User intention detection method, system and equipment based on artificial intelligence

Abstract

The invention relates to the technical field of intention detection, and discloses a user intention detection method, system and equipment based on artificial intelligence, and the method comprises the steps: obtaining the multi-modal user information of a target user in real time, and constructing a user portrait of the target user based on the multi-modal user information; obtaining a marketing target, and determining specific intention information based on the marketing target; judging whether the user portrait contains specific intention information or not, and if not, determining an information acquisition strategy based on the user portrait and the specific intention information; and detecting the specific intention of the target user by using the information acquisition strategy. According to the method, the marketing target and the user portrait are closely associated, artificial intelligence is utilized to identify whether the user portrait lacks specific intention information or not, and the specific information acquisition task is actively triggered on the basis, so that the acquisition efficiency and accuracy of the user intention information are greatly improved, the marketing strategy is optimized, and the marketing efficiency is improved. Powerful support is provided for fine customer operation and dynamic marketing of enterprises, and the competitiveness is enhanced.

Description

Technical Field

[0001] This invention relates to the field of intent detection technology, specifically to a user intent detection method, system, and device based on artificial intelligence. Background Technology

[0002] In a highly competitive business environment, effective communication between enterprises and customers, and accurate understanding of customers' deep and dynamic intentions, have become key factors in enhancing enterprise competitiveness. Existing technologies primarily rely on customers' proactive expressions (such as inquiries and purchases) or periodic, identical survey results, lacking the ability to proactively identify potential intentions based on real-time customer behavior and characteristics. This results in passive and delayed acquisition of intention information. While existing technologies can personalize push notifications to customers with different tags, the focus is on information delivery rather than proactively identifying customer intentions and triggering targeted information collection tasks based on marketing objectives. Relying on single questionnaires cannot capture customer intention signals from more natural and diverse interactions. The timing, content, and channel selection for intention probing often depend on the experience of operations personnel or pre-set rules, making it difficult to intelligently and in real-time adjust and execute probing strategies based on individual differences among customers and the dynamically changing state of their intentions.

[0003] Existing technologies generally lack a core mechanism for intelligently linking and executing marketing goals, user profiles, and proactive intent information collection tasks. This not only leads to low efficiency and inaccuracy in obtaining intent information, but also severely restricts enterprises' ability to conduct refined and dynamic customer operations. Summary of the Invention

[0004] In view of this, the present invention provides a user intention detection method, system and device based on artificial intelligence to solve the problems of low efficiency and low accuracy in obtaining user intention information.

[0005] In a first aspect, the present invention provides a user intent detection method based on artificial intelligence, the method comprising:

[0006] Acquire multimodal user information of target users in real time, and build user profiles of target users based on multimodal user information;

[0007] Obtain marketing objectives and determine specific intent messages based on those objectives;

[0008] Determine whether the user profile contains specific intent information. If not, determine the information collection strategy based on the user profile and the specific intent information.

[0009] Use information gathering strategies to detect the specific intentions of target users.

[0010] The user intent detection method based on artificial intelligence provided by this invention closely links marketing goals with user profiles, uses artificial intelligence to identify whether specific intent information is missing in the user profile, and proactively triggers targeted information collection tasks based on this, realizing a marketing goal-driven intent detection closed loop. This greatly improves the efficiency and accuracy of obtaining user intent information, optimizes marketing strategies, provides strong support for enterprises' refined customer operations and dynamic marketing, and enhances competitiveness.

[0011] In one optional implementation, the multimodal user information includes: static information data and dynamic behavioral data. The multimodal user information of the target user is acquired in real time, and a user profile of the target user is constructed based on the multimodal user information, including:

[0012] Static information data of target users is collected based on customer relationship management systems, and dynamic behavioral data of target users is obtained based on online platforms.

[0013] Multimodal user information is cleaned, structured, and standardized to obtain standard user information;

[0014] By using text analysis, behavioral sequence analysis, and interaction data analysis methods, standard user information is parsed to extract the intention information of target users and construct user profiles containing this intention information.

[0015] The user intent detection method based on artificial intelligence provided by this invention makes the profile more comprehensive and three-dimensional through multi-source data fusion, accurately reflecting user interests and potential needs. Artificial intelligence technology digs out deep intents, breaks through the limitations of surface data, and the dynamic update mechanism ensures the timeliness of the profile and adapts to changes in user behavior. It provides data support for the identification of missing intents and intelligent decision-making on collection strategies, helping enterprises to accurately grasp user dynamics, optimize operational strategies, and improve the efficiency of customer intent detection and the level of refined operation.

[0016] In one optional implementation, the information collection strategy includes: information collection channels and information interaction content;

[0017] Information collection strategies are determined based on user profiles and specific intent information, including:

[0018] Based on user profiles, determine the target users' content preferences, channel preferences, and historical interaction data;

[0019] Based on channel preferences and historical interaction data, a strategy selection model is used to determine the information collection channels with the highest probability of user response.

[0020] Based on specific intent information and the target user's content preferences, a strategy selection model is used to determine the content for information interaction.

[0021] In one alternative implementation, an information gathering strategy is used to detect the specific intentions of a target user, including:

[0022] Based on information collection channels, collect feedback data from target users regarding information interaction content;

[0023] Analyze the feedback data to infer the specific intentions of the target users.

[0024] The user intent detection method based on artificial intelligence provided by this invention selects appropriate information collection strategies based on user profiles and specific intent information, breaking through the limitations of traditional single channels and general content, realizing personalized customization of collection strategies, and significantly improving user response rate and information acquisition efficiency. By dynamically optimizing strategies with AI models, it ensures that channel and content selection matches users' real-time behavioral characteristics, accurately detects missing intent information, provides enterprises with more targeted user insights, supports refined marketing decisions, and enhances the initiative and accuracy of customer operations.

[0025] In one alternative implementation, the method further includes:

[0026] Based on the information collection strategy, the response rate of the target users is collected, and the matching degree between the target users' specific intentions and specific intention information is calculated.

[0027] The strategy selection model is optimized based on response rate and matching degree.

[0028] The user intention detection method based on artificial intelligence provided by this invention optimizes the strategy selection model by collecting response rates and calculating matching degrees. It can dynamically adjust the channel and content selection logic based on real feedback data, making the strategy more in line with user behavior characteristics, improving the response rate of collection tasks and the accuracy of information acquisition, forming a closed-loop iterative mechanism of "collection-analysis-optimization", and continuously enhancing the system's ability to detect users' dynamic intentions.

[0029] In one alternative implementation, the method further includes:

[0030] Update the multimodal user information database of the target users using the specific intention information of the target users;

[0031] User profiles are optimized based on the updated multimodal user information database.

[0032] The user intent detection method based on artificial intelligence provided by this invention updates the multimodal user information database and optimizes the profile using specific intent information. It can integrate newly acquired intent data into the user profile system, enrich the profile dimensions while ensuring data timeliness, making the profile more accurately reflect the user's current state, providing more reliable data support for subsequent intent detection and marketing decisions, and helping enterprises achieve more refined customer operations and dynamic marketing.

[0033] Secondly, the present invention provides an artificial intelligence-based user intent detection system, the system comprising:

[0034] The profile building module is used to acquire multimodal user information of target users in real time and build user profiles of target users based on multimodal user information;

[0035] The specific intent information determination module is used to obtain marketing objectives and determine specific intent information based on the marketing objectives;

[0036] The information collection strategy determination module is used to determine whether the user profile contains specific intention information. If not, the information collection strategy is determined based on the user profile and the specific intention information.

[0037] The intent detection module is used to detect the specific intentions of target users using information collection strategies.

[0038] Thirdly, the present invention provides a computer device, comprising: a memory and a processor, the memory and the processor being communicatively connected to each other, the memory storing computer instructions, and the processor executing the computer instructions to perform the method described in the first aspect or any corresponding embodiment thereof.

[0039] Fourthly, the present invention provides a computer-readable storage medium storing computer instructions for causing a computer to perform the method described in the first aspect or any corresponding embodiment thereof.

[0040] Fifthly, the present invention provides a computer program product, including computer instructions for causing a computer to perform the method described in the first aspect or any corresponding embodiment thereof. Attached Figure Description

[0041] To more clearly illustrate the specific embodiments of the present invention or the technical solutions in the prior art, the drawings used in the description of the specific embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are some embodiments of the present invention. For those skilled in the art, other drawings can be obtained from these drawings without creative effort.

[0042] Figure 1 This is a flowchart illustrating a user intent detection method based on artificial intelligence according to an embodiment of the present invention;

[0043] Figure 2 This is a schematic diagram of the process of collecting intention information based on marketing objectives in the AI-based user intention detection method according to an embodiment of the present invention;

[0044] Figure 3 This is a flowchart illustrating another user intent detection method based on artificial intelligence according to an embodiment of the present invention;

[0045] Figure 4 This is a structural block diagram of an artificial intelligence-based user intent detection system according to an embodiment of the present invention;

[0046] Figure 5 This is a schematic diagram of the hardware structure of a computer device according to an embodiment of the present invention. Detailed Implementation

[0047] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0048] This invention provides a user intent detection method, system, and device based on artificial intelligence. By closely linking marketing goals with user profiles, it proactively triggers targeted intent detection to improve the efficiency and accuracy of obtaining user intent information.

[0049] According to an embodiment of the present invention, an embodiment of a user intent detection method based on artificial intelligence is provided. It should be noted that the steps shown in the flowchart in the accompanying drawings can be executed in a computer system such as a set of computer-executable instructions. Furthermore, although a logical order is shown in the flowchart, in some cases, the steps shown or described may be executed in a different order than that shown here.

[0050] This embodiment provides an artificial intelligence-based user intent detection method, which can be used in the aforementioned computer equipment. Figure 1 This is a flowchart of an artificial intelligence-based user intent detection method according to an embodiment of the present invention, such as... Figure 1 As shown, the process includes the following steps:

[0051] Step S101: Obtain multimodal user information of the target user in real time, and construct a user profile of the target user based on the multimodal user information.

[0052] Specifically, target users are generally potential or past customers. By integrating manually maintained basic information (including but not limited to name, contact information, purchase history, etc.) and summarizing automatically collected customer behavior records across the entire platform (including but not limited to browsing history, click operations, purchase process, interactive feedback, etc.), multimodal user information of target users can be comprehensively obtained.

[0053] Artificial intelligence (AI) technology is used to deeply analyze the collected multimodal user information (including but not limited to chat logs, comments, questionnaire answers, and short video interaction behavior) to infer the target user's interests, consumption habits, potential purchase intentions, and emotional tendencies toward specific products / activities, etc., and to construct a user profile that includes intention / preference indicators. The user profile also includes the target user's basic characteristics and the current known or inferred user intention information status, providing a data foundation for subsequent intention detection.

[0054] Step S102: Obtain marketing objectives and determine specific intent information based on marketing objectives.

[0055] Specifically, the system acquires the marketing goals set by the enterprise, such as promoting new product X, improving user Y's retention rate, and detecting user Z's willingness to participate in activity A. These are just examples, but not the only possibilities. Operations personnel input specific marketing goals through the system interface, such as "detecting users' purchase intention for the upcoming smartphone model X," "identifying user groups interested in high-end skincare products but not yet purchasing," and "understanding churned users' suggestions for service improvement." The system transforms these marketing goals into an internal representation that the AI ​​model can understand. For example, it identifies the product / service type associated with the goal (smartphone X, high-end skincare products), the target user group (potential buyers, churned users), and the specific intention types to be detected (purchase intention, interest, suggestions). For specific intention information required for the current marketing goal, such as knowing whether users' specific needs for X-type products are gaming or office work, or knowing what dissatisfaction or expectations Y has regarding the current service, specific intention information could be potential buyers' purchase intention for smartphone X. This is just an example, but not the only possibility.

[0056] like Figure 2 The diagram illustrates the process of collecting intent information based on marketing objectives in an AI-based user intent detection method. It begins by breaking down the marketing objectives to determine the required number of users, then analyzing and segmenting them into subgroups such as A, B, and C. For each subgroup, missing information in their user profiles, such as consumption preferences and behavioral habits, is identified. Then, suitable content is generated for collection, such as questionnaires and advertorials, and information is assessed through user interaction. Data is then collected by reaching users through multiple channels, including WeChat private domains and mobile apps, ultimately supplementing the user profiles with new information to make them more accurate and support subsequent business decisions and marketing.

[0057] Step S103: Determine whether the user profile contains specific intent information. If not, determine the information collection strategy based on the user profile and the specific intent information.

[0058] Specifically, AI models (such as rule-based, decision tree, or more complex reinforcement learning models) are used to analyze dynamic user profiles based on marketing objectives, traversing or filtering users relevant to the marketing goals. For each user, the system intelligently identifies which target customer groups lack sufficient known or inferred intent information to support decisions regarding that marketing objective—that is, specific intent information is missing. Based on the user profile, it assesses whether the currently known or inferred intent information meets the marketing objective's need to know the user's intent. For example, if the objective is to probe purchase intention, does the profile contain clear information about the user's recent purchase plans, opinions on competitors, or budget range? If not, or if the information is uncertain, it is determined that intent information is missing.

[0059] Once missing intent information is identified, an intent information collection task for that user is automatically triggered. For users determined to have missing intent information, the system automatically generates a specific intent information collection task for that user. The task includes: target user ID, the type of intent information to be collected (e.g., purchase intention for product X), trigger time, etc. The system can prioritize tasks based on factors such as user value and the urgency of missing intent information. This is just an example and is not a limitation.

[0060] Based on the currently known intention / preference indicators, activity levels, and channel usage habits in the user profile, determine the user profile characteristics (e.g., which channels users prefer, which content formats they respond to better, and their active time). Determine the collection task requirements based on the specific intention information types to be collected, and then intelligently match the optimal collection and interaction strategy from the preset strategy library, including the optimal information collection channels (e.g., APP push, WeChat, stores, telephone, etc.) and the optimal collection content type / format (e.g., customized questionnaires, short videos on specific themes, targeted dialogue scripts, etc.).

[0061] For example, if it is necessary to detect a user's preference for a certain visual feature, a short video containing that feature may be pushed to them; if it is necessary to understand a user's opinion on the terms of service, a customized questionnaire may be pushed to them; if it is necessary to gain a deeper understanding of complex needs, a targeted dialogue script may be generated to trigger customer service communication. These are just examples, but are not limited to this.

[0062] Step S104: Use information collection strategies to detect the specific intentions of target users.

[0063] Specifically, the process involves receiving raw response data and performing structured processing. Advanced data analytics tools and AI algorithms (such as classification, regression, natural language understanding, and behavioral sequence analysis) are then used to conduct in-depth analysis of the response data. For example, statistical analysis and text mining are performed on questionnaire answers; for short video interaction data, behavioral analysis models and fine-grained feature extraction are used to infer users' interest in and potential intentions regarding specific content in the videos. Specific, missing, and specific intention information about users is inferred from the response data (e.g., "The user intends to use product X for gaming, not for office work," or "The user shows a high willingness to participate in activity A").

[0064] The AI-based user intent detection method provided in this embodiment closely links marketing goals with user profiles. It uses AI to identify whether specific intent information is missing in the user profile and proactively triggers targeted information collection tasks based on this. This achieves a marketing goal-driven intent detection closed loop, greatly improving the efficiency and accuracy of user intent information acquisition, optimizing marketing strategies, providing strong support for enterprises' refined customer operations and dynamic marketing, and enhancing competitiveness.

[0065] This embodiment provides an artificial intelligence-based user intent detection method, which can be used in the aforementioned computer system. Figure 3 This is a flowchart of an artificial intelligence-based user intent detection method according to an embodiment of the present invention, such as... Figure 3 As shown, the process includes the following steps:

[0066] Step S201: Obtain multimodal user information of the target user in real time, and construct a user profile of the target user based on the multimodal user information.

[0067] Specifically, the multimodal user information includes: static information data and dynamic behavioral data, and step S201 above includes:

[0068] Step S2011: Collect static information data of target users based on the customer relationship management system, and obtain dynamic behavior data of target users based on the online platform.

[0069] Specifically, it involves comprehensively collecting static information data and dynamic behavioral data of target users. This includes integrating static customer information and transaction records from the Customer Relationship Management (CRM) system; and collecting behavioral logs from online platforms (official website, APP, mini-program), such as browsing, clicks, searches, shopping carts, orders, reviews, customer service chat logs, social media interaction data, and historical intention collection interaction data (such as questionnaires and activity participation feedback).

[0070] Step S2012 involves cleaning, structuring, and standardizing the multimodal user information to obtain standard user information.

[0071] Specifically, data from different sources may have different types and formats. Multi-source heterogeneous data are cleaned, structured, and standardized in sequence to obtain standard user information, which is then collected into the data warehouse.

[0072] Step S2013: Use text analysis, behavior sequence analysis, and interaction data analysis methods to parse standard user information, extract the intention information of target users, and construct a user profile containing intention information.

[0073] Specifically, multimodal user information includes, but is not limited to, text data, behavioral sequences, and interaction data. AI technology is used for in-depth mining and analysis to initially infer the target user's interests, behavioral patterns, and potential intentions or inclinations towards specific areas, forming intention / preference indicators in the user profile. Through data fusion and cleaning, a user profile with rich dimensions and dynamic updates is constructed.

[0074] By using natural language processing (NLP) technology to analyze text data, such as chat logs, comments, and reviews, we can use methods such as sentiment analysis, topic modeling, and keyword extraction to identify the products, services, and activities mentioned by target users, determine their sentiment tendencies (positive, negative, neutral), extract their focus on specific topics, and predict their inclinations toward certain categories of products or services.

[0075] Behavioral sequence analysis techniques such as machine learning / deep learning (ML / DL) are used to analyze behavioral sequence data. Sequence models (such as RNN, Transformer) or association rule mining are applied to user browsing, click, search, and purchase sequence data to discover user behavioral patterns and infer their potential purchase intentions or preferences for specific categories / styles (for example, users who frequently browse high-end products and pay attention to new product releases may have a high willingness to try new things and purchasing power).

[0076] By using interactive data analysis methods such as machine learning, we can analyze historical questionnaire responses, activity participation records, and content interaction data (likes, shares, favorites) to directly or indirectly obtain some of the user's known preferences and intentions.

[0077] The extracted intention / preference pre-indicators are integrated with customer basic information and behavioral characteristics to form a user profile that includes the intention dimension. The system continuously monitors new data inflows and runs AI models periodically or in real time to update the profiles, ensuring their dynamism and timeliness.

[0078] The user intent detection method based on artificial intelligence provided in this embodiment makes the profile more comprehensive and three-dimensional through multi-source data fusion, accurately reflecting user interests and potential needs. Artificial intelligence technology digs out deep intents, breaks through the limitations of surface data, and the dynamic update mechanism ensures the timeliness of the profile and adapts to changes in user behavior. It provides data support for the identification of missing intents and intelligent decision-making on collection strategies, helping enterprises to accurately grasp user dynamics, optimize operational strategies, and improve the efficiency of customer intent detection and the level of refined operation.

[0079] Step S202: Obtain marketing objectives and determine specific intent information based on these objectives. See details below. Figure 1 Step S101 of the illustrated embodiment will not be described again here.

[0080] Step S203: Determine whether the user profile contains specific intent information. If not, determine the information collection strategy based on the user profile and the specific intent information.

[0081] Specifically, the information collection strategy includes: information collection channels and information interaction content, and step S203 above includes:

[0082] Step S2031: Determine the target user's content preferences, channel preferences, and historical interaction data based on the user profile.

[0083] Specifically, user profiles are used to determine the target user's content preferences, channel preferences, historical interaction data, etc., such as user profile data:

[0084] User profile A on an e-commerce platform shows: 25-year-old female, working in internet operations, with an upper-middle income; in the past 30 days, she has browsed over 10 beauty brand serum products, saved 2 serums that emphasize "anti-aging" and "moisturizing" effects, watched 5 related product review short videos, and left comments asking "Is it suitable for sensitive skin?" Based on this user profile, it can be determined that this user has a significant preference for products related to "beauty and skincare" and "anti-aging effects," and pays more attention to "ingredient safety" (suitability for sensitive skin) and "detailed functional reviews." Her preferred channel is short videos, which is only an example and not a limitation.

[0085] Step S2032: Based on channel preferences and historical interaction data, use a strategy selection model to determine the information collection channel with the highest probability of user response.

[0086] Specifically, a diversified interactive channel system is established, enabling interaction with customers through various means such as push notifications from user-end apps, WeChat official accounts / mini-programs, on-site communication at offline stores, and telephone calls.

[0087] Based on historical data, we learn the target users' open rates, interaction rates, and response rates across different channels, and select the channel most likely to succeed in the current task as the optimal information collection channel (e.g., push notifications to users who prefer to use the app, and phone calls or physical stores to elderly users). We then use the optimal information collection channel to perform the data collection and interaction tasks and collect customer response data in real time.

[0088] Step S2033: Based on specific intention information and the target user's content preferences, determine the information interaction content using a strategy selection model.

[0089] Specifically, based on the type of intent to be collected (e.g., whether direct answers or behavioral inference are needed) and content preferences in user profiles, the AI ​​can choose from formats such as questionnaires, targeted dialogue scripts, and short videos on specific themes. In addition to information collection channels and interactive content, a timing and frequency model can be established to select the optimal timing for push notifications or contacts based on user activity times and historical interaction frequencies. For example, to explore young users' preferences for the appearance of trendy product X, and given that their user profile shows a preference for short videos, AI might choose to push a short video showcasing different appearances of product X at night (collected content) via the app (channel). For understanding elderly users' risk preferences for financial products, AI might generate a simple and easy-to-understand questionnaire (collected content) and conduct interviews via telephone (channel).

[0090] The questionnaire / script generation process includes: a built-in question template library for different intention types (such as purchase intention, usage feedback, service satisfaction, and willingness to participate in activities). When the AI ​​decision-making module specifies that a certain type of intention information needs to be collected and selects a questionnaire / dialogue format, the content generation module selects or combines relevant questions from the template library based on user profiles and specific product / activity information. It also adjusts the language style and question complexity according to the user profile to generate personalized, targeted questionnaires or dialogue scripts. For example, a questionnaire to probe purchase intention might include: "How interested are you in [Product X]? (Options: Very interested, Somewhat interested, Not interested)", "When do you expect to consider purchasing [Product X]? (Options: Within 1 month, 1-3 months, After 3 months, Uncertain)", and "What are the main factors influencing your purchase of [Product X]? (Short answer / Multiple choice)". This is just an example and is not a limitation.

[0091] The probing short video generation process includes: the system has a built-in video material library with different product features and marketing themes (product close-ups, function demonstrations, user scenarios, promotional information, etc.). When the AI ​​decision-making module selects a short video as the collection format, the content generation module selects relevant materials (product silent demonstration clips) from the material library according to the specific intention type to be detected (e.g., detecting interest in the silent performance of product X), and edits, adds music, and adds text descriptions based on the user profile's visual preferences, video length preferences, etc. The key is to design one or more "intent stimuli" in the video (such as repeatedly showing a certain detail, or lingering longer after a function demonstration). The generated video is designed to trigger specific behavioral feedback from the user during viewing (watching a specific segment for a certain duration, pausing, repeating playback, commenting on specific content).

[0092] Step S204: Use information collection strategies to detect the specific intentions of target users.

[0093] Specifically, step S204 includes:

[0094] Step S2041: Collect feedback data from target users regarding information interaction content based on information collection channels.

[0095] Specifically, the multi-channel information collection and interaction module will collect raw response data (such as questionnaire submission results, the distribution of user viewing time for short videos, interaction records at specific times in the video, and dialogue text records). Based on marketing objectives, the specific intention information to be collected, and user profile characteristics, it will flexibly design personalized questionnaire content (including questions probing specific intentions), targeted dialogue scripts, etc., aiming to directly obtain customers' specific intention information. Strictly following the instructions issued by the AI ​​intention detection and decision-making module, it will accurately and efficiently distribute carefully designed interactive content / collection tasks for collecting specific intention information to target customers and collect customer response data (questionnaire answers, short video viewing and interaction data, dialogue content, etc.). Through in-depth analysis of user interests and preferences and the specific intentions to be detected, it will create and push short video content containing specific stimuli (such as showcasing specific features of product X, or posing questions about activity A). By leveraging fine-grained monitoring and analysis of user short video viewing and interaction behavior (such as viewing time, number of times specific segments are replayed, likes / comments on specific stimuli, etc.), it will infer users' reactions to the stimuli, thereby detecting their specific intentions and interests.

[0096] Step S2042: Analyze the feedback data to infer the specific intentions of the target users.

[0097] Specifically, statistical analysis is performed on the answers to structured questionnaires to calculate the distribution of responses to specific intention questions; NLP technology is used to perform semantic understanding on the dialogue text to extract the user's expressed views, needs, and concerns, and directly infer the user's specific intentions.

[0098] Using behavioral analysis models, fine-grained viewing behaviors of users in exploratory short videos are analyzed. For example, if a user repeatedly watches a demonstration clip of product X's silent performance, it is inferred that the user has a high level of attention and interest in the product's silent performance; if a user pauses or rewinds when seeing promotional information, it is inferred that the user is price-sensitive or interested in the promotion. These behaviors are converted into evidence of intent inference. Based on this inference evidence, the specific intentions of the target users are determined.

[0099] The AI-based user intent detection method provided in this embodiment selects appropriate information collection strategies based on user profiles and specific intent information, breaking through the limitations of traditional single channels and general content. It achieves personalized customization of collection strategies, significantly improving user response rates and information acquisition efficiency. By dynamically optimizing strategies with AI models, it ensures that channel and content selection matches users' real-time behavioral characteristics, accurately detects missing intent information, provides enterprises with more targeted user insights, supports refined marketing decisions, and enhances the initiative and accuracy of customer operations.

[0100] In some alternative implementations, the method further includes:

[0101] Based on the information collection strategy, the response rate of the target users is collected, and the matching degree between the target users' specific intentions and specific intention information is calculated.

[0102] The strategy selection model is optimized based on response rate and matching degree.

[0103] Specifically, based on large-scale intention inference results and collection efficiency data (such as response rate and collection success rate under different strategies), the strategy selection model can learn better intention information missing judgment rules, collection strategy selection weights, and improve the generation logic of collection content. For example, it can discover that some questions are not effective in questionnaires, and that some video stimuli are more effective in stimulating interaction, thereby optimizing the efficiency and accuracy of the entire intention detection process.

[0104] Based on the feedback from the intention inference results and the collection process, the strategies of the intention detection and decision-making modules are optimized. For example, the judgment logic for missing intentions is adjusted, the collection strategy selection model is optimized, and the collection content generation rules are improved to form a closed loop and continuously improve the efficiency and accuracy of intention detection.

[0105] The user intent detection method based on artificial intelligence provided in this embodiment optimizes the strategy selection model by collecting response rates and calculating matching degrees. It can dynamically adjust the channel and content selection logic based on real feedback data, making the strategy more in line with user behavior characteristics, improving the response rate of collection tasks and the accuracy of information acquisition, forming a closed-loop iterative mechanism of "collection-analysis-optimization", and continuously enhancing the system's ability to detect users' dynamic intentions.

[0106] In some alternative implementations, the method further includes:

[0107] Update the multimodal user information database of the target users using the specific intention information of the target users;

[0108] User profiles are optimized based on the updated multimodal user information database.

[0109] Specifically, specific intention information (such as "User [ID]'s purchase intention for product X: high, reason: silent performance" or "User [ID]'s participation intention for activity A: high, evidence: repeatedly watching the activity instruction video") is fed back to the dynamic user profile module in the form of structured tags or confidence scores to update the user's intention / preference indicators and improve the accuracy of the intention dimension of the user profile.

[0110] The AI-based user intent detection method provided in this embodiment updates the multimodal user information database and optimizes the user profile using specific intent information. It can integrate newly acquired intent data into the user profile system, enrich the profile dimensions while ensuring data timeliness, making the profile more accurately reflect the user's current state, providing more reliable data support for subsequent intent detection and marketing decisions, and helping enterprises achieve more refined customer operations and dynamic marketing.

[0111] This embodiment also provides an artificial intelligence-based user intent detection system, which is used to implement the above embodiments and preferred embodiments; details already described will not be repeated. As used below, the term "module" can be a combination of software and / or hardware that performs a predetermined function. Although the systems described in the following embodiments are preferably implemented in software, hardware implementations, or a combination of software and hardware, are also possible and contemplated.

[0112] This embodiment provides a user intent detection system based on artificial intelligence, such as Figure 4 As shown, the system includes:

[0113] The profile building module 401 is used to acquire multimodal user information of the target user in real time and build a user profile of the target user based on the multimodal user information.

[0114] The specific intent information determination module 402 is used to obtain marketing objectives and determine specific intent information based on the marketing objectives.

[0115] The information collection strategy determination module 403 is used to determine whether the user profile contains specific intention information. If not, it determines the information collection strategy based on the user profile and the specific intention information.

[0116] The intention detection module 404 is used to detect the specific intentions of a target user using information collection strategies.

[0117] In some alternative implementations, the image construction module 401 includes:

[0118] The data acquisition unit is used to collect static information data of target users based on the customer relationship management system and to acquire dynamic behavioral data of target users based on the online platform.

[0119] The data preprocessing unit is used to clean, structure, and standardize multimodal user information to obtain standard user information.

[0120] The data parsing unit is used to analyze standard user information using text analysis, behavioral sequence analysis, and interactive data analysis methods, extract the intention information of target users, and construct user profiles containing the intention information.

[0121] In some optional implementations, the information acquisition strategy determination module 403 includes:

[0122] The data determination unit is used to determine the target user's content preferences, channel preferences, and historical interaction data based on the user profile.

[0123] The channel selection unit is used to determine the information collection channel with the highest probability of user response based on channel preferences and historical interaction data using a strategy selection model.

[0124] The content determination unit is used to determine the information interaction content based on specific intention information and the content preferences of target users using a strategy selection model.

[0125] In some alternative implementations, the intention detection module 404 includes:

[0126] The feedback data collection unit is used to collect feedback data from target users regarding information interaction content based on information collection channels.

[0127] The intention inference unit is used to analyze feedback data and infer the specific intentions of the target user.

[0128] Further functional descriptions of the above modules and units are the same as those in the corresponding embodiments described above, and will not be repeated here.

[0129] In this embodiment, the user intent detection system based on artificial intelligence is presented in the form of functional units. Here, a unit refers to an ASIC (Application Specific Integrated Circuit), a processor and memory that execute one or more software or fixed programs, and / or other devices that can provide the above functions.

[0130] This invention also provides a computer device having the above-described features. Figure 3 The system shown is an AI-based user intent detection system.

[0131] Please see Figure 5 , Figure 5 This is a schematic diagram of the structure of a computer device provided in an optional embodiment of the present invention, such as... Figure 5 As shown, the computer device includes one or more processors 10, memory 20, and interfaces for connecting the components, including high-speed interfaces and low-speed interfaces. The components communicate with each other via different buses and can be mounted on a common motherboard or otherwise installed as needed. The processors can process instructions executed within the computer device, including instructions stored in or on memory to display graphical information of a GUI on external input / output devices (such as display devices coupled to the interfaces). In some alternative implementations, multiple processors and / or multiple buses can be used with multiple memories and multiple memory modules, if desired. Similarly, multiple computer devices can be connected, each providing some of the necessary operations (e.g., as a server array, a group of blade servers, or a multiprocessor system). Figure 5 Take a processor 10 as an example.

[0132] Processor 10 may be a central processing unit, a network processor, or a combination thereof. Processor 10 may further include a hardware chip. The hardware chip may be an application-specific integrated circuit (ASIC), a programmable logic device (PLD), or a combination thereof. The programmable logic device may be a complex programmable logic device (CAMP), a field-programmable gate array (FPGA), a general-purpose array logic (GDA), or any combination thereof.

[0133] The memory 20 stores instructions executable by at least one processor 10 to cause the at least one processor 10 to perform the method shown in the above embodiments.

[0134] The memory 20 may include a program storage area and a data storage area. The program storage area may store the operating system and applications required for at least one function; the data storage area may store data created based on the use of the computer device. Furthermore, the memory 20 may include high-speed random access memory and may also include non-transitory memory, such as at least one disk storage device, flash memory device, or other non-transitory solid-state storage device. In some alternative embodiments, the memory 20 may optionally include memory remotely located relative to the processor 10, and these remote memories may be connected to the computer device via a network. Examples of such networks include, but are not limited to, the Internet, intranets, local area networks, mobile communication networks, and combinations thereof.

[0135] The memory 20 may include volatile memory, such as random access memory; the memory may also include non-volatile memory, such as flash memory, hard disk or solid-state drive; the memory 20 may also include a combination of the above types of memory.

[0136] The computer device also includes a communication interface 30 for communicating with other devices or communication networks.

[0137] This invention also provides a computer-readable storage medium. The methods described above according to embodiments of the invention can be implemented in hardware or firmware, or implemented as computer code that can be recorded on a storage medium, or implemented as computer code downloaded via a network and originally stored on a remote storage medium or a non-transitory machine-readable storage medium and then stored on a local storage medium. Thus, the methods described herein can be processed by software stored on a storage medium using a general-purpose computer, a dedicated processor, or programmable or dedicated hardware. The storage medium can be a magnetic disk, optical disk, read-only memory, random access memory, flash memory, hard disk, or solid-state drive, etc.; further, the storage medium can also include combinations of the above types of memory. It is understood that computers, processors, microprocessor controllers, or programmable hardware include storage components capable of storing or receiving software or computer code, which, when accessed and executed by the computer, processor, or hardware, implements the methods shown in the above embodiments.

[0138] A portion of this invention can be applied as a computer program product, such as computer program instructions, which, when executed by a computer, can invoke or provide the methods and / or technical solutions according to the invention through the operation of the computer. Those skilled in the art will understand that the forms in which computer program instructions exist in a computer-readable medium include, but are not limited to, source files, executable files, installation package files, etc. Correspondingly, the ways in which computer program instructions are executed by a computer include, but are not limited to: the computer directly executing the instructions, or the computer compiling the instructions and then executing the corresponding compiled program, or the computer reading and executing the instructions, or the computer reading and installing the instructions and then executing the corresponding installed program. Here, the computer-readable medium can be any available computer-readable storage medium or communication medium accessible to a computer.

[0139] Although embodiments of the invention have been described in conjunction with the accompanying drawings, those skilled in the art can make various modifications and variations without departing from the spirit and scope of the invention, and such modifications and variations all fall within the scope defined by the appended claims.

Claims

1. A user intent detection method based on artificial intelligence, characterized in that, The method includes: Real-time acquisition of multimodal user information of target users, and construction of user profiles of target users based on the multimodal user information; Obtain marketing objectives and determine specific intent messages based on those objectives; Determine whether the user profile contains the specific intention information; if not, determine the information collection strategy based on the user profile and the specific intention information. The information collection strategy is used to detect the specific intentions of the target user.

2. The method according to claim 1, characterized in that, Multimodal user information includes: static information data and dynamic behavioral data. The real-time acquisition of multimodal user information of the target user, and the construction of a user profile of the target user based on the multimodal user information, includes: Static information data of target users is collected based on customer relationship management systems, and dynamic behavioral data of target users is obtained based on online platforms. The multimodal user information is cleaned, structured, and standardized to obtain standard user information; The standard user information is parsed using text analysis, behavioral sequence analysis, and interaction data analysis methods to extract the target user's intention information and construct a user profile containing the intention information.

3. The method according to claim 1, characterized in that, The information collection strategy includes: information collection channels and information interaction content; Determining an information collection strategy based on the user profile and the specific intent information includes: Based on user profiles, determine the target users' content preferences, channel preferences, and historical interaction data; Based on the channel preferences and historical interaction data, a strategy selection model is used to determine the information collection channel with the highest probability of user response. Based on the specific intent information and the target user's content preferences, a strategy selection model is used to determine the content for information interaction.

4. The method according to claim 3, characterized in that, Utilizing the information collection strategy to detect the specific intentions of the target user includes: Based on information collection channels, collect feedback data from target users regarding information interaction content; The feedback data is analyzed to infer the specific intentions of the target users.

5. The method according to claim 4, characterized in that, The method further includes: Based on the information collection strategy, the response rate of the target users is collected, and the matching degree between the target users' specific intentions and specific intention information is calculated. The strategy selection model is optimized based on the response rate and the matching degree.

6. The method according to claim 1, characterized in that, The method further includes: Update the target user's multimodal user information database using the target user's specific intention information; The user profile is optimized based on the updated multimodal user information database.

7. A user intent detection system based on artificial intelligence, characterized in that, The system includes: The profile building module is used to acquire multimodal user information of the target user in real time and build a user profile of the target user based on the multimodal user information. The specific intent information determination module is used to obtain marketing objectives and determine specific intent information based on the marketing objectives; The information collection strategy determination module is used to determine whether the user profile contains the specific intention information. If it does not contain it, the information collection strategy is determined based on the user profile and the specific intention information. An intention detection module is used to detect the specific intentions of the target user using the information collection strategy.

8. A computer device, characterized in that, include: A memory and a processor, the memory and the processor being communicatively connected to each other, the memory storing computer instructions, the processor executing the computer instructions to perform the method of any one of claims 1 to 6.

9. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores computer instructions for causing a computer to perform the method of any one of claims 1 to 6.

10. A computer program product, characterized in that, Includes computer instructions for causing a computer to perform the method of any one of claims 1 to 6.

Images