Network advertisement effect evaluation method based on eye movement tracking

Abstract

The invention provides a network advertisement effect evaluation method based on eye movement tracking, and aims to improve the advertisement putting accuracy and the effect evaluation accuracy through integrating eye movement data and emotion data, and the method comprises the steps of capturing an eye movement track by using an infrared camera, extracting eye movement feature points, and calculating a fixation point and time. Meanwhile, emotion data are collected through a facial expression recognition technology and a physiological sensor, a visual attention model and an emotion analysis model are constructed through a deep neural network, the reverse emotion and cognitive avoidance state of the user is analyzed, and advertisement content display is dynamically adjusted; according to the method, the user position and time data are combined, personalized advertisement putting is achieved, the advertisement effect is optimized through multi-dimensional data analysis, the user experience is improved, and a scientific basis is provided for advertisement decision making.

Description

Technical Field

[0001] The present invention relates to the technical field of advertising evaluation, and more specifically, to a network advertising effect evaluation method based on eye tracking. Background Art

[0002] As Internet advertising becomes an important part of online marketing, the demand for evaluation of advertising effectiveness is growing. Internet advertising has become one of the main online marketing methods due to its wide coverage, strong interactivity and low cost. Personalized recommendations and reasonable configuration of advertisements can improve user experience, increase customer satisfaction and improve the performance of advertising systems. Eye tracking technology, as a mature technology, can obtain human visual information and perform visualization processing, providing a method to observe user unconscious behavior information. This technology has the advantages of objectivity, synchronization and ecological validity, and has been applied to many fields such as business, industrial design, and software evaluation. Applying eye tracking technology to the study of online advertising effectiveness can effectively make up for the shortcomings of less user information and inaccurate information.

[0003] Most of the existing studies on eye tracking technology only study user behavior through information sources such as eye movement parameters, and consider different types of eye movement parameters separately. This method has a single information source and is not comprehensive enough to comprehensively evaluate the effectiveness of online advertising. Summary of the invention

[0004] In order to overcome the above technical defects of the prior art, the present invention designs a network advertising effect evaluation method based on eye tracking, which can effectively solve the above technical problems.

[0005] In order to solve the above technical problems, the technical solution of the present invention is as follows:

[0006] A network advertising effect evaluation method based on eye tracking comprises the following steps:

[0007] Get the user's eye movement data;

[0008] Constructing a visual attention model based on a deep neural network, and obtaining the user's attention area in the information flow advertisement according to the visual attention model;

[0009] Get user's emotion data;

[0010] Constructing a deep neural network-based emotion analysis model, and judging the user's adverse emotion data according to the emotion analysis model;

[0011] According to the user's adverse emotion data and cognitive avoidance state data, searching for the attention area related to the information flow advertisement from the visual attention model;

[0012] Advertisement delivery is adjusted based on the area of ​​interest.

[0013] Preferably, the step of obtaining the user's eye movement data further comprises:

[0014] Use infrared cameras to capture the user's eye movements;

[0015] The eye movement feature points are extracted through image processing technology, and the user's gaze point and gaze time are calculated using the eye movement model algorithm;

[0016] The gaze point is converted into screen coordinates to determine the user's focus area in the information flow advertisement.

[0017] Preferably, the step of constructing a visual attention model based on a deep neural network further comprises:

[0018] A convolutional neural network is used to model the user's visual attention, including:

[0019] Input the preprocessed eye movement data into the convolutional neural network;

[0020] Extract visual features through multiple convolutional layers and pooling layers;

[0021] Mapping the visual features to a user's visual attention distribution map using a fully connected layer;

[0022] The user’s visual attention sequence is analyzed through a recurrent neural network, including:

[0023] Using the visual attention distribution map output by the convolutional neural network as an input sequence of the recurrent neural network;

[0024] Using LSTM units to capture dynamic changes of the visual attention distribution map in time series;

[0025] The attention mechanism is used to dynamically adjust the user's visual focus, including:

[0026] Adjusting the weight of the visual attention model according to the user's real-time gaze point;

[0027] The visual attention distribution map is dynamically updated to reflect changes in the user's visual focus.

[0028] Preferably, the step of obtaining the user's emotion data specifically includes:

[0029] Use facial expression recognition technology to capture the user's facial expression changes;

[0030] Collecting users’ physiological data through heart rate monitors and skin conductance sensors;

[0031] Facial expressions and physiological data are fed into emotion recognition algorithms to identify the user's emotional state.

[0032] Preferably, the step of constructing a sentiment analysis model based on a deep neural network specifically includes:

[0033] Use sentiment analysis algorithms to perform semantic analysis on users’ sentiment expressions, including:

[0034] By inputting the user's text data or voice data into the sentiment analysis model;

[0035] Use natural language processing techniques to extract emotion-related features;

[0036] Use the emotion classification model to classify the user's emotional state, including:

[0037] Training the emotion classification model to recognize different categories of emotions;

[0038] Classify users’ emotional data in real time;

[0039] Real-time monitoring of users’ emotional responses through physiological signal analysis technology, including:

[0040] Collect the user's heart rate and skin conductance physiological signals;

[0041] Physiological signals are analyzed using machine learning models to identify changes in emotional states.

[0042] Preferably, the step of searching for the focus area related to the information flow advertisement according to the user's adverse emotion data and cognitive avoidance state data specifically includes:

[0043] Machine learning models are used to analyze users’ negative emotion data and cognitive avoidance state data to identify areas of concern related to advertising, including:

[0044] Input the user's adverse emotion data and cognitive avoidance state data into the machine learning model;

[0045] Train machine learning models to identify characteristics of user emotional states that correlate with ad effectiveness;

[0046] The visual attention model is adjusted according to the machine learning model output to find the attention area related to the information flow advertisement.

[0047] Preferably, the step of adjusting the advertisement delivery according to the focus area specifically includes:

[0048] Dynamically adjust the display order and display duration of advertising content based on the user's visual attention area and emotional state, including:

[0049] Analyze the user's visual attention area and determine the user's focus;

[0050] Adjust the display strategy of advertising content according to the user's focus and emotional state;

[0051] Based on the user's location and time data, selectively display advertising content that is relevant to the user's current environment and activities, including:

[0052] Collect users’ GPS location data and timestamps;

[0053] Analyze user location and time data to determine user behavior patterns;

[0054] According to the user's behavior pattern, relevant advertising content is pushed.

[0055] An electronic device, comprising:

[0056] A memory storing executable program code;

[0057] a processor coupled to the memory;

[0058] The processor calls the executable program code stored in the memory to execute the above-mentioned method for evaluating the effect of online advertising based on eye tracking.

[0059] A computer storage medium stores computer instructions, which, when called, are used to execute the above-mentioned method for evaluating the effect of online advertising based on eye tracking.

[0060] Compared with the prior art, the beneficial effects of the present invention are as follows: through multi-source data fusion, the present invention can more comprehensively understand the user's reaction to advertisements and avoid the limitations of a single information source. The application of deep learning models improves the recognition accuracy of user visual attention and emotional state, making the evaluation of advertising effects more reliable. The advertising content is dynamically adjusted according to the user's attention area and emotional state, thereby improving the pertinence and attractiveness of the advertisement. By real-time monitoring of the user's emotions and attention, the present invention helps to improve the quality of advertising content and reduce the user's adverse emotions towards advertisements, thereby improving the user experience. The method of the present invention can provide real-time feedback on advertising effects, provide data support for advertisers, and facilitate timely adjustment of advertising strategies.

[0061] In order to more clearly illustrate the implementation methods of the present invention or the technical solutions in the prior art, the drawings required for use in the implementation methods or the description of the prior art will be briefly introduced below. Obviously, the drawings in the following description are only exemplary, and for ordinary technicians in this field, other implementation drawings can be derived based on the provided drawings without paying any creative work.

[0062] Figure 1 This is a step-by-step diagram of a method for evaluating the effectiveness of online advertising based on eye tracking. DETAILED DESCRIPTION

[0063] The drawings are for illustrative purposes only and should not be construed as limiting the present patent;

[0064] In order to better illustrate the present embodiment, some parts in the drawings may be omitted, enlarged or reduced, and do not represent the size of the actual product;

[0065] It is understandable to those skilled in the art that some well-known structures and their descriptions may be omitted in the drawings.

[0066] The technical solution of the present invention is further described below in conjunction with the accompanying drawings and embodiments.

[0067] Example

[0068] A method for evaluating the effectiveness of online advertising based on eye tracking, such as Figure 1 As shown, the following steps are included:

[0069] Get the user's eye movement data;

[0070] Constructing a visual attention model based on a deep neural network, and obtaining the user's attention area in the information flow advertisement according to the visual attention model;

[0071] Get user's emotion data;

[0072] Constructing a deep neural network-based emotion analysis model, and judging the user's adverse emotion data according to the emotion analysis model;

[0073] According to the user's adverse emotion data and cognitive avoidance state data, searching for the attention area related to the information flow advertisement from the visual attention model;

[0074] Advertisement delivery is adjusted based on the area of ​​interest.

[0075] The step of obtaining the user's eye movement data further includes:

[0076] Use an infrared camera to capture the user's eye movements; when the user browses the information flow advertisement, start an infrared camera specially designed to track eye movements.

[0077] The infrared camera monitors the user's eye movements by emitting infrared light and capturing the reflected light, thereby recording the eye movement trajectory.

[0078] Eye movement feature points are extracted through image processing technology, and the eye movement model algorithm is used to calculate the user's gaze point and gaze time; image processing technologies such as edge detection, feature extraction and pattern recognition are used to identify the feature points of the eyeball, such as the center position of the pupil, from the captured eye movement images.

[0079] Apply eye movement model algorithms, such as the Pearson correlation coefficient algorithm or the Kalman filter, to process these feature point data and calculate the user's gaze point, that is, the point where the user's eyes are focused, and the gaze time, that is, the time the user stays on a specific point.

[0080] The computation of the gaze point typically involves determining the location of the pupil center in consecutive frames and estimating the gaze direction corresponding to these locations.

[0081] The gaze points are converted into screen coordinates to determine the user's focus area in the information flow advertisement. Once the gaze points are calculated, these points need to be converted from the camera coordinate system to the screen coordinate system. This usually involves a calibration process in which the user needs to look at several known points on the screen so that the system can establish a mapping relationship between the camera coordinate system and the screen coordinate system.

[0082] Through this conversion, the system can accurately know where the user looks at and for how long on the screen, thereby determining the user's attention area in the information flow ad.

[0083] The data of these attention areas can then be used to analyze the user's interest in the advertising content and the effectiveness of the advertising design.

[0084] The step of constructing a visual attention model based on a deep neural network further includes:

[0085] A convolutional neural network (CNN) is used to model the user's visual attention, including:

[0086] Input the preprocessed eye movement data into the convolutional neural network;

[0087] Extract visual features through multiple convolutional layers and pooling layers;

[0088] Mapping the visual features to a user's visual attention distribution map using a fully connected layer;

[0089] The user’s visual attention sequence is analyzed through a recurrent neural network (RNN), including:

[0090] Using the visual attention distribution map output by the convolutional neural network as an input sequence of the recurrent neural network;

[0091] Use LSTM (or GRU) units to capture the dynamic changes of the visual attention distribution map in time series;

[0092] The attention mechanism is used to dynamically adjust the user's visual focus, including:

[0093] Adjusting the weight of the visual attention model according to the user's real-time gaze point;

[0094] The visual attention distribution map is dynamically updated to reflect changes in the user's visual focus.

[0095] Inputting preprocessed eye movement data,First, the acquired eye movement data, such as gaze point coordinates and gaze time,,are preprocessed, which may include normalization, denoising, and format conversion,,so as to facilitate input into the convolutional neural network.

[0096] Feature extraction of convolutional layer and pooling layer, input the preprocessed data into the convolutional neural network, and extract the visual features of the image through multiple convolutional layers and pooling layers. The convolutional layer identifies the local features in the image through convolution operations, and the pooling layer reduces the spatial dimension of the features while retaining important information.

[0097] Feature mapping of the fully connected layer,The fully connected layer at the end of the convolutional neural network maps the extracted visual features,to a high-dimensional space to generate a user's visual attention distribution,which indicates the user's visual attention to the advertisement.

[0098] The user’s visual attention sequence is analyzed through a recurrent neural network (RNN).

[0099] Input sequence to RNN, pass the visual attention distribution map output by convolutional neural network as input sequence to recurrent neural network, where LSTM (Long Short-Term Memory Network) or GRU (Gated Recurrent Unit) is used to process sequence data.

[0100] LSTM / GRU units capture dynamic changes: LSTM or GRU units can capture the dynamic changes of visual attention distribution maps in time series and understand the continuity and change trend of user visual attention.

[0101] Use the attention mechanism to dynamically adjust the user's visual focus:

[0102] Adjust the model weights and use the attention mechanism to dynamically adjust the weights of the visual attention model based on the user's real-time gaze point, so that the model can pay more attention to the area that the user is currently focusing on.

[0103] Dynamically update the visual attention distribution map: By continuously adjusting the weights, the model dynamically updates the visual attention distribution map to reflect the user's visual focus changes in real time. This method can more accurately simulate the user's visual attention process and provide more refined data for advertising effectiveness evaluation.

[0104] The step of obtaining the user's emotion data specifically includes:

[0105] Use facial expression recognition technology to capture the user's facial expression changes;

[0106] Collecting users’ physiological data through heart rate monitors and skin conductance sensors;

[0107] Facial expressions and physiological data are fed into emotion recognition algorithms to identify the user's emotional state.

[0108] The steps of constructing a sentiment analysis model based on a deep neural network specifically include:

[0109] Use sentiment analysis algorithms to perform semantic analysis on users’ sentiment expressions, including:

[0110] By inputting the user's text data or voice data into the sentiment analysis model;

[0111] Use natural language processing techniques to extract emotion-related features;

[0112] Sentiment algorithms use a lexicon-based approach, using sentiment lexicons such as SentiWordNet, LIWC (Linguistic Inquiry and Word Count), or AFINN to assess the positive or negative sentiment score of each word or phrase in the text.

[0113] Calculate the sentiment score for the entire text, usually by adding up and averaging the sentiment scores of all the words in the text.

[0114] Use the emotion classification model to classify the user's emotional state, including:

[0115] Training the emotion classification model to recognize different categories of emotions;

[0116] Classify users’ emotional data in real time;

[0117] Real-time monitoring of users’ emotional responses through physiological signal analysis technology, including:

[0118] Collect the user's heart rate and skin conductance physiological signals;

[0119] Physiological signals are analyzed using machine learning models to identify changes in emotional states.

[0120] The step of searching for the focus area related to the information flow advertisement according to the user's adverse emotion data and cognitive avoidance state data specifically includes:

[0121] Machine learning models are used to analyze users’ negative emotion data and cognitive avoidance state data to identify areas of concern related to advertising, including:

[0122] Input the user's adverse emotion data and cognitive avoidance state data into the machine learning model;

[0123] Train machine learning models to identify characteristics of user emotional states that correlate with ad effectiveness;

[0124] The visual attention model is adjusted according to the machine learning model output to find the attention area related to the information flow advertisement.

[0125] The step of adjusting the advertisement delivery according to the focus area specifically includes:

[0126] Dynamically adjust the display order and display duration of advertising content based on the user's visual attention area and emotional state, including:

[0127] Analyze the user's visual attention area and determine the user's focus;

[0128] Adjust the display strategy of advertising content according to the user's focus and emotional state;

[0129] Based on the user's location and time data, selectively display advertising content that is relevant to the user's current environment and activities, including:

[0130] Collect users’ GPS location data and timestamps;

[0131] Analyze user location and time data to determine user behavior patterns;

[0132] According to the user's behavior pattern, relevant advertising content is pushed.

[0133] An electronic device, comprising:

[0134] A memory storing executable program code;

[0135] a processor coupled to the memory;

[0136] The processor calls the executable program code stored in the memory to execute the above-mentioned method for evaluating the effect of online advertising based on eye tracking.

[0137] A computer storage medium stores computer instructions. When the computer instructions are called, they are used to execute the above-mentioned network advertising effect evaluation method based on eye tracking. These computer instructions can be stored in storage media such as hard disks, solid-state drives, USB flash drives, optical disks, etc. for electronic devices to read and execute.

[0138] Obtain the user's eye movement data and use a high-precision infrared camera to capture the user's eye movement trajectory. The camera must have sufficient sensitivity and resolution to ensure that subtle changes in the user's eye movements are captured; extract eye movement feature points through image processing techniques such as edge detection and feature point extraction, and use eye movement model algorithms such as pupil center positioning and gaze point calculation to calculate the user's gaze point and gaze time; convert the calculated gaze point into screen coordinates to determine the user's focus area in the information flow advertisement. This step needs to consider factors such as screen resolution and user distance to ensure the accuracy of coordinate conversion.

[0139] A visual attention model based on deep neural network is constructed. Convolutional neural network (CNN) is used to model the user's visual attention. The preprocessed eye movement data is used as input, and visual features are extracted through multiple convolutional layers and pooling layers. Then, a fully connected layer is used to map the visual features to the user's visual attention distribution map. The recurrent neural network (RNN), especially the LSTM unit, is used to analyze the user's visual attention sequence. The visual attention distribution map output by CNN is used as the input sequence of RNN to capture the dynamic changes of the visual attention distribution map in the time series. The attention mechanism is used to dynamically adjust the user's visual focus. The weight of the visual attention model is adjusted according to the user's real-time gaze point, and the visual attention distribution map is dynamically updated to reflect the changes in the user's visual focus.

[0140] Obtain the user's emotional data and use facial expression recognition technology to capture the user's facial expression changes, including smiling, frowning, etc. Use heart rate monitors and skin conductance sensors to collect the user's physiological data, such as heart rate, skin conductance, etc., to reflect the user's emotional state. Input the facial expression and physiological data into the emotion recognition algorithm, and combine it with contextual information, such as advertising content, user historical behavior, etc., to identify the user's emotional state, such as pleasure, anger, rebellion, etc.

[0141] Construct a sentiment analysis model based on a deep neural network, use a sentiment analysis algorithm to perform semantic analysis on the user's emotional expression, input the user's text data, such as comments, likes, etc. or voice data into the sentiment analysis model, use natural language processing technology to extract emotion-related features, use a sentiment classification model to classify the user's emotional state, train the sentiment classification model to identify different categories of emotions, and classify the user's emotional data in real time, monitor the user's emotional response in real time through physiological signal analysis technology, collect the user's heart rate and skin conductance physiological signals, use a machine learning model to analyze these signals to identify changes in emotional state, find the focus areas related to information flow advertising based on the user's adverse emotion data and cognitive avoidance state data, use machine learning models such as support vector machines, random forests, etc. to analyze the user's adverse emotion data and cognitive avoidance state data, input these data into the machine learning model, train the model to identify the user's emotional state features related to advertising effectiveness, and adjust the visual attention model based on the output of the machine learning model to find the focus areas related to information flow advertising. These areas may be areas that users pay special attention to or avoid, which have an important impact on advertising effectiveness.

[0142] Adjust advertising delivery based on the attention area. Dynamically adjust the display order and duration of advertising content based on the user's visual attention area and emotional state. Analyze the user's visual attention area, determine the user's focus, and adjust the advertising content display strategy based on the focus and emotional state. Based on the user's location and time data, selectively display advertising content related to the user's current environment and activities. Collect the user's GPS location data and timestamp, analyze the user's location and time data, determine the user's behavior patterns, such as on the way to work, during breaks, etc., and push relevant advertising content based on the behavior patterns.

[0143] In the specific implementation, suppose a user is browsing the social media information stream. First, the user's eye movement trajectory is captured through an infrared camera to identify the user's attention to a certain advertisement. Then, the visual attention model is used to further analyze the changes in the user's visual focus, and combined with facial expression recognition, heart rate monitoring and skin conductance data, the user's emotional response to the advertisement, such as interest and disgust, is judged. If the user is detected to show rebellious emotions or cognitive avoidance, it will adjust the visual attention model to identify other potential attention areas related to the advertisement, and dynamically adjust the display strategy of the advertising content accordingly, such as reducing the display time of the advertisement or replacing it with more relevant advertising content. At the same time, it will also push advertisements related to the user's current environment, such as workplace, leisure time and activities, such as shopping and travel, based on the user's location and time data, to increase the attractiveness and conversion rate of the advertisement.

[0144] The same or similar reference numerals correspond to the same or similar components;

[0145] The terms used in the drawings to describe positional relationships are only used for illustrative purposes and should not be construed as limiting this patent;

[0146] Obviously, the above embodiments of the present invention are merely examples for clearly illustrating the present invention, and are not limitations on the implementation methods of the present invention. For ordinary technicians in the relevant field, other different forms of changes or modifications can be made on the basis of the above description. It is not necessary and impossible to list all the implementation methods here. Any modifications, equivalent substitutions and improvements made within the spirit and principles of the present invention should be included in the protection scope of the claims of the present invention.

Claims

1. A method for evaluating the effect of online advertising based on eye tracking, characterized in that: The following steps are involved: Get the user's eye movement data; Constructing a visual attention model based on a deep neural network, and obtaining the user's attention area in the information flow advertisement according to the visual attention model; Get user's emotion data; Constructing a deep neural network-based emotion analysis model, and judging the user's adverse emotion data according to the emotion analysis model; According to the user's adverse emotion data and cognitive avoidance state data, searching for the attention area related to the information flow advertisement from the visual attention model; Advertisement delivery is adjusted based on the area of ​​interest.

2. The method for evaluating the effect of online advertising based on eye tracking according to claim 1, characterized in that: The step of obtaining the user's eye movement data further includes: Use infrared cameras to capture the user's eye movements; The eye movement feature points are extracted through image processing technology, and the user's gaze point and gaze time are calculated using the eye movement model algorithm; The gaze point is converted into screen coordinates to determine the user's focus area in the information flow advertisement.

3. The method for evaluating the effect of online advertising based on eye tracking according to claim 1, characterized in that: The step of constructing a visual attention model based on a deep neural network further includes: A convolutional neural network is used to model the user's visual attention, including: Input the preprocessed eye movement data into the convolutional neural network; Extract visual features through multiple convolutional layers and pooling layers; Mapping the visual features to a user's visual attention distribution map using a fully connected layer; The user’s visual attention sequence is analyzed through a recurrent neural network, including: Using the visual attention distribution map output by the convolutional neural network as an input sequence of the recurrent neural network; Using LSTM units to capture dynamic changes of the visual attention distribution map in time series; The attention mechanism is used to dynamically adjust the user's visual focus, including: Adjusting the weight of the visual attention model according to the user's real-time gaze point; The visual attention distribution map is dynamically updated to reflect changes in the user's visual focus.

4. The method for evaluating the effect of online advertising based on eye tracking according to claim 1, characterized in that: The step of obtaining the user's emotion data specifically includes: Use facial expression recognition technology to capture the user's facial expression changes; Collecting users’ physiological data through heart rate monitors and skin conductance sensors; Facial expressions and physiological data are fed into emotion recognition algorithms to identify the user's emotional state.

5. The method for evaluating the effect of online advertising based on eye tracking according to claim 1, characterized in that: The steps of constructing a sentiment analysis model based on a deep neural network specifically include: Use sentiment analysis algorithms to perform semantic analysis on users’ sentiment expressions, including: By inputting the user's text data or voice data into the sentiment analysis model; Use natural language processing techniques to extract emotion-related features; Use the emotion classification model to classify the user's emotional state, including: Training the emotion classification model to recognize different categories of emotions; Classify users’ emotional data in real time; Real-time monitoring of users’ emotional responses through physiological signal analysis technology, including: Collect the user's heart rate and skin conductance physiological signals; Physiological signals are analyzed using machine learning models to identify changes in emotional states.

6. The method for evaluating the effect of online advertising based on eye tracking according to claim 1, characterized in that: The step of searching for the focus area related to the information flow advertisement according to the user's adverse emotion data and cognitive avoidance state data specifically includes: Machine learning models are used to analyze users’ negative emotion data and cognitive avoidance state data to identify areas of concern related to advertising, including: Input the user's adverse emotion data and cognitive avoidance state data into the machine learning model; Train machine learning models to identify characteristics of user emotional states that correlate with ad effectiveness; The visual attention model is adjusted according to the machine learning model output to find the attention area related to the information flow advertisement.

7. The method for evaluating the effect of online advertising based on eye tracking according to claim 1, characterized in that: The step of adjusting the advertisement delivery according to the focus area specifically includes: Dynamically adjust the display order and display duration of advertising content based on the user's visual attention area and emotional state, including: Analyze the user's visual attention area and determine the user's focus; Adjust the display strategy of advertising content according to the user's focus and emotional state; Based on the user's location and time data, selectively display advertising content that is relevant to the user's current environment and activities, including: Collect users’ GPS location data and timestamps; Analyze user location and time data to determine user behavior patterns; According to the user's behavior pattern, relevant advertising content is pushed.

8. An electronic device, characterized in that: The electronic device comprises: A memory storing executable program code; a processor coupled to the memory; The processor calls the executable program code stored in the memory to execute the network advertising effect evaluation method based on eye tracking as described in any one of claims 1-7.

9. A computer storage medium, characterized in that The computer storage medium stores computer instructions, which, when called, are used to execute the network advertising effect evaluation method based on eye tracking as described in any one of claims 1 to 7.

Images