waving gestures. All interactive behavior data is transmitted in real time to a backend data server via the device's underlying API and stored as a structured log file containing fields such as the user's unique identifier, interaction time, action type, and coordinate parameters.

[0028] Step S12: obtaining the number of interaction behaviors corresponding to each interaction behavior action of the user in the new media art design work through the corresponding interaction behavior data of the user in the new media art design work; In an embodiment of the present invention, a data grouping and statistical algorithm is used to extract the number of user interaction behaviors in new media art and design works from interaction behavior logs. First, the log data is doubly grouped based on the user's unique identifier and the interaction behavior type (click, touch, swipe), generating a nested data structure with the user ID as the key and the action type as the subkey. For example, if user "U001" generates 20 clicks, 15 swipes, and 5 long presses (a touch lasting more than 2 seconds is considered a long press) within 30 minutes, the system automatically categorizes these actions under the corresponding subkeys and counts them. For touch behaviors, a distinction is further made between single-point touch and multi-point touch, and the number of touch points is determined by coordinate data. If the number of touch points is 1, it is counted as a single-point touch; if it is ≥2, it is counted as a multi-point touch. A real-time data stream processing framework (such as Apache Flink) is used in the statistical process to perform incremental calculations on the newly added interaction logs per second to ensure real-time updates of the number of interaction behaviors.

[0029] Step S13: Based on the number of interaction behaviors corresponding to each interaction behavior action of the user in the new media art design work, the first interaction time point and the last interaction time point corresponding to each interaction behavior action are obtained during the user's interaction process with the new media art design work; In an embodiment of the present invention, by obtaining the first and last interaction time points based on the number of interaction behaviors, two timestamp variables are maintained for each interaction behavior type of each user: first_interaction_time and last_interaction_time. Initially, these two variables are set to null. When the user generates the first interaction action (such as the first click), first_interaction_time is updated to the timestamp of the action; each subsequent time a similar action is triggered, last_interaction_time is updated in real time to the timestamp of the current action. For example, user "U002" swipes the screen for the first time at 14:00:05, and the system records first_interaction_time as 14:00:05. Subsequently, he swipes again at 14:00:10 and 14:00:15, and last_interaction_time is finally updated to 14:00:15. For different types of interactions occurring simultaneously (such as clicks and swipes triggered within the same millisecond), the system updates the timestamps of the corresponding action types in the chronological order of the event queue to ensure the accuracy of the time point records.

[0030] Step S14: determining the interaction duration corresponding to each interactive action of the user in the new media art design work according to the first interaction time point and the last interaction time point corresponding to each interactive action; In this embodiment of the present invention, when determining the duration of a behavioral interaction, the system first verifies whether first_interaction_time and last_interaction_time are valid timestamps (i.e., non-null). If both the first and last times of a certain interactive behavior type (such as sliding) exist, the time interval between the two is calculated using a time difference calculation function (such as datetime.diff() in Python), with the unit of seconds being uniform. For example, the first click behavior of user "U003" is at 10:00:00 and the last time is at 10:00:30, resulting in a behavioral interaction duration of 30 seconds. If a certain action type has only one interaction (i.e., the first and last times are the same), the duration is recorded as 1 second (to avoid division by zero errors in subsequent frequency calculations). For cross-session interactions (such as a user exiting and then re-entering a work), the system uses the session ID to distinguish interactions in different time periods and only counts the time range within a single session to ensure that the duration calculation does not include inactive periods.

[0031] Step S15: Based on the behavioral interaction duration corresponding to each interactive behavior action of the user in the new media art design work, the behavioral interaction frequency analysis is performed on the number of interactive behaviors corresponding to each interactive behavior action of the user in the new media art design work to obtain the interaction frequency data corresponding to each interactive behavior action of the user in the new media art design work.

[0032] In an embodiment of the present invention, when performing behavioral action interaction frequency analysis, the calculation formula of "number of interactions / interaction duration" is adopted, and the result unit is times / second. First, the behavioral interaction duration is converted from seconds to hours (such as 30 seconds = 0.0083 hours), and then the number of interactions of the action type is divided by the duration to obtain standardized interaction frequency data. For example, user "U004" performed 10 touch operations within 20 seconds, and the interaction frequency was 10 times / 20 seconds = 0.5 times / second. For abnormal data with a duration of 0, the system automatically ignores or marks it as invalid. Finally, each interactive behavior type of each user generates a frequency data point, which is stored as a three-dimensional data table containing user ID, action type, and interaction frequency (times / second). This data can be used for subsequent user behavior clustering analysis. For example, high-frequency sliding users are more interested in dynamic interactive elements, while low-frequency clicking users prefer static content. Furthermore, step S2 includes the following steps: Step S21: obtaining the user's corresponding interactive browsing time in the new media art design work through the background operation log; Step S22: determining the corresponding page stay jump time points on each page in the new media art design work based on the user's corresponding interactive browsing time mark in the new media art design work; Step S23: analyzing the page dwell time of the corresponding interactive browsing time based on the corresponding page dwell jump time points on each page in the new media art design work, so as to obtain the corresponding page browsing dwell time of the user on each page in the new media art design work; Step S24: Based on the corresponding page browsing dwell time of the user on each page in the new media art design work, the user's page jump evaluation in the new media art design work is performed to obtain the corresponding page interaction jump probability of the user in the new media art design work.

[0033] As an embodiment of the present invention, refer to Figure 3 As shown, Figure 1 Detailed step flow diagram of step S2 in the embodiment, step S2 includes the following steps: Step S21: obtaining the user's corresponding interactive browsing time in the new media art design work through the background operation log; In an embodiment of the present invention, when obtaining the user interactive browsing time by parsing the background operation log, the log data structure is first defined, requiring each log to contain a user unique identifier (UserID), a page unique identifier (PageID), a page entry timestamp (EnterTime), a page exit timestamp (ExitTime) and a session ID (SessionID). The background server uses a distributed log system (such as ELK Stack) to collect user operation data in real time, wherein EnterTime is generated when the user triggers a page loading event, and ExitTime is generated when the user clicks a page jump button or closes the page. If the user does not actively trigger the exit, the session timeout time (such as 30 minutes of no operation) is used as ExitTime. For example, user "U005" enters the "Home" (PageID: P001) at 15:00:10 in the session with SessionID "S20231001-001", and clicks to jump to the "Work Details Page" (PageID: P002) at 15:05:30. The browsing time of this homepage is 320 seconds. The system uses ETL tools (such as Apache NiFi) regularly cleans log data and removes abnormal records where the EnterTime is later than the ExitTime to ensure the validity of the interactive browsing time.

[0034] Step S22: determining the corresponding page stay jump time points on each page in the new media art design work based on the user's corresponding interactive browsing time mark in the new media art design work; In an embodiment of the present invention, by marking the page stop and jump time points, a page event sequence arranged in chronological order is generated for each user session based on the cleaned log data. By using an event-driven architecture, when it is detected that the EnterTime of the same user with different PageIDs has time overlap, the EnterTime triggered first shall prevail, and the later one shall be regarded as a page jump event. For example, user "U006" enters the "Interactive Tutorial Page" (P003) at 16:00:00, and clicks the "Back" button to enter the "Home Page" (P001) at 16:00:15 without exiting the page. The system marks 16:00:00 as the start time of the stay at P003, and 16:00:15 as the end time of the stay at P003 and the start time of the jump to P001. For a single-page session (only entering but not jumping), the default end time of the stay is ExitTime, and the jump time point is marked as null. By establishing a page event table (UserID, SessionID, PageID, StayStartTime, StayEndTime, JumpToPageID), accurate marking of the stay and jump time points of each page can be achieved.

[0035] Step S23: analyzing the page dwell time of the corresponding interactive browsing time based on the corresponding page dwell jump time points on each page in the new media art design work, so as to obtain the corresponding page browsing dwell time of the user on each page in the new media art design work; In an embodiment of the present invention, when performing page dwell time analysis, the dwell time is calculated for each record in the page event table using the formula: dwell time = StayEndTime - StayStartTime. The result is converted to seconds. If StayEndTime is null (i.e., the user has not redirected and the session has not timed out), the log collection deadline is used as the default end time. For example, the StayStartTime of user "U007" on the "Authoring Tool Page" (P004) is 17:30:00. Due to a system failure, the ExitTime is not recorded. During log analysis, the dwell time is calculated as 1800 seconds using 18:00:00 (the log deadline). For visits to the same page across sessions (such as when a user closes and reopens the browser), different sessions are distinguished by SessionID to avoid merging and calculating non-continuous dwell times. Ultimately, a page dwell time table is generated for each user, containing fields such as UserID, PageID, TotalStayTime (total stay time, in seconds), and VisitCount (number of visits), for subsequent redirect evaluation.

[0036] Step S24: Based on the corresponding page browsing dwell time of the user on each page in the new media art design work, the user's page jump evaluation in the new media art design work is performed to obtain the corresponding page interaction jump probability of the user in the new media art design work.

[0037] In an embodiment of the present invention, when evaluating the page interaction jump probability, a page jump matrix (JumpMatrix) is constructed, where the rows represent source pages (FromPageID) and the columns represent target pages (ToPageID). The matrix values are the number of jumps from the source page to the target page. First, the JumpToPageID field in the page event table is grouped and counted. For example, the number of jumps from source page P001 to P002 is 50 times, the number of jumps to P003 is 30 times, and the number of no jumps (i.e., session end) is 20 times. Then, the jump probability is calculated: Jump probability = number of jumps to target page / The total number of visits to the source page. Assuming that the total number of visits to P001 is 100, the jump probability from P001 to P002 is 50%, to P003 is 30%, and the probability of session end is 20%. For pages that do not jump (such as error pages), the default jump probability is 0. The final generated page jump probability table contains FromPageID, ToPageID, and JumpProbability (jump probability, retaining two decimal places), providing data support for designers to optimize the page navigation structure. For example, pages with high jump probability can add quick entry points, and pages with low jump probability need to check the interaction fluency.

[0038] Furthermore, step S24 includes the following steps: Step S241: performing a user engagement depth analysis based on the page browsing dwell time of the user on each page of the new media art design work to obtain the page dwell engagement depth of the user on each page of the new media art design work; In the embodiment of the present invention, when conducting a user engagement depth analysis, a quantitative model of page dwell engagement depth is first established to divide the user's browsing dwell time on a single page into three levels: dwell time less than 60 seconds is defined as "shallow engagement", 60 seconds to 300 seconds as "medium engagement", and more than 300 seconds as "deep engagement", and are assigned depth values of 1, 2, and 3 respectively. For example, user "U008"'s page browsing dwell time on the "Artist Introduction Page" (P005) is 200 seconds, which belongs to medium engagement and the depth value is 2; The user stays on the "Product Purchase Page" (P006) for 400 seconds, which is deep engagement, and the depth value is recorded as 3. By traversing the page dwell time table, each page visit record of each user is graded and a page dwell engagement depth table is generated, which contains fields such as UserID, PageID, StayDepth (stay depth value), and DepthLevel (engagement level). This table provides data on the degree of user engagement on different pages for subsequent analysis. For example, pages with deep engagement usually carry core interactive content, and the interaction fluency needs to be optimized.

[0039] Step S242: tracking the user's jump behavior between various pages in the new media art design work, and determining the jump frequency and direction based on the user's jump behavior between various pages, so as to obtain the corresponding jump frequency and jump direction of the user on each page in the new media art design work; In an embodiment of the present invention, when tracking user page jump behavior, the JumpToPageID field is extracted from a previously generated page event table to construct a user jump behavior dataset. For each page access sequence in each session, such as "P001→P002→P001→P003", the jump relationships between adjacent pages are recorded one by one, and directed edges (FromPageID, ToPageID) are generated. Jump relationships are stored using a graph database (such as Neo4j). The nodes are PageIDs, and the attributes of the edges include the number of jumps and timestamps. For example, user "U009" jumped from the "Homepage" (P001) to the "Work Details Page" (P002) 15 times and to the "Creation Community Page" (P007) 8 times in 5 sessions. The system automatically counts the jump frequency of P001 as 23 times, and records the main jump directions as P002 (accounting for 65%) and P007 (accounting for 35%). In this way, a jump frequency list and direction matrix are generated for each page, clarifying the user's interaction path preference between pages.

[0040] Step S243: performing jump frequency statistics based on the jump frequencies of the user on each page in the new media art design work, to obtain the jump frequencies of the user on each page in the new media art design work; In an embodiment of the present invention, when performing jump frequency statistics, the jump frequency is calculated by taking VisitCount (number of visits) in the page event table as the denominator and the number of jumps with the page as the source page as the numerator. The formula is: Jump frequency = number of source page jumps / number of source page visits, and the result is rounded to three decimal places. For example, the "Authorization Tool Page" (P004) was visited 200 times during the log period, of which the number of times users jumped from this page to other pages was 80 times. The jump frequency is 80 / 200 = 0.400 times / visit. For pages that did not jump (such as the Error404 page, which was visited 50 times and JumpToPageID was null), the jump frequency was recorded as 0. The system integrates the jump frequency data into the page interaction feature table, and associates it with data such as page stay participation depth and jump direction to form a multi-dimensional user interaction behavior data set, providing a quantitative basis for subsequent probability evaluation.

[0041] Step S244: performing a page jump tendency evaluation based on the corresponding jump directions of the user on each page in the new media art design work, so as to obtain the corresponding inter-page jump tendency degree of the user on each page in the new media art design work; In an embodiment of the present invention, when evaluating the jump tendency between pages, the ratio of the number of target page jumps to the total number of source page jumps is calculated for each source page's jump direction. The formula is: Jump tendency = number of target page jumps / total number of source page jumps. The result is expressed as a percentage with one decimal place retained. For example, the source page "Work Details Page" (P002) has a total of 200 jumps, of which 120 jumps to the "Related Works Recommendation Page" (P008) and 80 jumps to the "User Comments Page" (P009). The jump tendency from P002 to P008 is 60.0%, and to P009 is 40.0%. For pages with only a single jump direction (e.g., P003 can only jump to P001), the tendency is recorded as 100.0%. By constructing a jump tendency matrix, the interactive dependency between pages can be intuitively presented. For example, high-tendency page combinations can be designed with linkage interaction effects, while low-tendency jump paths require additional guidance prompts.

[0042] Step S245: Based on the user's corresponding page stay participation depth on each page in the new media art design work and the tendency to jump between pages and combined with the corresponding jump frequency, the user's jump probability in the new media art design work is evaluated and calculated to obtain the user's corresponding page interaction jump probability in the new media art design work.

[0043] In the embodiment of the present invention, a weighted comprehensive model is used when performing jump probability evaluation calculation, and the formula is: Page Interaction Jump Probability = 0.4×Dwell Engagement Depth + 0.3×Jump Propensity + 0.2×Jump Frequency + 0.1×Base Probability, where the base probability is the average jump probability of all pages (calculated based on historical data, such as a default value of 0.2). For example, when user "U010" visits the "Homepage" (P001), the dwell engagement depth is 2 (staying for 180 seconds), the jump propensity is 50.0% (mainly jumping to P002), and the jump frequency is 0.6 times / visit. Substituting into the formula, we can obtain: Jump Probability = 0.4×2+0.3×0.5+0.2×0.6+0.1×0.2=0.8+0.15+0.12+0.02=1.09. Since the probability value needs to be normalized to the [0, 1] interval, the final result takes the min value. (1.09, 1) = 1.00 (i.e., 100%). For combinations of pages that have not been visited, the Laplace smoothing algorithm is used to add 1 to the numerator and denominator to avoid zero probability values. This model comprehensively considers user engagement depth, jump preferences, and frequency to generate accurate probability values for each jump relationship between pages, helping designers predict user behavior and optimize page navigation logic. For example, a preloading mechanism can be added to pages with high jump probabilities to enhance the interactive experience.

[0044] Furthermore, step S3 includes the following steps: Step S31: determining the user's interaction hotspots in the new media art design work based on the interaction frequency data corresponding to each interactive behavior action of the user in the new media art design work and combining the corresponding page interaction jump probability to obtain the user's new media work interaction hotspots; In the embodiment of the present invention, when determining the interactive hotspot based on the interaction frequency data and the page interaction jump probability, a two-dimensional evaluation model is first established to normalize the interaction frequency data (unit: times / second) and the page interaction jump probability (value range [0, 1]) corresponding to each interactive behavior of the user in the new media art design work, and the Z-score normalization method is used to eliminate the dimension effect, and the hotspot determination threshold is set: the interaction frequency is greater than or equal to 1.5 times the standard deviation, and the page interaction jump probability is greater than or equal to 0.7. For example, the user " User U011's click interaction frequency on the "Virtual Gallery Page" (P010) is 0.8 times per second (1.5 standard deviations higher than the average frequency for all users), and the jump probability from this page to the "Work Collection Page" (P011) is 0.75, meeting the hotspot determination criteria. P010 is identified as an interactive hotspot. The system traverses the interaction frequency and jump probability data for all pages to generate a hotspot list. This list includes fields such as PageID, interaction frequency Z value, jump probability value, and hotspot level (level 1 / level 2), providing target areas for subsequent sentiment analysis.

[0045] Step S32: acquiring user interaction action emotion data corresponding to the corresponding interaction hotspot by combining camera capture; In an embodiment of the present invention, when capturing user interaction action and emotion data of interactive hotspots through cameras, multi-angle cameras are deployed in the physical space or virtual interface corresponding to the hotspots. For physical interaction devices, a depth camera (such as Intel RealSense) is used to collect user body movements and facial expression images in real time, with a resolution of 1920×1080 and a frame rate of 30fps. For virtual interface interactions, the user's facial video stream is obtained through the browser camera API and transmitted to the emotion analysis server after encryption. For example, when a user performs a touch and slide operation in the "virtual sculpture interactive area" (P012), the camera deployed above the device synchronously captures the user's micro-expressions such as frowning and smiling, as well as body movement data such as arm swing amplitude. The server uses a video stream processing framework (such as FFmpeg) to pre-process the raw data, including face detection (using the MTCNN algorithm) and action key point recognition (OpenPose model), extracting the coordinates of 68 facial key points and 18 body joint points to form a user interaction action and emotion dataset.

[0046] Step S33: quantifying the user interaction emotion according to the user interaction action emotion data to obtain a user interaction emotion score value; In an embodiment of the present invention, when quantifying emotion based on the emotional data of user interaction actions, the user's emotional state sequence and corresponding score values in the hot topics are first obtained through corresponding entity extraction and state analysis. The emotion scoring model uses a weighted integral method: the basic polarity values of "happy," "depressed," and "anxious" are preset to +1, -0.5, and -1, respectively. The comprehensive score is calculated by combining the duration of the emotional state (unit: seconds) and the fluctuation adjustment factor (factor = 1.2 for severe fluctuations and factor = 1.0 for gentle fluctuations). For example, the user's emotional state on the "Interactive Game Competition Page" (P023) is: happy (lasting 40 seconds, no fluctuation) → anxious (lasting 15 seconds, fluctuation amplitude -2). The score is 1 × 40 × 1.0 + (-1) × 15 × 1.2 = 40 - 18 = 22. The user emotion scores of all hot topics are normalized and mapped to the range [-100, +100], where positive values represent positive emotion, negative values represent negative emotion, and 0 represents neutral emotion.

[0047] Step S34: performing interaction pattern recognition analysis on the corresponding user new media work interaction hotspot section based on the user interaction sentiment score value to obtain the user's interaction behavior pattern in the corresponding hotspot section.

[0048] In the embodiment of the present invention, when the interaction mode is identified based on the user interaction emotion score value, three modes are divided according to the preset score range standard: quick exit mode (i.e., score value -100 to -1), browsing mode (0), and active participation mode (1-100). The system matches the score interval through the rule engine: if the user's emotion score value on the "Artwork Mall Page" (P024) is 15, it belongs to the quick exit mode, which is characterized by low emotional investment (score <30) and high exit rate (exit probability within 30 seconds >60%); if the score is 55, it belongs to the browsing mode, which is characterized by medium emotional investment (31≤score≤70), average stay time 2-5 minutes, and jump The probability of switching to a related page is 40%-60%. If the score is 85, it belongs to the active participation mode, characterized by high emotional investment (score>70), dwell time>5 minutes, and high frequency of interactive functions (such as likes and shares, frequency>3 times / minute). The user interaction data of the hot section (including score value, dwell time, and function usage frequency) is verified through the K-means clustering algorithm to ensure the accuracy of the mode division. Finally, an interactive mode analysis report is generated, marking the mainstream interaction mode and user proportion of each hot section, providing a basis for designers to optimize the interaction process, such as adding deep interaction functions for users in the active participation mode and simplifying the operation interface for users in the quick exit mode.

[0049] Furthermore, step S31 includes the following steps: Based on the interaction frequency data corresponding to each user's interactive behavior in the new media art design work, the interaction heat of each user's interactive behavior in the new media art design work is analyzed in time and space, so as to accurately capture the interaction heat of each user's interactive behavior corresponding to each page area in the new media art design work according to the corresponding interaction frequency, and generate the interaction heat distribution corresponding to each interactive behavior in each page area; In an embodiment of the present invention, when performing a spatiotemporal analysis of interaction heat based on interaction frequency data, the page of the new media art design work is first divided into a grid area of 100×100 pixels. Each grid is assigned a unique region identifier (RegionID). Through the front-end tracking technology, the coordinate data of the user's interactive behaviors such as clicks, touches, and slides in each grid area are collected in real time. Combined with the interaction frequency data (unit: times / second), a spatiotemporal two-dimensional matrix is constructed. The time dimension uses 5 minutes as a time window, and the spatial dimension records the interaction frequency of each grid within the window. For example, user "U012" clicked 8 times on the coordinates (300, 400) of the "Work Details Page" (PageID: P015) between 14:00 and 14:05. The coordinates correspond to the grid RegionID: R007, the calculated interaction frequency of the grid in this time window is 8 times / 300 seconds = 0.0267 times / second. The system uses the Gaussian kernel density estimation method to smooth the interaction frequency of the full-page grid, and finally generates an interaction behavior heat distribution heat map represented by a color gradient. The darker the color, the higher the interaction frequency, achieving precise temporal and spatial capture of the heat of user interaction behavior.

[0050] Preferably, according to the interaction heat distribution corresponding to each interactive action in each page area, page area cluster analysis is performed according to the corresponding interaction heat, so as to identify page area clusters corresponding to frequent user interaction behaviors, and obtain a user interaction heat cluster area division map; In this embodiment of the present invention, the DBSCAN (Density-Based Spatial Clustering Application) algorithm is used to perform page region cluster analysis based on the distribution of interaction behavior heat. A density threshold of 5 times / second (i.e., each grid region has an interaction frequency of ≥5 times per unit time) and a minimum number of neighboring grids of 3 is set. The algorithm traverses all grid regions, marking grids with interaction frequencies ≥ the threshold as core points. An ε-neighborhood search (ε = 2 grid spacing) connects the core points and their reachable points to form clusters of page regions with frequent user interactions. For example, on the "Virtual Exhibition Page" (Page ID: P016), the interaction frequencies of the three adjacent grids R012, R013, and R014 are 6 times / second, 7 times / second, and 5 times / second, respectively, meeting the density and neighborhood conditions and forming a single cluster. After clustering, a user interaction heat clustering map is generated. Different colored areas in the map represent different clusters, and each cluster is annotated with parameters such as average interaction frequency and number of covered grids, providing a spatial distribution basis for subsequent hotspot screening.

[0051] Preferably, the interaction jump path density between each page area in the user interaction heat cluster area division diagram is determined by combining the corresponding page interaction jump probability, and the page area corresponding to the largest value is screened out according to the interaction jump path density as the corresponding hot section in the new media art design work to obtain the user new media work interaction hot section.

[0052] In an embodiment of the present invention, when determining the interactive jump path density by combining the page interactive jump probability, the page area clusters generated by each cluster are regarded as nodes in graph theory, and the edge weight between the nodes is defined as the ratio of the number of jumps between the two areas to the jump path length (unit: times / meter). The calculation formula is: Jump path density = number of jumps from area A to area B / geometric center distance between area A and area B. The jump relationship of each regional cluster is stored in a graph database (such as Neo4j), and all edge weights are traversed to screen out the top 10% nodes with the largest density value as candidate hotspots. For example, the geometric center distance between regional clusters C001 (including R021-R025 grids) and C002 (including R031-R035 grids) is 0.5 meters, 1 The number of jumps within the hour is 200, and the jump path density is 200 times / 0.5 meters = 400 times / meter; the density of C001 and C003 is 300 times / meter, so the path density formed by C001 and C002 is higher. Finally, the top three regional clusters with the highest jump path density are selected as the hot spots for user interaction with new media works. A three-dimensional analysis report is generated that includes the location of the hot spots, interaction frequency, and jump density, providing data support for designers to optimize the interaction layout.

[0053] Furthermore, step S33 includes the following steps: Extract facial expression and body movement entities from the user's interactive action emotion data to obtain the corresponding facial expression entities and body movement entities during the user's interaction process; In the embodiment of the present invention, when extracting facial expressions and body movements from the emotional data of user interaction actions, multimodal data processing technology is used to deploy a depth camera (such as Azure Kinect) at the hardware level. The system uses a DK (Dark Knob) and an RGB camera to capture user interaction video streams at a resolution of 1280×720 (depth map) and 1920×1080 (color map), respectively, with a frame rate of 30 fps. At the software level, face detection and alignment are performed using the MTCNN (Multi-Task Cascaded Convolutional Network) algorithm, extracting the coordinates of 68 facial key points, including the contour points of the eyebrows, eyes, nose, and mouth. Simultaneously, the OpenPose model is used to detect the coordinates of 18 limb joints, including those of the shoulders, elbows, wrists, and hips. For example, when a user is operating the "Virtual Pottery Interactive Zone," the camera captures the user's mouth corners raised (the difference in the coordinates of the mouth corner key points is greater than 15 pixels) and their arms naturally drooping (the angle of the elbow joints is greater than 160 degrees). The system labels these features as a "smiling" facial expression entity and a "relaxing" body movement entity. All entity data is stored in time series, forming a structured dataset containing frame numbers, facial key point matrices, and body joint point matrices.

[0054] Preferably, an interactive emotional state analysis is performed based on the user's corresponding facial expression entity and body movement entity during the interaction process to obtain the user's corresponding interactive emotional state interval during the interaction process, including happiness, frustration and anxiety; In an embodiment of the present invention, by analyzing the interactive emotional state based on facial expression entities and body movement entities, an emotional state mapping rule library is constructed. In terms of facial expressions, the ResNet-50 model trained based on the FER+ dataset classifies the coordinates of 68 key points, outputs 7 categories of emotional probability values such as "happy", "depressed", and "anxious", and takes the category with probability > 0.6 as the dominant expression; in terms of body movements, "positive movements" (such as waving, nodding, and the joint movement amplitude > 20 pixels / frame), "negative movements" (such as hugging the arms, lowering the head, and the joint movement amplitude < 5 pixels / frame and lasting > 2 seconds), and "neutral movements" (no obvious body movements) are defined. The system uses rule fusion to determine the emotional state: if the facial expression is "happy" and the body movement is "active", the emotional state is determined to be "happy"; if the facial expression is "frown" (the distance between the eyebrows and eyes is less than 20 pixels) and the body movement is "crossing arms" (the angle of the arms crossed is less than 90 degrees), it is determined to be "depressed"; if the facial expression is "excessively open eyes" (the vertical distance between the eyes is greater than 18 pixels) and the body movement is "frequent hand raising" (raising hands more than 5 times per minute), it is determined to be "anxious". For example, if the user continuously smiles (facial expression probability 0.8) and nods (1 time per second) in the "interactive narrative page", the system determines their emotional state to be "happy".

[0055] Preferably, the user's emotion fluctuation amplitude and the user's emotion duration corresponding to different emotion states during the interaction process are determined based on the user's interaction emotion state interval corresponding to the user during the interaction process; In an embodiment of the present invention, when determining the amplitude and duration of emotional fluctuations based on interactive emotional state intervals, a time series segmentation algorithm is adopted. First, the emotional state data is sorted by timestamp, and the dynamic time warping (DTW) algorithm is used to detect state transition points. When the emotional state changes in three consecutive frames, it is marked as a fluctuation event. The emotional fluctuation amplitude is defined as the difference in emotional polarity between adjacent states (preset "happy" polarity is +1, "depressed" polarity is -0.5, and "anxious" polarity is -1). For example, when changing from "happy" to "anxious", the fluctuation amplitude is -2; the duration is the number of consecutive frames of a single emotional state × the frame interval time (33 milliseconds). For example, the emotional state sequence of a user on the "creation tool page" is: [happy (10 seconds) → anxious (5 seconds) → happy (15 seconds)], the first fluctuation amplitude is -2 (1→-1), and the duration is 5 seconds; the second fluctuation amplitude is +2 (-1→1), and the duration is 15 seconds. The system generates an emotional state timeline for each user, marking the start / end time, fluctuation amplitude, and duration of each state interval.

[0056] Preferably, the user interaction emotion is quantified according to the user emotion fluctuation amplitude and the user emotion duration corresponding to different emotion states of the user during the interaction process, so as to obtain the corresponding user interaction emotion score value.

[0057] In the embodiment of the present invention, when quantifying emotions according to the amplitude and duration of emotional fluctuations, a weighted integral model is adopted, and the calculation formula is: User interaction emotion score value = Σ(emotion polarity × duration × fluctuation adjustment factor) Polarity is +1, "frustration" is -0.5, and "anxiety" is -1. The fluctuation adjustment factor is dynamically adjusted according to the fluctuation amplitude of the adjacent states. When the amplitude is greater than 1, the factor is 1.2, and when the amplitude is ≤1, the factor is 1.0. For example, the user's emotional state on the "virtual exhibition page" is: happy (lasting 30 seconds, no fluctuation) → frustrated (lasting 20 seconds, fluctuation amplitude of -1.5) → anxious (lasting 10 seconds, fluctuation amplitude of -0.5), then the score value = 1×30×1.0+(-0.5)×20×1.2+(-1)×10×1.0=30-12-10=8. For the neutral emotional state (no valid expression and action), the score value is attenuated by 50% of the previous state. The final generated emotional score value range is [-100, +100]. Positive values indicate positive emotions, negative values indicate negative emotions, and 0 is neutral. This provides a quantitative basis for subsequent interaction pattern recognition.

[0058] Furthermore, the interaction pattern recognition analysis described in step S34 is specifically to identify the interaction behavior pattern of the hot spot section of the user's new media work interaction according to the scoring range standard corresponding to the user interaction emotion score value, so as to determine the interaction behavior pattern of the hot spot section corresponding to the user interaction emotion score value in the low range of -100 to -1 as the quick exit mode, determine the interaction behavior pattern of the hot spot section when the user interaction emotion score value is 0 as the browsing mode, and determine the interaction behavior pattern of the hot spot section corresponding to the user interaction emotion score value in the high range of 1-100 as the active participation mode.

[0059] Furthermore, step S4 includes the following steps: Step S41: Obtaining corresponding design elements in the new media art design work, including graphics, colors, layout, and sound effects; In an embodiment of the present invention, a multi-dimensional parsing technology is used to obtain design elements in new media art design works. Graphic elements are extracted through a vector graphics parser (such as an SVG Path parsing library) to identify icons, illustrations, shapes, etc. in the work, and record the geometric properties (such as vertex coordinates, curvature radius) and visual properties (such as stroke width, fill style) of the graphics. Color elements use a color extraction algorithm (such as K-means clustering) to analyze the RGB pixel distribution of the work, extract the main color, secondary color and color contrast data. For example, the main color of a certain work is #FF6B6B (red), the secondary color is #4ECDC4 (cyan), and the contrast ratio is 7:1. Layout elements are obtained through a page layout parsing tool (such as the ComputedStyle API of Chrome DevTools), and the position coordinates (X / Y axis), size (width / height), hierarchical relationship (Z-index) and white space ratio of each element are recorded, such as the "virtual sculpture display page". The main model occupies 60% of the center of the page, with a blank rate of 35%. Sound effect elements are extracted through audio spectrum analysis tools (such as Librosa), analyzing the frequency range (such as 20Hz-20kHz), rhythm speed (BPM) and sound effect triggering conditions (such as playing high-frequency prompt sounds when clicking) of the background music. Finally, a sound effect feature dataset containing waveforms and power spectrum density is generated. All design elements are classified and stored by PageID to form a structured data table containing element type, attribute parameters, and trigger conditions.

[0060] Step S42: performing correlation analysis between elements and interactive behaviors of users in corresponding hot spots based on corresponding design elements in the new media art design work, so as to obtain the degree of interactive correlation between each design element and each interactive behavior pattern in the hot spots; In an embodiment of the present invention, when performing correlation analysis of interactive behavior patterns based on design elements, the canonical correlation analysis (CCA) algorithm is used to calculate the degree of correlation between element attributes and interactive patterns. Taking the "active participation mode" as an example, hot sections with a user share of more than 40% in this mode (such as the "3D painting creation page" P025) are extracted, and their design elements are analyzed: in terms of graphics, the interactive brush icon uses a high-contrast outline (stroke width 4px); in terms of color, the canvas background is low-saturation gray (#F5F5F5); in terms of layout, the tool buttons are concentrated in the 15% area on the right side of the screen; in terms of sound effects, the brushstroke sound effect is 600 Hz high-frequency sound, CCA is used to calculate the correlation coefficient between each element attribute and pattern characteristics (stay time > 5 minutes, function usage frequency > 3 times / minute). For example, the correlation coefficient between the brush icon outline contrast and function usage frequency is 0.82, and the correlation coefficient between background color saturation and stay time is -0.65. The degree of interactive correlation is expressed as the absolute value of the correlation coefficient, ranging from [0, 1]. The larger the value, the stronger the correlation. Finally, a correlation matrix is generated. The rows represent design elements (such as graphic complexity, color brightness, layout concentration, and sound effect frequency), the columns represent interaction modes (quick exit, browsing, and active participation), and the matrix values represent the corresponding degree of correlation.

[0061] Step S43: Based on the degree of interactive correlation between each design element and each interactive behavior pattern under the hot section, a design-assisted optimization analysis is performed on the new media art design work, so as to make a comparative judgment between a preset correlation threshold and the degree of interactive correlation. If the degree of interactive correlation is greater than or equal to the preset correlation threshold, a positive feedback relationship exists between the corresponding design element and the corresponding interactive behavior pattern, and the design element corresponding to the positive feedback relationship is retained in the corresponding new media art design work; if the degree of interactive correlation is less than the preset correlation threshold, a negative feedback relationship exists between the corresponding design element and the corresponding interactive behavior pattern, and the design element corresponding to the negative feedback relationship is found in the corresponding new media art design work, and corresponding improvement and optimization suggestions are proposed for the problematic design element, including adjusting color matching and optimizing layout structure, to generate a design element interaction optimization suggestion scheme corresponding to the new media art work.

[0062] In an embodiment of the present invention, when performing design-assisted optimization analysis based on the degree of interactive correlation, a preset correlation threshold is 0.5 (calibrated by historical data). The correlation matrix is traversed. If the correlation degree between an element and the interaction mode is ≥0.5, it is determined to be a positive feedback relationship. For example, the outline contrast of the brush icon in the "active participation mode" (correlation degree 0.82) is ≥ the threshold, indicating that the high-contrast graphic design promotes users to use functions frequently, and it is retained. If the correlation degree is <0.5, it is determined to be a negative feedback relationship. For example, the button layout concentration of a page in the "quick exit mode" (correlation degree 0.35) is < the threshold. Analysis shows that the buttons are scattered around the page (average spacing > 200px). ), resulting in low user operation efficiency. To address this problem, specific optimization solutions are proposed: ① Adjust the color scheme and change the secondary function button from high-saturation red (#FF0000) to medium-saturation blue (#4A90E2) to reduce visual interference; ② Optimize the layout structure and use an F-shaped layout to concentrate the core buttons in the upper left area of the page (spacing ≤ 50px), and increase the visual white space by 15px to improve the recognition of the clickable area. All optimization suggestions are accompanied by quantitative indicators, such as an expected 20% increase in operation efficiency and an extension of the dwell time by 1.5 minutes. Finally, a design element interaction optimization suggestion plan containing problem elements, improvement plans, and expected effects is generated for designers to refer to and implement. Furthermore, the present invention also provides a new media art design auxiliary analysis system based on interactive behavior, which is used to execute the new media art design auxiliary analysis method based on interactive behavior as described above. The new media art design auxiliary analysis system based on interactive behavior includes: The action interaction frequency analysis module is used to collect the user's corresponding interactive behavior data in the new media art design work in real time, and perform action interaction frequency analysis based on the user's corresponding interactive behavior data in the new media art design work to obtain the interaction frequency data corresponding to each interactive behavior action of the user in the new media art design work; A page jump evaluation module is used to obtain the user's corresponding interactive browsing time in the new media art design work through the background operation log, and evaluate the user's page jump in the new media art design work based on the interactive browsing time, thereby obtaining the user's corresponding page interactive jump probability in the new media art design work; The hotspot pattern recognition module is used to determine the user's interactive hotspots in the new media art design work based on the interaction frequency data corresponding to each interactive behavior action of the user in the new media art design work and the corresponding page interaction jump probability, so as to obtain the user's new media work interactive hotspots; by combining the camera to capture the corresponding user interaction action emotion data, and based on the user interaction action emotion data, perform interaction pattern recognition analysis on the user's new media work interactive hotspots, thereby obtaining the user's interactive behavior pattern in the corresponding hotspot; The work design element optimization module is used to obtain the corresponding design elements in new media art design works, including graphics, colors, layouts and sound effects, and conduct design-assisted optimization analysis on the new media art design works based on the corresponding design elements in the new media art design works and combined with the user's interactive behavior patterns in the corresponding hot sections, thereby generating corresponding design element interaction optimization suggestions for the new media art works.

[0063] The present invention is therefore intended to be illustrative and non-restrictive in all respects, with the scope of the invention being defined by the appended claims rather than the foregoing description, and all changes that come within the meaning and range of equivalents of the application documents are intended to be embraced therein.

[0064] The foregoing description is intended only to provide specific embodiments of the present invention, which will enable those skilled in the art to understand and implement the present invention. Various modifications to these embodiments will be readily apparent to those skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of the present invention. Therefore, the present invention is not intended to be limited to the embodiments shown herein, but is to be construed in the widest possible manner consistent with the principles and novel features disclosed herein.

Claims

1. A new media art design auxiliary analysis method based on interactive behavior, characterized by: The following steps are involved: Step S1: collecting the user's corresponding interactive behavior data in the new media art design work in real time, and performing behavioral action interaction frequency analysis based on the user's corresponding interactive behavior data in the new media art design work to obtain the interaction frequency data corresponding to each interactive behavior action of the user in the new media art design work; Step S2: Obtaining the user's corresponding interactive browsing time in the new media art design work through the background operation log, and evaluating the user's page jump in the new media art design work based on the interactive browsing time to obtain the user's corresponding page interactive jump probability in the new media art design work; Step S3: Determine the user's interaction hotspots in the new media art design work based on the interaction frequency data corresponding to each interactive behavior action of the user in the new media art design work and in combination with the corresponding page interaction jump probability, so as to obtain the user's new media work interaction hotspots; By combining camera capture to obtain the corresponding user interaction action emotion data, and based on the user interaction action emotion data, the interaction pattern recognition analysis of the user's new media work interaction hot spots is carried out to obtain the user's interaction behavior pattern in the corresponding hot spots; Step S4: Obtain the corresponding design elements in the new media art design work, including graphics, colors, layout and sound effects, and conduct design-assisted optimization analysis on the new media art design work based on the corresponding design elements in the new media art design work and combined with the user's interactive behavior pattern in the corresponding hot section, and generate a design element interaction optimization suggestion plan corresponding to the new media art work.

2. The new media art design auxiliary analysis method based on interactive behavior according to claim 1 is characterized in that: Step S1 includes the following steps: Step S11: collecting the user's corresponding interactive behavior data in the new media art design work in real time, including the interactive behavior information of the user's corresponding click, touch and slide actions with the new media art design work; Step S12: obtaining the number of interaction behaviors corresponding to each interaction behavior action of the user in the new media art design work through the corresponding interaction behavior data of the user in the new media art design work; Step S13: Based on the number of interaction behaviors corresponding to each interaction behavior action of the user in the new media art design work, the first interaction time point and the last interaction time point corresponding to each interaction behavior action are obtained during the user's interaction process with the new media art design work; Step S14: determining the interaction duration corresponding to each interactive action of the user in the new media art design work according to the first interaction time point and the last interaction time point corresponding to each interactive action; Step S15: Based on the behavioral interaction duration corresponding to each interactive behavior action of the user in the new media art design work, the behavioral interaction frequency analysis is performed on the number of interactive behaviors corresponding to each interactive behavior action of the user in the new media art design work to obtain the interaction frequency data corresponding to each interactive behavior action of the user in the new media art design work.

3. The new media art design auxiliary analysis method based on interactive behavior according to claim 1 is characterized in that: Step S2 includes the following steps: Step S21: obtaining the user's corresponding interactive browsing time in the new media art design work through the background operation log; Step S22: determining the corresponding page stay jump time points on each page in the new media art design work based on the user's corresponding interactive browsing time mark in the new media art design work; Step S23: analyzing the page dwell time of the corresponding interactive browsing time based on the corresponding page dwell jump time points on each page in the new media art design work, so as to obtain the corresponding page browsing dwell time of the user on each page in the new media art design work; Step S24: Based on the corresponding page browsing dwell time of the user on each page in the new media art design work, the user's page jump evaluation in the new media art design work is performed to obtain the corresponding page interaction jump probability of the user in the new media art design work.

4. The interactive behavior-based new media art design auxiliary analysis method according to claim 3 is characterized in that: Step S24 includes the following steps: Step S241: performing a user engagement depth analysis based on the page browsing dwell time of the user on each page of the new media art design work to obtain the page dwell engagement depth of the user on each page of the new media art design work; Step S242: tracking the user's jump behavior between various pages in the new media art design work, and determining the jump frequency and direction based on the user's jump behavior between various pages, so as to obtain the corresponding jump frequency and jump direction of the user on each page in the new media art design work; Step S243: performing jump frequency statistics based on the jump frequencies of the user on each page in the new media art design work, to obtain the jump frequencies of the user on each page in the new media art design work; Step S244: performing a page jump tendency evaluation based on the corresponding jump directions of the user on each page in the new media art design work, so as to obtain the corresponding inter-page jump tendency degree of the user on each page in the new media art design work; Step S245: Based on the user's corresponding page stay participation depth on each page in the new media art design work and the tendency to jump between pages and combined with the corresponding jump frequency, the user's jump probability in the new media art design work is evaluated and calculated to obtain the user's corresponding page interaction jump probability in the new media art design work.

5. The interactive behavior-based new media art design auxiliary analysis method according to claim 1, characterized in that: Step S3 includes the following steps: Step S31: determining the user's interaction hotspots in the new media art design work based on the interaction frequency data corresponding to each interactive behavior action of the user in the new media art design work and combining the corresponding page interaction jump probability to obtain the user's new media work interaction hotspots; Step S32: acquiring user interaction action emotion data corresponding to the corresponding interaction hotspot by combining camera capture; Step S33: quantifying the user interaction emotion according to the user interaction action emotion data to obtain a user interaction emotion score value; Step S34: performing interaction pattern recognition analysis on the corresponding user new media work interaction hotspot section based on the user interaction sentiment score value to obtain the user's interaction behavior pattern in the corresponding hotspot section.

6. The interactive behavior-based new media art design auxiliary analysis method according to claim 5, characterized in that: Step S31 includes the following steps: Based on the interaction frequency data corresponding to each user's interactive behavior in the new media art design work, the interaction heat of each user's interactive behavior in the new media art design work is analyzed in time and space, so as to accurately capture the interaction heat of each user's interactive behavior corresponding to each page area in the new media art design work according to the corresponding interaction frequency, and generate the interaction heat distribution corresponding to each interactive behavior in each page area; Based on the interaction heat distribution of each interactive action in each page area, perform page area cluster analysis according to the corresponding interaction heat to identify page area clusters corresponding to frequent user interaction behaviors and obtain a user interaction heat cluster area division map; By combining the corresponding page interaction jump probability, the interaction jump path density between each page area in the user interaction heat cluster area division diagram is determined, and the page area with the largest value is selected according to the interaction jump path density as the corresponding hot section in the new media art design works to obtain the user new media work interaction hot section.

7. The interactive behavior-based new media art design auxiliary analysis method according to claim 5, characterized in that: Step S33 includes the following steps: Extract facial expression and body movement entities from the user's interactive action emotion data to obtain the corresponding facial expression entities and body movement entities during the user's interaction process; The interactive emotional state analysis is performed based on the user's corresponding facial expression entities and body movement entities during the interaction process to obtain the user's corresponding interactive emotional state range during the interaction process, including happiness, frustration and anxiety; Determine the user's emotion fluctuation amplitude and the user's emotion duration corresponding to different emotion states during the interaction process based on the user's corresponding interaction emotion state interval during the interaction process; The user interaction emotion is quantified according to the user emotion fluctuation amplitude and user emotion duration corresponding to different emotional states during the interaction process to obtain the corresponding user interaction emotion score value.

8. The interactive behavior-based new media art design auxiliary analysis method according to claim 5, characterized in that: The interaction pattern recognition analysis described in step S34 is specifically to identify the interaction behavior pattern of the hot spot section of the user's new media work interaction according to the scoring range standard corresponding to the user interaction emotion score value, so as to determine the interaction behavior pattern of the hot spot section corresponding to the user interaction emotion score value in the low range of -100 to -1 as the quick exit mode, determine the interaction behavior pattern of the hot spot section when the user interaction emotion score value is 0 as the browsing mode, and determine the interaction behavior pattern of the hot spot section when the user interaction emotion score value is in the high range of 1-100 as the active participation mode.

9. The interactive behavior-based new media art design auxiliary analysis method according to claim 1, characterized in that: Step S4 includes the following steps: Step S41: Obtaining corresponding design elements in the new media art design work, including graphics, colors, layout, and sound effects; Step S42: performing correlation analysis between elements and interactive behaviors of users in corresponding hot spots based on corresponding design elements in the new media art design work, so as to obtain the degree of interactive correlation between each design element and each interactive behavior pattern in the hot spots; Step S43: Based on the degree of interactive correlation between each design element and each interactive behavior pattern under the hot section, a design-assisted optimization analysis is performed on the new media art design work, so as to make a comparative judgment between a preset correlation threshold and the degree of interactive correlation. If the degree of interactive correlation is greater than or equal to the preset correlation threshold, a positive feedback relationship exists between the corresponding design element and the corresponding interactive behavior pattern, and the design element corresponding to the positive feedback relationship is retained in the corresponding new media art design work; if the degree of interactive correlation is less than the preset correlation threshold, a negative feedback relationship exists between the corresponding design element and the corresponding interactive behavior pattern, and the design element corresponding to the negative feedback relationship is found in the corresponding new media art design work, and corresponding improvement and optimization suggestions are proposed for the problematic design element, including adjusting color matching and optimizing layout structure, to generate a design element interaction optimization suggestion scheme corresponding to the new media art work.

10. A new media art design auxiliary analysis system based on interactive behavior, characterized in that: For executing the interactive behavior-based new media art design auxiliary analysis method according to claim 1, the interactive behavior-based new media art design auxiliary analysis system comprises: The action interaction frequency analysis module is used to collect the user's corresponding interactive behavior data in the new media art design work in real time, and perform action interaction frequency analysis based on the user's corresponding interactive behavior data in the new media art design work to obtain the interaction frequency data corresponding to each interactive behavior action of the user in the new media art design work; A page jump evaluation module is used to obtain the user's corresponding interactive browsing time in the new media art design work through the background operation log, and evaluate the user's page jump in the new media art design work based on the interactive browsing time, thereby obtaining the user's corresponding page interactive jump probability in the new media art design work; The hotspot pattern recognition module is used to determine the user's interactive hotspots in the new media art design work based on the interaction frequency data corresponding to each interactive behavior action of the user in the new media art design work and the corresponding page interaction jump probability, so as to obtain the user's new media work interactive hotspots; by combining the camera to capture the corresponding user interaction action emotion data, and based on the user interaction action emotion data, perform interaction pattern recognition analysis on the user's new media work interactive hotspots, thereby obtaining the user's interactive behavior pattern in the corresponding hotspot; The work design element optimization module is used to obtain the corresponding design elements in new media art design works, including graphics, colors, layouts and sound effects, and conduct design-assisted optimization analysis on the new media art design works based on the corresponding design elements in the new media art design works and combined with the user's interactive behavior patterns in the corresponding hot sections, thereby generating corresponding design element interaction optimization suggestions for the new media art works.

Images