Interactive network video advertisement promotion method and system

Abstract

The invention discloses an interactive network video advertisement promotion method and system, and relates to the technical field of data processing. Comprising the following steps: acquiring operation data of a user based on an interactive network video, wherein the operation data comprises a behavior mode and behavior time; based on the operation data, network video content attached with the user operation data is called in the network; constructing an association retrieval model, matching the retrieved network video content with user personalization, and retrieving a predicted network video based on user operation data in the network through real-time data analysis; and constructing a subject acquisition model. According to the method, the interaction switching model is constructed, the user is allowed to trigger different feature node branches through clicking and sliding operations, the setting of network video plot development is promoted, the participation sense and experience sense of the user are enhanced, advertisement information is more vivid and interesting, attention and resonance of the user are more easily aroused, and the user experience is improved. And the drainage capability of the video advertisement is enhanced to a certain extent.

Description

Technical Field

[0001] The present invention relates to the technical field of data processing, and in particular to an interactive network video advertising promotion method and system. Background Art

[0002] The interactive online video advertising promotion method is an innovative data processing technology that combines traditional video advertising with users' interactive experience through an online platform to achieve more efficient and accurate advertising promotion.

[0003] A Chinese invention patent with publication number "CN118134554A" was retrieved, which disclosed "an advertising promotion method based on big data analysis". This application can select advertisements that are more in line with user interests and needs by deeply understanding the interests, preferences, behaviors and other information of user groups. In addition, it can effectively improve the efficiency and returns of advertising by accurately positioning target users, increasing exposure and coverage, improving click and conversion rates, and real-time monitoring and optimization strategies.

[0004] The Chinese invention patent with publication number "CN111353821A" discloses "an advertising promotion system". The application includes a server and a client system. The server includes a database, a management module and a points redemption module; the client system is equipped with an advertisement selection module, a video playback module and a points module; the advertisement selection module and the video playback module are both connected to the points module. After the advertisement video is selected by the advertisement selection module and the advertisement is played by the video playback module, the points module obtains corresponding points, and the video playback module consumes points by playing film and television information. Through the points earning and consumption mode, the number and frequency of advertisement playbacks are greatly increased, so that the advertisements can be promoted more.

[0005] However, when the methods and systems disclosed in the above-mentioned patents and similar patents are actually run, the interactivity between users and advertisements is limited, so that users often lack sufficient sense of participation and immersion when watching video promotion advertisements, making it difficult to stimulate users' interest and resonance in the advertising content, and the traffic-generating effect is limited. Summary of the invention

[0006] The purpose of the present invention is to provide an interactive network video advertising promotion method and system to solve the problems raised in the above background technology.

[0007] In the first aspect, in order to achieve the above-mentioned purpose, the present invention provides the following technical solutions: an interactive network video advertising promotion method, comprising:

[0008] Obtaining user operation data based on interactive network videos, including: behavior patterns and behavior time;

[0009] Based on the operation data, retrieve network video content attached with the user operation data in the network;

[0010] Build an associated retrieval model to match retrieved online video content with user personalization, and retrieve predicted online videos based on user operation data within the network through real-time data analysis;

[0011] Construct a subject acquisition model, which is based on the predicted network video features and outputs promotional products that match the predicted network video features;

[0012] Create online promotional videos with feature node branches based on the features of the promoted products;

[0013] Constructing a trigger node, where the trigger node is set on a feature branch node in the online promotion video based on the features of the promoted product;

[0014] Construct an interactive switching model, which is used for users to select different feature node branches in the online promotion video by interactively triggering the promotion product, so as to promote the development of the online promotion video plot;

[0015] The interactive switching model includes:

[0016] Obtain scene data in the online promotion video, and set multiple trigger points corresponding to the promotion products in the online promotion video based on the scene data;

[0017] Each trigger point corresponds to the user's behavior pattern in the online promotion video;

[0018] When the user triggers the corresponding trigger point, the storyline in the online promotion video will develop according to the set feature node branches, and the storyline is used to display the promotion product information related to the trigger point.

[0019] As a further preferred embodiment of the present technical solution, the scene data includes: physical data and behavior data;

[0020] Physical data is used to record the environment, layout and props in the online promotion video;

[0021] Behavioral data is used to record the user's actions, time, and interactions with other users while watching online promotional videos;

[0022] The trigger point setting methods include:

[0023] Scan online promotional videos, obtain key frames based on scene data, and extract significant features from the key frames. The significant features include: the appearance of objects and the actions of characters in the scene;

[0024] Match the promotion product features with the preset promotion product database, identify the trigger promotion product corresponding to the significant features, and set the trigger point at the key frame based on the appearance of the object and the action of the scene character.

[0025] As a further preferred embodiment of the present technical solution, the method for creating an online promotion video includes:

[0026] Based on the characteristics of the promoted product, search for relevant video footage about the promoted product on the Internet, and extract scene clips that match the characteristics of the promoted product;

[0027] A combined insertion model is constructed, which is used to splice and combine the extracted scene fragments based on the promotional products to construct one or more coherent online promotion video drafts.

[0028] As a further preferred embodiment of the present technical solution, the method for creating an online promotion video also includes:

[0029] Based on the characteristics of the promoted product, search the network for relevant text materials about the promoted product;

[0030] Based on the text material, manually polish the storyline to match the characteristics of the promoted product;

[0031] Match the polished storyline with the draft online promotion video, and adjust the scene clips, rhythm and soundtrack of the draft online promotion video;

[0032] Acquire storyline content, prepare auxiliary videos based on the content, and merge the auxiliary videos with the online promotion video draft to form the online promotion video.

[0033] As a further preferred embodiment of the present technical solution, the associated retrieval model includes:

[0034] Obtain metadata information of online video content, including the subject, tags, and keywords of the video, analyze the metadata information based on machine learning algorithms, and extract the subject features of the video content and user preference features;

[0035] Constructing a feature vector space, mapping the theme features of the video content and the user preference features into the feature vector space, and forming feature vectors of the video content and the user preference;

[0036] Construct a comparison model, which calculates the similarity between the video content feature vector and the user preference feature vector, and outputs the comparison result;

[0037] Based on the comparison results, the predicted network videos based on the user operation data are retrieved within the network.

[0038] As a further preferred embodiment of the present technical solution, the user personalized matching includes:

[0039] Obtain users’ online behavior data, including users’ browsing history, search history, and the types and duration of videos watched;

[0040] Based on big data analysis, we analyze network behavior data to extract users’ interest preferences and demand characteristics;

[0041] Based on user interest preferences and demand characteristics, build user IP, which is used to identify the user's unique attributes and needs;

[0042] Based on the promotion product list, retrieve the characteristics of the user's IP, filter out promotion products that match the user's interest preferences and demand characteristics, and form a personalized promotion product list.

[0043] As a further preferred embodiment of the present technical solution, the subject acquisition model includes:

[0044] Collect and predict network video features, including the theme, style, emotion and potential audience information of the video;

[0045] Based on the collected predicted network video features, a feature mapping table is constructed, the feature mapping table is used to associate the video features with the promotion product features to form a mapping relationship between the predicted network video and the promotion product, and the promotion product matching the predicted network video features is selected. Based on the mapping relationship, the selected promotion products are sorted according to the preset promotion strategy to form a promotion product list;

[0046] Based on the user's behavior patterns and behavior time in the online promotional video, the collection order of the promotional product list is dynamically adjusted.

[0047] As a further preferred embodiment of the present technical solution, the trigger node includes two modes: user behavior trigger and time trigger;

[0048] User behavior triggers include:

[0049] When users complete behavioral operations or meet preset conditions in the online promotional video, the trigger node is activated, which in turn drives the plot of the online video to the display of promotional product information related to the trigger point;

[0050] Time triggering means that when the online promotion video is played to a preset time point, the node is automatically triggered to embed the promotional product information into the video.

[0051] In the second aspect, in order to improve the above technical solution, an interactive network video advertising promotion system is proposed. The system uses the above interactive network video advertising promotion method and includes:

[0052] User behavior analysis module, used to monitor and analyze user behavior data in online promotional videos in real time;

[0053] Promotional product database module, used to provide data support for the selection and matching of promotional products;

[0054] Video content generation module: Based on the characteristics of the promoted product, the video content generation module is used to automatically retrieve and extract relevant video materials on the Internet, build a coherent online promotion video draft through the combination insertion model, and finally generate a complete online promotion video by polishing and adjusting the storyline;

[0055] A correlation retrieval module is used to analyze metadata information of video content using machine learning algorithms;

[0056] The personalized matching module is used to analyze the user's network behavior data, extract the user's interest preferences and demand characteristics, construct the user's IP, and based on the promotion product list, screen out promotion products that match the user's interest preferences and demand characteristics to form a personalized promotion product list;

[0057] The trigger node management module is used to manage the trigger nodes in the interactive switching model, including user behavior triggering and time triggering, to ensure that the trigger nodes are activated at the appropriate time and under the appropriate conditions, and to promote the development of the online video plot to the display of promotional product information related to the trigger point;

[0058] The user interface display module is used to display the generated online promotion videos and personalized promotion product lists to users, and at the same time provides a user interaction interface, allowing users to trigger different feature node branches through clicks and slides, promote the development of the online video plot, and enhance the user's sense of participation and experience.

[0059] As a further preferred embodiment of the present technical solution, the user interface display module further includes:

[0060] Real-time feedback module, used to collect real-time feedback data from users during the process of watching online promotional videos, including users’ viewing time, number of views, likes, comments, and shares;

[0061] Data analysis unit, used to conduct in-depth analysis of real-time feedback data to extract users' viewing habits, preferences, and acceptance of promoted products;

[0062] The content optimization unit optimizes the content of existing online promotion videos based on the user feedback information extracted by the data analysis unit, including adjusting the video rhythm, adding new scene clips, and changing the soundtrack;

[0063] Personalized recommendation unit, user IP built based on user feedback information and personalized matching module.

[0064] Compared with the prior art, the present invention has the following beneficial effects:

[0065] The interactive online video advertising promotion method and system, through in-depth analysis of user network behavior data, screens out promotional products that are highly matched with users based on these characteristics, and forms a personalized promotional product list, which not only improves the click-through rate and conversion rate of advertisements, but also enhances the user's sense of participation and experience effect;

[0066] In addition, the interactive online video advertising promotion method proposed in the present invention, by constructing an interactive switching model, allows users to trigger different feature node branches by clicking and sliding operations, and promotes the setting of the development of the online video plot, which not only enhances the user's sense of participation and experience, but also makes the advertising information more vivid and interesting, and is more likely to attract the attention and resonance of users, thereby enhancing the traffic-generating ability of video advertising to a certain extent;

[0067] It should be added that the present invention can timely discover the user's preferences and preferences for advertising content and the degree of acceptance of the promoted products by collecting and analyzing the user's feedback data during the online promotion video viewing process in real time, and based on these feedback information, the existing online promotion video can be optimized in content, the video rhythm can be adjusted, new scene clips can be added, and the soundtrack can be changed;

[0068] Finally, the present invention realizes the efficient use of relevant film and television footage and promotional product information in the network by constructing an associated retrieval model and a feature mapping table, and constructs a coherent and attractive network promotion video, which not only reduces the cost and time of advertising production, but also improves the quality and effect of advertising. BRIEF DESCRIPTION OF THE DRAWINGS

[0069] Figure 1 It is the operation logic diagram of the method of the present invention;

[0070] Figure 2 It is the operation logic diagram of the interactive switching model of the present invention;

[0071] Figure 3 A diagram showing a method for setting a trigger point of the present invention;

[0072] Figure 4 A diagram of a method for creating a network promotion video of the present invention;

[0073] Figure 5 It is the operation logic diagram of the association retrieval model of the present invention;

[0074] Figure 6 It is the operation logic diagram of the main acquisition model of the present invention;

[0075] Figure 7 It is a module composition diagram of the system of the present invention. DETAILED DESCRIPTION

[0076] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0077] When the interactive network video advertising promotion method proposed in the present invention is actually run, the original video materials may gradually lose their appeal as the times change and user needs continue to change. Therefore, it is necessary to regularly update the video material library and introduce new and timely footage to ensure that the content of the network promotion video is always synchronized with the interests and needs of users. In addition, during actual operation, it can also be combined with social media platforms to achieve social promotion of advertisements. For example, through social interactions such as sharing, liking and commenting, the advertising content can be shared with friends, thereby expanding the scope of advertising dissemination, increasing the user's sense of participation and experience, and improving the advertising effect.

[0078] like Figure 1-Figure 6 As shown, the present invention provides a technical solution: an interactive network video advertising promotion method, comprising: obtaining user operation data based on interactive network videos, the operation data comprising: behavior patterns and behavior time, based on the operation data, retrieving network video content attached with user operation data in the network, building an associated retrieval model, matching the retrieved network video content with user personalization, and retrieving predicted network videos based on user operation data in the network through real-time data analysis, building a subject acquisition model, the subject acquisition model outputs promotion products matching the predicted network video features based on predicted network video features, creating a network promotion video with feature node branches based on promotion product features, building a trigger node, the trigger node is set on a feature branch node in the network promotion video based on the promotion product features, building an interactive switching model, the interactive switching model is used for users to select different feature node branches in the network promotion video in a manner of interactively triggering promotion products, and promoting the development of the network promotion video plot.

[0079] As a preferred embodiment, refer to Figure 2 It can be seen that in this embodiment, the interactive switching model includes: obtaining scene data in the online promotion video, setting multiple trigger points corresponding to the promotional products in the online promotion video based on the scene data, each trigger point corresponds to the user's behavior pattern in the online promotion video, and when the user triggers the corresponding trigger point, the storyline in the online promotion video will develop according to the set feature node branches, and the storyline is used to display the promotional product information related to the trigger point.

[0080] It should be added that, in the present embodiment, each trigger point is not only matched with a specific user behavior pattern, but also associated with a specific time window to ensure the smoothness and continuity of the user's experience in the process of watching the online promotional video. For example, if a user stays for a long time in the introduction part of a promotional product but does not perform any operation, the system may automatically trigger an in-depth introduction related to the product or the display of special discount information to attract further attention from the user.

[0081] It should also be added that, in this embodiment, the scene data includes: physical data and behavioral data; the physical data is used to record the environment, layout and props in the online promotion video, and the behavioral data is used to record the user's actions, time and interaction with other users during the process of watching the online promotion video;

[0082] Among them reference Figure 3 It can be seen that in this embodiment, the method for setting the trigger point includes:

[0083] Scan online promotional videos and obtain key frames based on scene data, extract significant features from the key frames, the significant features include: the appearance of objects and the actions of scene characters, match the promotional product features with the preset promotional product database, identify the triggered promotional products corresponding to the significant features, and set the trigger points at the key frames for the triggering promotional products based on the appearance of objects and the actions of scene characters.

[0084] It should be noted that, in this embodiment, the image acquisition technology in the prior art is used to extract key frames in online promotional videos. This image acquisition technology identifies key frames based on the complexity and change rate of video frames. Specifically, in the video stream, by comparing the pixel differences, color histograms and motion vectors between adjacent frames, those frames with large content changes and high information content are found as key frames. It should be added that, in this embodiment, the key frames contain important information of the video and the core features of the promoted products.

[0085] As a preferred embodiment, refer to Figure 4 It can be seen that in this embodiment, the method for creating an online promotional video includes: based on the characteristics of the promoted product, retrieving relevant film and television materials about the promoted product on the Internet, and extracting scene clips that match the characteristics of the promoted product, and constructing a combined insertion model. The combined insertion model is used to splice and combine the extracted scene clips based on the promoted product to construct one or more coherent online promotional video drafts.

[0086] It should be added that, as a supplement to the method for creating an online promotional video, the method for creating an online promotional video also includes: based on the characteristics of the promoted product, retrieving relevant text materials about the promoted product on the Internet, based on the text materials, manually polishing a storyline that conforms to the characteristics of the promoted product, matching the polished storyline with the draft online promotion video, and adjusting the scene clips, rhythm and music of the draft online promotion video to obtain the storyline content, preparing an auxiliary video based on the content, and merging the auxiliary video with the draft online promotion video to form the online promotion video.

[0087] It should be noted that in this embodiment, in order to further improve the effect of the online promotion video, a user feedback mechanism can also be introduced. That is, after the user watches the online promotion video, the user feedback mechanism will collect the user's feedback data, including viewing time, number of views, number of likes and comment content. These data will be used as the basis for evaluating the effect of the online promotion video and provide a reference for subsequent promotion video optimization.

[0088] Since each user group has different needs for promotional videos, the user groups can be segmented and exclusive promotional videos can be customized for each segment. For example, for young user groups, a more fashionable and dynamic video style and soundtrack can be chosen, while for middle-aged and elderly user groups, a more stable and traditional video style can be chosen.

[0089] It should also be added that, in the present embodiment, the combined insertion model conducts an in-depth analysis of the core features of the promoted product, including the product's functions, target user groups, and market positioning. Then, the combined insertion model will quickly screen out scene clips that are highly relevant to the features of the promoted product from a large amount of online film and television footage through keyword search and tag matching technology. During the screening process, the model will take into account the visual effects, emotional atmosphere, and user acceptance of the scene clips to ensure that the selected clips can intuitively display the characteristics and advantages of the promoted product. At the same time, the combined insertion model will also optimize the duration, rhythm, and music of the scene clips to ensure that they are in harmony with the overall style of the promoted product. After selecting suitable scene clips, the combined insertion model will splice and combine these clips. During the splicing process, the model will take into account the coherence of the storyline and the smoothness of user viewing to ensure that the final generated online promotional video can attract the user's attention and resonate with them.

[0090] In addition, it should be added that the combined insertion model also supports real-time preview and adjustment of the generated promotional video. If the user or the promotion team is not satisfied with a part of the video, they can use the tools provided by the model to quickly modify and optimize it until a satisfactory effect is achieved.

[0091] As a preferred embodiment, refer to Figure 5 It can be seen that in this embodiment, the association retrieval model includes: obtaining metadata information of network video content, the metadata information includes the theme, tags and keywords of the video, analyzing the metadata information based on a machine learning algorithm, extracting the theme features of the video content and the user preference features, constructing a feature vector space, mapping the theme features of the video content and the user preference features into the feature vector space, and forming feature vectors of the video content and user preferences, constructing a comparison model, the comparison model calculates the similarity between the video content feature vector and the user preference feature vector, and outputs the comparison result, and based on the comparison result, retrieving the predicted network video based on the user operation data in the network.

[0092] It should be noted that the comparison model in this embodiment is based on the cosine similarity algorithm, and measures the similarity of two feature vectors by calculating the cosine value of the angle between them. The closer the cosine value of the angle is to 1, the closer the directions of the two feature vectors are, that is, the more similar the theme features of the video content and the user preference features are. Conversely, the closer the cosine value of the angle is to -1, the more opposite the directions of the two feature vectors are, that is, the more dissimilar the theme features of the video content and the user preference features are. The cosine similarity algorithm is a common data processing comparison algorithm at this stage, so the comparison model proposed in this application is also a mature computing model existing at this stage, so the comparison model will not be explained in detail.

[0093] As a preferred implementation, in this implementation, user personalized matching includes: obtaining the user's network behavior data, including the user's browsing history, search history, type and duration of video viewing, analyzing the network behavior data based on big data analysis, extracting the user's interest preferences and demand characteristics, constructing a user IP based on the user's interest preferences and demand characteristics, and the user IP is used to identify the user's unique attributes and needs; based on the promotion product list, retrieving the characteristics of the user IP, screening out promotion products that match the user's interest preferences and demand characteristics, and forming a personalized promotion product list.

[0094] As a preferred embodiment, refer to Figure 6 It can be seen that in this embodiment, the main collection model includes: collecting predicted network video features, the predicted network video features include the theme, style, emotion and potential audience group information of the video, based on the collected predicted network video features, constructing a feature mapping table, the feature mapping table is used to associate the network video features with the promotion product features to form a mapping relationship between the predicted network video and the promotion product, selecting promotion products that match the predicted network video features, based on the mapping relationship, sorting the selected promotion products according to the preset promotion strategy to form a promotion product list, and dynamically adjusting the collection order of the promotion product list based on the user's behavior patterns and behavior time in the online promotion video.

[0095] In addition, it should be added that in the present application, the trigger nodes include two methods: user behavior triggering and time triggering. The user behavior triggering includes: when the user completes the behavioral operation or meets the preset conditions in the online promotion video, the trigger node is activated, thereby promoting the development of the online video plot to the display of promotional product information related to the trigger point. The time triggering means that when the online promotion video is played to the preset time point, the node is automatically triggered to embed the promotional product information into the video.

[0096] As a preferred embodiment, refer to Figure 7 It can be seen that the present invention provides a technical solution: an interactive network video advertising promotion system, which uses the above-mentioned interactive network video advertising promotion method and includes: a user behavior analysis module for real-time monitoring and analysis of user behavior data in network promotion videos, a promotion product database module for providing data support for the selection and matching of promotion products, a video content generation module, which is based on the promotion product features and is used to automatically retrieve and extract relevant video materials in the network, construct a coherent network promotion video draft through a combined insertion model, and finally generate a complete network promotion video in combination with the polishing and adjustment of the storyline, an associated retrieval module for analyzing the metadata information of the video content using a machine learning algorithm, and a personalized matching module for analyzing the content of the video content. Analyze the user's network behavior data, extract the user's interest preferences and demand characteristics, build the user IP, and based on the promotion product list, screen out the promotion products that match the user's interest preferences and demand characteristics to form a personalized promotion product list. The trigger node management module is used to manage the trigger nodes in the interactive switching model, including user behavior triggering and time triggering. It ensures that the trigger node is activated at the appropriate time and under the appropriate conditions, and promotes the development of the network video plot to the promotion product information display related to the trigger point. The user interface display module is used to display the generated network promotion video and personalized promotion product list to the user, and at the same time provide a user interaction interface, allowing the user to trigger different feature node branches through clicks and sliding operations, promote the development of the network video plot, and enhance the user's sense of participation and experience effect.

[0097] It should be added that the user interface display module also includes: a real-time feedback module, which is used to collect real-time feedback data of users during the process of watching online promotional videos, including users' viewing time, number of views, likes, comments and shares; a data analysis unit, which is used to conduct in-depth analysis of real-time feedback data to extract users' viewing habits, preferences and acceptance of promoted products; a content optimization unit, which optimizes the content of existing online promotional videos based on the user feedback information extracted by the data analysis unit, including adjusting the video rhythm, adding new scene clips and changing the soundtrack to improve user experience and advertising effectiveness; a personalized recommendation unit, which is based on user feedback information and user IP constructed by the personalized matching module.

[0098] It should be emphasized that in the interactive network video advertising promotion system disclosed in the present embodiment, during actual operation, the user behavior analysis module monitors the user's behavior data in the network promotion video in real time, such as clicks, viewing time, pauses and playbacks, and transmits these data to the data analysis unit. The data analysis unit uses big data analysis and machine learning algorithms to conduct in-depth analysis of the user behavior data, so as to more accurately understand the user's interest preferences and demand characteristics. At the same time, the data analysis unit will also combine the data of the promotion product database module to screen out promotion products that match the user's interest preferences and demand characteristics. After receiving the promotion product selection instruction, the video content generation module will automatically retrieve and extract the relevant content in the network according to the characteristics of the promotion product. Regarding the stock footage, these materials are processed by the combination insertion model to form a coherent and attractive online promotion video draft. Then, the video content generation module will combine the polishing and adjustment of the storyline, as well as possible user feedback data, to generate the final online promotion video. After that, the association retrieval module uses machine learning algorithms to analyze the metadata information of the video content during the video content generation process to ensure the close association between the video content and the promotion product. At the same time, the association retrieval module will also dynamically adjust the selection and matching of promotion products based on the data of the user behavior analysis module. In addition, when performing personalized matching, the personalized matching module extracts the user's interest preferences and demand characteristics by analyzing the user's network behavior data, and constructs the user IP. Based on this user IP, the personalized matching module will filter out promotion products that match the user's interest preferences and demand characteristics from the promotion product list to form a personalized promotion product list.

[0099] The trigger node management module is responsible for managing the trigger nodes in the online video during the online video generation process. When triggered by user behavior or time, the trigger node is activated to promote the development of the online video plot to the display of promotional product information related to the trigger point.

[0100] Finally, the user interface display module displays the generated online promotion video and personalized promotion product list to users. Users can trigger different feature node branches by clicking and sliding, promote the development of the online video plot, and achieve in-depth interaction with the advertisement.

[0101] Although embodiments of the present invention have been shown and described, it will be appreciated by those skilled in the art that various changes, modifications, substitutions and variations may be made to the embodiments without departing from the principles and spirit of the present invention, and that the scope of the present invention is limited by the attached embodiments and their equivalents.

Claims

1. An interactive network video advertising promotion method, characterized in that: include: Obtaining user operation data based on interactive network videos, including: behavior patterns and behavior time; Based on the operation data, retrieve network video content attached with the user operation data in the network; Build an associated retrieval model to match retrieved online video content with user personalization, and retrieve predicted online videos based on user operation data within the network through real-time data analysis; Construct a subject acquisition model, which is based on the predicted network video features and outputs promotional products that match the predicted network video features; Create online promotional videos with feature node branches based on the features of the promoted products; Constructing a trigger node, where the trigger node is set on a feature branch node in the online promotion video based on the features of the promoted product; Construct an interactive switching model, which is used for users to select different feature node branches in the online promotion video by interactively triggering the promotion product, so as to promote the development of the online promotion video plot; The interactive switching model includes: Obtain scene data in the online promotion video, and set multiple trigger points corresponding to the promotion products in the online promotion video based on the scene data; Each trigger point corresponds to the user's behavior pattern in the online promotion video; When the user triggers the corresponding trigger point, the storyline in the online promotion video will develop according to the set feature node branches, and the storyline is used to display the promotion product information related to the trigger point.

2. The interactive network video advertising promotion method according to claim 1, characterized in that: The scene data includes: physical data and behavioral data; Physical data is used to record the environment, layout and props in the online promotion video; Behavioral data is used to record the user's actions, time, and interactions with other users while watching online promotional videos; The trigger point setting methods include: Scan online promotional videos, obtain key frames based on scene data, and extract significant features from the key frames. The significant features include: the appearance of objects and the actions of characters in the scene; Match the promotion product features with the preset promotion product database, identify the trigger promotion product corresponding to the significant features, and set the trigger point at the key frame based on the appearance of the object and the action of the scene character.

3. The interactive network video advertising promotion method according to claim 1, characterized in that: Methods for creating online promotional videos include: Based on the characteristics of the promoted product, search for relevant video footage about the promoted product on the Internet, and extract scene clips that match the characteristics of the promoted product; A combined insertion model is constructed, which is used to splice and combine the extracted scene fragments based on the promotional products to construct one or more coherent online promotion video drafts.

4. The interactive network video advertising promotion method according to claim 3, characterized in that: The creation methods of online promotion videos also include: Based on the characteristics of the promoted product, search the network for relevant text materials about the promoted product; Based on the text material, manually polish the storyline to match the characteristics of the promoted product; Match the polished storyline with the draft online promotion video, and adjust the scene clips, rhythm and soundtrack of the draft online promotion video; Acquire storyline content, prepare auxiliary videos based on the content, and merge the auxiliary videos with the online promotion video draft to form the online promotion video.

5. The interactive network video advertising promotion method according to claim 1, characterized in that: The associated retrieval model includes: Obtain metadata information of online video content, including the subject, tags, and keywords of the video, analyze the metadata information based on machine learning algorithms, and extract the subject features of the video content and user preference features; Constructing a feature vector space, mapping the theme features of the video content and the user preference features into the feature vector space, and forming feature vectors of the video content and the user preference; Construct a comparison model, which calculates the similarity between the video content feature vector and the user preference feature vector, and outputs the comparison result; Based on the comparison results, the predicted network videos based on the user operation data are retrieved within the network.

6. The interactive network video advertising promotion method according to claim 5, characterized in that: The user personalized matching includes: Obtain users’ online behavior data, including users’ browsing history, search history, and the types and duration of videos watched; Based on big data analysis, we analyze network behavior data to extract users’ interest preferences and demand characteristics; Based on user interest preferences and demand characteristics, build user IP, which is used to identify the user's unique attributes and needs; Based on the promotion product list, retrieve the characteristics of the user's IP, filter out promotion products that match the user's interest preferences and demand characteristics, and form a personalized promotion product list.

7. The interactive network video advertising promotion method according to claim 1, characterized in that: The subject acquisition model includes: Collect and predict network video features, including the theme, style, emotion and potential audience information of the video; Based on the collected predicted network video features, a feature mapping table is constructed, the feature mapping table is used to associate the video features with the promotion product features to form a mapping relationship between the predicted network video and the promotion product, and the promotion product matching the predicted network video features is selected. Based on the mapping relationship, the selected promotion products are sorted according to the preset promotion strategy to form a promotion product list; Based on the user's behavior patterns and behavior time in the online promotional video, the collection order of the promotional product list is dynamically adjusted.

8. The interactive network video advertising promotion method according to claim 1, characterized in that: The trigger node includes two modes: user behavior trigger and time trigger; User behavior triggers include: when a user completes a behavioral operation or meets preset conditions in an online promotional video, the trigger node is activated, thereby driving the plot of the online video to the display of promotional product information related to the trigger point. Time triggering means that when the online promotional video is played to a preset time point, the node is automatically triggered to embed the promotional product information into the video.

9. An interactive network video advertising promotion system, using an interactive network video advertising promotion method according to any one of claims 1 to 8, characterized in that: include: User behavior analysis module, used to monitor and analyze user behavior data in online promotional videos in real time; Promotional product database module, used to provide data support for the selection and matching of promotional products; Video content generation module, used to automatically retrieve and extract relevant video materials from the Internet; A correlation retrieval module is used to analyze metadata information of video content using machine learning algorithms; Personalized matching module, used to analyze the user's network behavior data, extract the user's interest preferences and demand characteristics, build the user's IP, and based on the promotion product list, screen out promotion products that match the user's interest preferences and demand characteristics, and form a personalized promotion product list; A trigger node management module is used to manage the trigger nodes in the interactive switching model; The user interface display module is used to display the generated online promotion video and personalized promotion product list to the user.

Images