Content recommendation method and device based on user interests, equipment and storage medium

Abstract

The invention provides a content recommendation method and device based on user interests, equipment and a storage medium, and relates to the field of artificial intelligence, in particular to the field of large model and intelligent recommendation. According to the specific implementation scheme, current scene information and a pre-stored interest model are obtained; wherein the current scene information represents a space-time scene where the user is currently located, and the interest model is a user portrait model representing a field of interested content of the user; according to the current scene information, updating the interest model to obtain an updated interest model; and according to the updated interest model and the equipment type of the terminal equipment currently used by the user, determining to-be-recommended contents, and pushing the to-be-recommended contents to the user. The content recommendation precision is improved, and the user experience is improved.

Description

Technical Field

[0001] The present disclosure relates to the field of big models and intelligent recommendation in the field of artificial intelligence, and in particular to a method, device, equipment and storage medium for content recommendation based on user interests. Background Art

[0002] Users' interests will change over time and in different scenarios, and with different interests, the content they want to watch will also be different.

[0003] Traditional content recommendation models analyze user needs based on static historical behaviors and often cannot flexibly respond to situations where user interests shift. The accuracy of content recommendations is low, which affects user experience. Summary of the invention

[0004] The present disclosure provides a content recommendation method, apparatus, device and storage medium based on user interests.

[0005] According to a first aspect of the present disclosure, a method for content recommendation based on user interests is provided, comprising:

[0006] Acquire current scene information and a pre-stored interest model; wherein the current scene information represents the spatiotemporal scene in which the user is currently located, and the interest model is a user portrait model representing the field of the user's interested content;

[0007] According to the current scene information, the interest model is updated to obtain an updated interest model;

[0008] According to the updated interest model and the device type of the terminal device currently used by the user, the content to be recommended is determined, and the content to be recommended is pushed to the user.

[0009] According to a second aspect of the present disclosure, there is provided a content recommendation device based on user interests, comprising:

[0010] An acquisition unit, used to acquire current scene information and a pre-stored interest model; wherein the current scene information represents the spatiotemporal scene in which the user is currently located, and the interest model is a user portrait model representing the field of the user's interested content;

[0011] An updating unit, configured to update the interest model according to the current scene information to obtain an updated interest model;

[0012] The recommendation unit is used to determine the content to be recommended according to the updated interest model and the device type of the terminal device currently used by the user, and push the content to be recommended to the user.

[0013] According to a third aspect of the present disclosure, there is provided an electronic device, including:

[0014] at least one processor; and

[0015] a memory communicatively coupled to the at least one processor;

[0016] The memory stores instructions that can be executed by the at least one processor, and the instructions are executed by the at least one processor to enable the at least one processor to perform the method described in the first aspect of the present disclosure.

[0017] According to a fourth aspect of the present disclosure, a non-transitory computer-readable storage medium storing computer instructions is provided, wherein the computer instructions are used to cause the computer to execute the method according to the first aspect of the present disclosure.

[0018] According to a fifth aspect of the present disclosure, a computer program product is provided, comprising a computer program, which implements the steps of the method described in the first aspect of the present disclosure when executed by a processor.

[0019] According to the technology disclosed in the present invention, the accuracy of content recommendation is improved and the user experience is enhanced.

[0020] It should be understood that the content described in this section is not intended to identify the key or important features of the embodiments of the present disclosure, nor is it intended to limit the scope of the present disclosure. Other features of the present disclosure will become easily understood through the following description. BRIEF DESCRIPTION OF THE DRAWINGS

[0021] The accompanying drawings are used to better understand the present solution and do not constitute a limitation of the present disclosure.

[0022] Figure 1 is a flowchart of a method for content recommendation based on user interests provided according to an embodiment of the present disclosure;

[0023] Figure 2 is a schematic diagram of an optimization process of an interest model provided according to an embodiment of the present disclosure;

[0024] Figure 3 is a flowchart of a method for content recommendation based on user interests provided according to an embodiment of the present disclosure;

[0025] Figure 4 is a schematic diagram of an updating process of an interest model provided according to an embodiment of the present disclosure;

[0026] Figure 5 is a flowchart of a method for content recommendation based on user interests provided according to an embodiment of the present disclosure;

[0027] Figure 6is a schematic diagram of a content recommendation process based on device type provided according to an embodiment of the present disclosure;

[0028] Figure 7 is a flowchart of a method for content recommendation based on user interests provided according to an embodiment of the present disclosure;

[0029] Figure 8 is a structural block diagram of a content recommendation device based on user interests provided according to an embodiment of the present disclosure;

[0030] Fig. 9 is a structural block diagram of a content recommendation device based on user interests provided according to an embodiment of the present disclosure;

[0031] Fig.10 is a block diagram of an electronic device for implementing the content recommendation method based on user interests according to an embodiment of the present disclosure;

[0032] Fig.11 The block diagram is a block diagram of an electronic device used to implement the content recommendation method based on user interests in an embodiment of the present disclosure. DETAILED DESCRIPTION

[0033] The following is a description of exemplary embodiments of the present disclosure in conjunction with the accompanying drawings, including various details of the embodiments of the present disclosure to facilitate understanding, which should be considered as merely exemplary. Therefore, it should be recognized by those of ordinary skill in the art that various changes and modifications may be made to the embodiments described herein without departing from the scope and spirit of the present disclosure. Similarly, for the sake of clarity and conciseness, descriptions of well-known functions and structures are omitted in the following description.

[0034] With the development of artificial intelligence technology, when users use terminal devices, it is possible to predict the content that users are interested in and automatically recommend content to users to improve user experience. However, user interests change with time and space, and traditional recommendation methods cannot flexibly cope with the instantaneous drift of user interests.

[0035] Currently, many recommendation systems analyze user needs based on their historical behaviors, which often fail to reflect users' real-time interest shifts. If users' interests change rapidly in a short period of time, traditional methods cannot capture them quickly. For example, current recommendation systems are difficult to cope with the changes in interests caused by users frequently switching between different time, space, and devices, which reduces the accuracy of content recommendations and affects user experience.

[0036] The present invention provides a method, device, equipment and storage medium for content recommendation based on user interests, which are applied to the large model and intelligent recommendation fields in the field of artificial intelligence to improve the accuracy and flexibility of content recommendation and enhance user experience.

[0037] It should be noted that the data in this embodiment is not targeted at a specific user and cannot reflect the personal information of a specific user. It should be noted that the data in this embodiment comes from a public data set.

[0038] In the technical solution of the present disclosure, the collection, storage, use, processing, transmission, provision and disclosure of user personal information involved are in compliance with the provisions of relevant laws and regulations and do not violate public order and good morals.

[0039] In order to enable readers to more deeply understand the implementation principle of the present disclosure, the following Figure 1-Figure 11 The embodiment is further refined.

[0040] Figure 1 FIG. 1 is a flow chart of a method for recommending content based on user interests according to an embodiment of the present disclosure. The method can be performed by a device for recommending content based on user interests. Figure 1 As shown, the method comprises the following steps:

[0041] S101. Obtain current scene information and a pre-stored interest model; wherein the current scene information represents the space-time scene in which the user is currently located, and the interest model is a user portrait model that represents the field of content that the user is interested in.

[0042] Exemplarily, a user can browse various contents of interest using a terminal device, for example, articles, pictures, videos, etc. Each user can have his or her own interest model, which is a user portrait model that can characterize the areas of content that the user is interested in. A user portrait can be constructed based on information such as the user's location, occupation, and historical browsing history. For example, if the user is a software development engineer, the interest model may include the technical field of software development. The user's interest model is stored to facilitate the acquisition of the interest model at any time and to make targeted content recommendations to the user.

[0043] The user's current scene information can be obtained in real time, and the current scene information can represent the user's current spatiotemporal scene, that is, determine the user's current time and space scene. For example, the current time scene is a weekday, and the space scene is an office. That is, the user's current geographic location information and time information can be obtained, and the user's current geographic location information and time information are determined as the current scene information.

[0044] S102. Update the interest model according to the current scene information to obtain an updated interest model.

[0045] Exemplarily, the current scene information may change at any time, and the currently acquired interest model may be updated in real time according to the current scene information, thereby obtaining an updated interest model. The updated interest model may be stored, that is, the stored interest model acquired each time is the latest interest model.

[0046] Based on the current scene information, the areas that the user may be currently interested in can be predicted, and based on the predicted areas, the existing areas in the interest model can be updated. For example, if the predicted area is not in the interest model, the predicted area can be added to the interest model to obtain an updated interest model. For example, if the current scene information indicates that the user is currently off work and at home, it can be predicted that the user's current area of ​​interest is leisure and entertainment, rather than technology. Leisure and entertainment areas can be added to the interest model to facilitate the recommendation of leisure and entertainment content to the user.

[0047] Different fields may be pre-associated with different scenes, and the fields associated with the current scene information may be searched according to the current scene information as the fields that the user may currently be interested in. The fields that the user is currently interested in may also be predicted according to the fields of the content that the user browsed under the current scene information in the historical time period. For example, the fields of the content that the user browsed under the same scene may be used as the fields that the user is interested in in the scene.

[0048] In this embodiment, the method also includes: obtaining the user's historical behavior data; wherein the historical behavior data includes the user's behavior data within a preset first time period and the user's behavior data within a preset second time period, the preset first time period is greater than the preset second time period, and the behavior data characterizes the user's operation behavior and operation content; based on the historical behavior data, the interest model is updated to obtain an updated interest model.

[0049] Specifically, in addition to updating the interest model according to the current scene information, the user's historical behavior data may also be obtained, and the interest model may be updated according to the user's historical behavior data.

[0050] A first time period and a second time period are preset, and the first time period can be greater than the second time period. For example, the first time period is a time period starting from the past year, and the second time period is a time period starting from the past month. That is, the span of the first time period can be greater than the span of the second time period, and the first time period can include the second time period. The historical behavior data corresponding to the first time period is the user's long-term behavior data, and the historical behavior data corresponding to the second time period is the user's short-term behavior data. For example, if the first time period is the past year and the second time period is the past month, all the user's behavior data in the past year and all the user's behavior data in the past month are obtained.

[0051] Behavioral data can characterize the user's operation behavior and operation content. Operation behavior may include web browsing, video clicking, searching, collecting, liking, etc. Operation content refers to the object on which the operation behavior acts, for example, it may be the web page browsed by the user, the post collected, etc.

[0052] Based on the historical behavior data obtained, the user's possible areas of interest are predicted, so as to update the current interest model, obtain the updated interest model, and store the newly updated interest model. For example, if a user frequently browses fitness videos within a month, the short-term behavior data can be used to determine that the user is interested in the fitness field, and there is no need to process the long-term behavior data, thus reducing the amount of calculation; for another example, if a user browses fitness videos three times a month, the user's interest in fitness videos cannot be determined only through short-term behavior data. By combining long-term behavior data, the user's browsing patterns can be accurately found, so that the fitness field can be added to the interest model.

[0053] The beneficial effect of this setting is that by obtaining long-term and short-term behavioral data, the user's areas of interest can be predicted efficiently and accurately, the accuracy of the interest model can be improved, and then the accuracy of content recommendations can be improved, thereby enhancing the user experience.

[0054] In this embodiment, the interest model includes multiple keywords and weights corresponding to each keyword, the keywords represent the field, and the weights represent the user's interest level; based on the historical behavior data, the interest model is updated to obtain an updated interest model, including: performing keyword extraction processing on the historical behavior data to obtain keywords corresponding to the historical behavior data, which are historical words; based on the historical words, the interest model is updated to obtain an updated interest model.

[0055] Specifically, a user may be interested in multiple fields, that is, multiple fields may be represented in the interest model. Different fields may correspond to different keywords, and the interest model may include multiple keywords, which are used to represent fields. For example, the keyword in the fitness field is "fitness". Each keyword in the interest model may also correspond to its own weight, which represents the degree of interest of the user in the field. The larger the weight, the greater the degree of interest.

[0056] After obtaining the historical behavior data, whether it is long-term behavior data or short-term behavior data, keywords can be extracted. For example, the user's web page reading record can be obtained, keywords can be extracted from the content of the web page read, and the reading time can be captured to infer the user's interest preferences. The user's click interaction record can also be obtained, and the user's interest in specific content can be counted by combining the user's likes, sharing, and page jump behaviors, and keywords in the specific content can be extracted. The user's search history can also be obtained, and the keywords searched by the user at a specific time or place can be analyzed as extracted keywords. The keywords extracted from the historical behavior data are determined as historical words.

[0057] Large models can be used to continuously track users' historical behavior data, such as web browsing, video clicks, keyword searches, favorites, and likes. By deeply analyzing these data and extracting historical words through text and video embedding vectors, the data can be used to continuously track users' historical behavior data, such as web browsing, video clicks, keyword searches, favorites, and likes.

[0058] Specifically, a deep learning model in NLP (Natural Language Processing) can be pre-built to extract keywords. For example, a BERT (Bidirectional Encoder Representations from Transformers) model can be used. Historical behavior data is input into the BERT model. For example, the articles visited or the content clicked by the user can be input into the model to extract keywords from these contents. In this embodiment, a hybrid architecture can be used to update the interest model. Specifically, an architecture combining deep learning and time series tracking can be used. For example, keywords can be extracted from text or multimodal data through a large language model based on BERT; short-term behavioral deviations can be dynamically monitored using LSTM (Long Short-Term Memory); TF-IDF (Term Frequency-Inverse Document Frequency) and collaborative filtering algorithms are combined with long-term user behavior data to achieve full-cycle coverage of user portraits. In this embodiment, the model architecture of the large model is not specifically limited.

[0059] When the large model is found to have changed the areas of interest of users after learning in different periods, the weight of the corresponding keywords can be gradually lowered, and newly generated keywords can be added to achieve timely updates of the model. That is, after obtaining historical words, the interest model can be updated according to the existing keywords in the interest model. For example, historical words can be added to the interest model, or the weight of the existing keywords in the interest model can be updated.

[0060] The beneficial effect of this setting is that the LSTM dynamically tracks the short-term interest drift of users, thereby flexibly adjusting the weights of keywords in the interest model; using TF-IDF and collaborative filtering algorithms, combined with historical browsing, search behavior and collection behavior, the interest model is updated to tap into the user's potential long-term interests. This achieves comprehensive confirmation of long-term and short-term interests through large models and natural language processing technology, improves the accuracy of the interest model, and thus improves the accuracy of content recommendations.

[0061] In this embodiment, the interest model is updated according to the historical words to obtain an updated interest model, including: if the historical words exist in the interest model, the weights corresponding to the historical words in the interest model are adjusted to obtain an updated interest model.

[0062] Specifically, it is determined whether the historical word exists in the interest model. If so, it is determined that content in the field related to the historical word can be recommended to the user based on the interest model. In order to enable the user to obtain content in the field more quickly, the weight of the historical word in the interest model can be adjusted, for example, the weight of the historical word can be increased. When recommending content, content in the field with a large weight is recommended first.

[0063] The beneficial effect of such a setting is that the weights in the interest model can be adjusted without rebuilding the interest model, thereby improving the adjustment efficiency and the accuracy of content recommendation.

[0064] In this embodiment, the interest model is updated according to the historical words to obtain an updated interest model, including: if the historical words do not exist in the interest model, the historical words are added to the interest model, and a preset initial weight is assigned to the historical words to obtain an updated interest model.

[0065] Specifically, if it is determined that the historical word extracted is not in the interest model, it is considered that the interest model cannot recommend content in the field of the historical word to the user. The historical word can be added to the interest model as a new keyword.

[0066] Each keyword in the interest model has its own weight. Therefore, it is necessary to assign weights to the historical words added to the interest model. A preset weight can be assigned to the historical words as the initial weight. The initial weight can be adjusted during the update of the interest model. The historical words can also be assigned weights based on their frequency of occurrence in the historical behavior data. For example, the higher the frequency of occurrence, the greater the weight, so that the content in the field of the historical words is recommended to the user first.

[0067] The beneficial effect of such a setting is that by adding keywords to the interest model, the interest model can accurately represent the areas of interest of the user and improve the accuracy of content recommendation.

[0068] S103: Determine the content to be recommended based on the updated interest model and the device type of the terminal device currently used by the user, and push the content to be recommended to the user.

[0069] Exemplarily, the device type of the terminal device currently used by the user is determined, for example, the device type may be a mobile terminal, a PC (Personal Computer), etc. Devices of different device types may be adapted to different content, for example, for mobile devices, it is more suitable to display short video content; for PC devices, it is more suitable to display long article content.

[0070] According to the updated interest model, the user's current area of ​​interest is determined, and content in this area is searched through online or offline means. According to the device type of the terminal device currently used by the user, content adapted to the device can be found from these contents as the content to be recommended. The content to be recommended is pushed to the user's terminal device for the user to view. For example, if the interest model indicates that the user's current content of interest is entertainment news, the current entertainment news or hot searches can be searched. If the terminal device currently used by the user is a mobile device, short video content can be found from the current entertainment news and hot searches and pushed to the user.

[0071] In this embodiment, it also includes: obtaining the user's operational behavior towards the recommended content, and determining the keywords corresponding to the content to be recommended; adjusting the weights of the keywords corresponding to the content to be recommended in the updated interest model according to the user's operational behavior towards the recommended content, to obtain an adjusted model; wherein the adjusted model represents the updated interest model after the weights are adjusted; determining new content to be recommended according to the adjusted model and the device type of the terminal device currently used by the user, and pushing the new content to be recommended to the user.

[0072] Specifically, after the recommended content is pushed to the user, the user's operation behavior on the recommended content can be monitored to determine whether the recommended content is what the user is interested in. If so, the content recommendation is considered accurate and there is no need to change the interest model; if not, the content recommendation is considered inaccurate and the interest model needs to be updated to recommend the content to the user again.

[0073] Through the real-time interaction between the user and the content to be recommended, the user's operation behavior on the recommended content can be captured. For example, the operation behavior may include click behavior, dwell time, like, share, exit, etc. Dwell time refers to the dwell time of the user after clicking on the content, so as to judge whether the user is satisfied with the content.

[0074] Determine the keywords of the content to be recommended, and determine the user's interest in the field of the keyword based on the user's operation behavior, so as to adjust the weight of the keyword in the interest model. For example, the longer the stay time, the higher the interest level, and the weight can be appropriately increased; the shorter the stay time, the lower the interest level, and the weight can be appropriately reduced. The interest model after weight adjustment is determined as the adjusted model. Based on the adjusted model and the device type of the terminal device, determine the new content to be recommended, continue to recommend it to the user, and continue to provide feedback on the user's operation behavior.

[0075] For example, for feedback on click behavior, it can be determined that the user has clicked on certain content with high frequency, and it is determined that the user has a high degree of attention to this content, and the keyword weight of the field of the content can be dynamically increased; for feedback on long-term browsing behavior, if the user stays on a certain content for a long time, it can be inferred that the user maintains a strong attention in this type of content and expects more relevant recommendations, and the weight of keywords in related fields can be increased; for feedback on skipping or sliding over behavior, it can be considered that the user is not interested in such content, and the corresponding weight can be reduced through a decreasing algorithm to gradually exclude these uninteresting fields. In this embodiment, time series deep learning models, such as RNN (Recurrent Neural Network), LSTM, etc., can be used to adjust weights in real time to adapt to the rapid changes in short-term needs. For example, a large language model can be used to dynamically capture and parse the feedback generated by users over time, place, and device, and real-time learning iteration can be performed by acquiring user feedback data, and content recommendation strategies can be gradually optimized in a closed-loop form to improve the accuracy of recommendations.

[0076] The beneficial effect of this setting is that the interest model is continuously optimized and adjusted according to the user's behavioral feedback to generate an increasingly accurate user personalized model. Each change in interest preference will trigger the interest model to be continuously fine-tuned and learned, so that the content pushed subsequently is more and more consistent with the user's current interests, improving the accuracy of content recommendations and enhancing user experience.

[0077] Figure 2 Schematic diagram of the optimization process of the interest model. Figure 2In the example, for the content pushed to the user, the user's operation behavior is obtained. For click behavior, the weight of the keyword corresponding to the content can be appropriately increased, so as to continuously recommend relevant content to the user; for long-term stay behavior, the weight of the keyword corresponding to the content can be appropriately increased, so as to continuously recommend relevant content to the user; for closing or sliding behavior, the weight of the keyword corresponding to the content can be appropriately weakened, so as to reduce the recommendation of relevant content to the user. According to the adjusted interest model, the recommendation list is updated. The recommendation list can include multiple contents that the user may be interested in, and new content is selected from the recommendation list to be recommended to the user. For example, the content ranked first in the recommendation list can be recommended to the user.

[0078] In the disclosed embodiment, the spatiotemporal scene in which the user is currently located can be determined, thereby determining the user's current scene information. The user's current interest model is obtained, and the interest model can characterize the areas of content that the user is interested in. According to the current scene information, the interest model is updated in real time so that the updated interest model can reflect the user's current areas of interest. According to the updated interest model and the device type of the terminal device currently used by the user, the content to be recommended is determined and pushed to the user. This allows the user to view content that is of interest and adapted to the terminal device, improves the accuracy and flexibility of content recommendations, and enhances the user experience.

[0079] Figure 3 A flowchart of a method for content recommendation based on user interests provided in an embodiment of the present disclosure.

[0080] In this embodiment, the interest model is updated according to the current scene information to obtain an updated interest model, including: predicting the user's preference areas according to the current scene information; wherein the preference areas represent the predicted areas of interest to the user; and updating the interest model according to the preference areas to obtain an updated interest model.

[0081] This embodiment is based on the above embodiment. Figure 3 As shown, the method comprises the following steps:

[0082] S301. Obtain current scene information and a pre-stored interest model; wherein the current scene information represents the space-time scene in which the user is currently located, and the interest model is a user portrait model that represents the field of content that the user is interested in.

[0083] Exemplarily, this step may refer to the above-mentioned step S101 and will not be described in detail.

[0084] S302. Predict the user's preferred areas based on current scene information; wherein the preferred areas represent the predicted areas of interest to the user.

[0085] Exemplarily, based on the current scene information, the user's current area of ​​interest is predicted as the preferred area. For example, if it is determined that the user is currently in a shopping mall, it can be predicted that the user's areas of interest are preferential activities, shopping, etc., and preferential activities, shopping, etc. are used as the current preferred areas.

[0086] In this embodiment, the current scene information includes time information and space information; predicting the user's preference area based on the current scene information includes: determining the scene category in which the user is currently located based on the time information and space information; wherein the scene category represents the user's schedule in the current scene; predicting the user's preference area based on the scene category in which the user is currently located.

[0087] Specifically, the current scene information may include current time information and space information. The time information may represent the current date and time, etc., and the space information may represent the current location and place, etc. According to the current time information and space information, the scene category in which the user is currently located is determined. The scene category may represent the user's schedule in the current scene. That is, in different categories of scenes, the user may have different schedules. For example, in a travel category scene, the user's schedule may be to visit scenic spots; in an office category scene, the user's schedule may be to check information.

[0088] Different time information and space information may correspond to different scene categories. A plurality of scene categories are preset. For example, if the time information is eight o'clock in the morning on a weekday and the space information is the gate of a residential area, the scene category may be a commuting scene.

[0089] In different categories of scenes, users may have different preference areas. According to the scene category that the user is currently in, the area that the user may be interested in can be determined as the preference area. For example, in the scene of going to work, the user may be interested in the road congestion situation, and the predicted preference area may be the road traffic area.

[0090] The beneficial effect of this setting is that, based on the current scene information, it is possible to determine what category of scene the user is currently in, thereby updating the user's interest preferences in real time based on the user's actual scene, flexibly responding to scene changes, and improving the accuracy of content recommendations.

[0091] In this embodiment, the scene category in which the user is currently located is determined based on time information and space information, including: determining the time category in which the user is currently located based on time information; wherein the time category represents the state of the user at the current time; determining the space category in which the user is currently located based on space information; wherein the space category represents the state of the user in the current place; determining the scene category in which the user is currently located based on the time category and the space category.

[0092] Specifically, multiple time categories are preset, and users can have different statuses in different time categories. For example, time categories can be divided into categories such as weekdays, weekends, mornings, and evenings. According to the current time information, the time category in which the user is currently located is determined. For example, the current time can be detected at any time through a preset time synchronization component, and it can be determined which time category the current time belongs to. That is, the association relationship between different dates and time categories, and / or the association relationship between different time points and time categories can be preset, and the corresponding time category can be determined according to the current time information.

[0093] There are multiple spatial categories preset, and users can have different states under different spatial categories. For example, spatial categories can be divided into office areas, travel areas, homes, and other categories. According to the spatial information, the spatial category in which the user is currently located is determined. For example, a preset geographic data matching tool, such as a high-precision map, can be used to map the current location coordinates to different spatial categories. That is, the association relationship between different location coordinates and spatial categories can be preset, and the corresponding spatial category can be determined according to the current spatial information.

[0094] Combine the time category and the space category to determine the scene category the user is currently in. You can preset different time categories and space categories, and the association relationship between the corresponding scene categories. According to the preset association relationship, determine the scene category corresponding to the time category and space category. For example, if the time category is weekdays and the space category is office area, then the scene category can be a scene of going to work. For another example, if the time category is weekends and the space category is outdoor, then the scene category can be a scene of an outing.

[0095] The beneficial effect of this setting is that it combines time information and space information to determine scene information, improves the accuracy of scene category determination, and adapts to changes in recommendation needs in different time periods and location environments. It also improves the update accuracy of the interest model and ensures accurate content recommendations for users.

[0096] In this embodiment, the user's preferred domain is predicted based on the scene category in which the user is currently located, including: according to a preset first association relationship, determining that the domain corresponding to the scene category in which the user is currently located is the preferred domain; wherein the preset first association relationship represents the association relationship between the scene category and the domain.

[0097] Specifically, a first association relationship is preset and stored, and the first association relationship can represent the association relationship between the scene category and the field. After determining the scene category that the user is currently in, according to the preset first association relationship, the field corresponding to the current scene category can be found as the user's current preferred field. For example, if the current scene category is a work scene, the preferred field can be a technology field; if the current scene category is an outing scene, the preferred field can be a travel-related field.

[0098] The beneficial effect of such a setting is that, according to the preset association relationship, the preferred field can be quickly determined, the efficiency of determining the preferred field is improved, and then the efficiency of content recommendation is improved.

[0099] S303: Update the interest model according to the preferred field to obtain an updated interest model.

[0100] Exemplarily, determine whether the preference field is the field represented in the interest model. If so, it is considered that the interest model can recommend content that the user is interested in; if not, it is considered that the interest model cannot recommend content that the user is interested in, and the preference field can be added to the interest model to recommend relevant content in the preference field to the user.

[0101] In this embodiment, based on the time and space scene the user is currently in, the user's current interest in the field is predicted, and the interest model is updated. Taking into account that the user's interest preferences change with the change of time and space scenes, the interest model is flexibly adjusted so that the recommended content follows the changes of the user's interests and improves the user experience.

[0102] In this embodiment, the interest model includes multiple keywords and weights corresponding to each keyword, the keywords represent the field, and the weights represent the user's interest level; according to the preference field, the interest model is updated to obtain an updated interest model, including: according to a preset second association relationship, determining the keyword corresponding to the preference field as the target word; wherein the preset second association relationship represents the association relationship between the field and the keyword; according to the target word, the interest model is updated to obtain an updated interest model.

[0103] Specifically, the interest model may include multiple keywords, and different keywords represent different fields. Each keyword has its own weight, which is used to indicate the user's interest in the keyword's field. The larger the weight, the higher the interest; the smaller the weight, the lower the interest.

[0104] A second association relationship is preset, and the second association relationship can represent the association relationship between different fields and keywords. After determining the preferred field, the keyword corresponding to the preferred field can be determined as the target word according to the second association relationship. The current interest is updated according to the target word, for example, the target word can be added to the interest model, or the weight of the keyword in the interest model can be adjusted according to the target word.

[0105] The beneficial effect of such a setting is that the keywords to be updated can be quickly determined as target words, thereby improving the updating efficiency of the interest model and achieving rapid content recommendation for users.

[0106] In this embodiment, the interest model is updated according to the target word to obtain an updated interest model, including: if the target word exists in the interest model, the weight corresponding to the target word in the interest model is adjusted to obtain an updated interest model.

[0107] Specifically, it is determined whether the target word exists in the interest model. If so, it is determined that content in the field related to the target word can be recommended to the user based on the interest model. In order to enable the user to obtain content in the field more quickly, the weight of the target word in the interest model can be adjusted. For example, the weight of the target word can be increased according to a preset incremental algorithm. When recommending content, content in fields with large weights is recommended first.

[0108] The beneficial effect of such a setting is that the weights in the interest model can be adjusted without rebuilding the interest model, thereby improving the adjustment efficiency and the accuracy of content recommendation.

[0109] In this embodiment, the interest model is updated according to the target word to obtain an updated interest model, including: if the target word does not exist in the interest model, the target word is added to the interest model, and a preset initial weight is assigned to the target word to obtain an updated interest model.

[0110] Specifically, if it is determined that the target word is not determined in the interest model, it is considered that the interest model cannot recommend content in the field of the target word to the user. The target word can be added to the interest model as a new keyword.

[0111] Each keyword in the interest model has its own weight. Therefore, it is necessary to assign weights to the target words added to the interest model. A preset weight can be assigned to the target words as the initial weight. The initial weight can be adjusted during the updating process of the interest model. In this embodiment, the target word represents the field that the user is currently interested in, that is, the user is most likely to be recommended content in this field. Therefore, a larger initial weight can be preset to give priority to recommending content in the field of the target word to the user.

[0112] The beneficial effect of such a setting is that by adding keywords to the interest model, the interest model can accurately represent the areas of interest of the user and improve the accuracy of content recommendation.

[0113] Figure 4 Schematic diagram of the updating process of the interest model. Figure 4 In the process, the user's current spatial information and the user's current time information are obtained through GPS (Global Positioning System). The spatial information and time information are determined as the current scene information. According to the current scene information, the user's current spatial category and time category are matched. Combined with the spatial category and time category, the keywords and weights in the interest model are adjusted, and the updated interest model is output.

[0114] S304: Determine the content to be recommended based on the updated interest model and the device type of the terminal device currently used by the user, and push the content to be recommended to the user.

[0115] Exemplarily, this step may refer to the above-mentioned step S103 and will not be described in detail.

[0116] In the disclosed embodiment, the spatiotemporal scene in which the user is currently located can be determined, thereby determining the user's current scene information. The user's current interest model is obtained, and the interest model can characterize the areas of content that the user is interested in. According to the current scene information, the interest model is updated in real time so that the updated interest model can reflect the user's current areas of interest. According to the updated interest model and the device type of the terminal device currently used by the user, the content to be recommended is determined and pushed to the user. This allows the user to view content that is of interest and adapted to the terminal device, improves the accuracy and flexibility of content recommendations, and enhances the user experience.

[0117] Figure 5 A flowchart of a method for content recommendation based on user interests provided in an embodiment of the present disclosure.

[0118] In this embodiment, the content to be recommended is determined based on the updated interest model and the device type of the terminal device currently used by the user, including: determining candidate content from a preset content database based on the updated interest model; wherein the candidate content is the content that the user is interested in; determining the device type of the terminal device currently used by the user, and determining the content to be recommended from the candidate content based on the device type of the terminal device.

[0119] This embodiment is based on the above embodiment. Figure 5 As shown, the method comprises the following steps:

[0120] S501. Obtain current scene information and a pre-stored interest model; wherein the current scene information represents the space-time scene in which the user is currently located, and the interest model is a user portrait model that represents the field of content that the user is interested in.

[0121] Exemplarily, this step may refer to the above-mentioned step S101 and will not be described in detail.

[0122] S502: Update the interest model according to the current scene information to obtain an updated interest model.

[0123] Exemplarily, this step may refer to the above-mentioned step S102 and will not be described in detail.

[0124] S503: Determine candidate content from a preset content database according to the updated interest model; wherein the candidate content is content that the user is interested in.

[0125] Exemplarily, the updated interest model can characterize the field to which the content that the user is currently interested in belongs, and the content corresponding to the field can be searched online or offline based on the updated interest model. In this embodiment, a content database can be preset, and the content database can store content in various fields. According to the updated interest model, the keywords contained in the model and the weight of each keyword are determined. The keywords can be sorted according to the weights of the keywords. According to the sorting position of the keywords, the content in the field corresponding to the keywords is searched from the preset content database. For example, the content in the field corresponding to the first two keywords can be searched.

[0126] The number of contents can be preset to avoid finding too much content. The contents in the content database can be screened according to the number of views, likes, and reposts of each content, so as to find out the content that may be recommended to the user. According to the preset number of contents, the found contents are stored in a preset content list. For example, if the preset number of contents is 10, these 10 contents can be stored in the content list according to the number of views. The contents found from the content database are determined as candidate contents, and the candidate contents are the contents that the user may be interested in. In this embodiment, the method for searching the candidate contents is not specifically limited. For example, according to the release time of the content, real-time popular contents can be given priority as candidate contents. For another example, the contents that the user has been reading for a long time or related to work can be determined as candidate contents, so as to make dynamic trade-offs and adjustments in the short-term interest fluctuations and long-term interest precipitation of the user.

[0127] S504: Determine the device type of the terminal device currently used by the user, and determine the content to be recommended from the candidate content according to the device type of the terminal device.

[0128] For example, a user uses a terminal device to view content, and the device type of the terminal device currently used by the user is determined, for example, the device type can be a mobile terminal or a PC terminal. When users use different types of devices, there are differences in usage habits and behavior patterns. Therefore, different content needs to be recommended to users for different types of devices.

[0129] The content viewed by users using different types of devices during the historical time period can be recorded. For example, when using the PC, users tend to read professional reports and detailed articles; when using the mobile terminal, users tend to watch short videos or read-once content. Based on the content viewed by the user using the current type of device during the historical period, the candidate content is searched for content with high similarity as the current content to be recommended.

[0130] In this embodiment, the device type that the user is using is detected, for example, the device resolution, device ID, etc. are directly obtained through an API (Application Program Interface) to determine the device type. Different content to be recommended is determined for different device types, making the recommendation more contextualized and in line with the device usage requirements, greatly improving the consistency of the recommendation experience across devices.

[0131] In this embodiment, content to be recommended is determined from candidate content according to the device type of the terminal device, including: determining a content format corresponding to the device type of the terminal device as a target format according to a preset third association relationship; wherein the preset third association relationship represents an association relationship between the device type and the content format; and determining candidate content of the target format as content to be recommended.

[0132] Specifically, different types of devices are suitable for displaying different formats of content, and a third association relationship is pre-set and stored, which can represent the association relationship between the device type and the content format. For example, for a PC device, the content suitable for display is the content in the long article format; for a mobile device, the content suitable for display is the content in the short video format.

[0133] After determining the device type of the terminal device currently used by the user, determine the content format corresponding to the device type of the terminal device according to the preset third association relationship as the target format. Obtain the content in the target format from the candidate content as the current content to be recommended. For example, you can search from the content list of the candidate content from top to bottom to see whether there is content in the target format, and use the first content in the target format as the content to be recommended. If there is no content in the target format in the candidate content, the content ranked first in the candidate content can be used as the content to be recommended.

[0134] Figure 6Schematic diagram of the content recommendation process based on device type. Figure 6 In the process, the attribute information of the device is first obtained, for example, the attribute information may include resolution, device ID, user operation behavior in a historical time period, etc. The device type of the device is determined based on the attribute information. For different device types, the content to be recommended is determined specifically and output on the device.

[0135] The beneficial effect of such a setting is that, through automatic detection of the device used by the user, the type of content to be recommended can be adjusted accordingly, content switching and optimization across devices can be achieved, and matching personalized services can be provided on different devices, thereby improving the user experience.

[0136] In the disclosed embodiment, the spatiotemporal scene in which the user is currently located can be determined, thereby determining the user's current scene information. The user's current interest model is obtained, and the interest model can characterize the areas of content that the user is interested in. According to the current scene information, the interest model is updated in real time so that the updated interest model can reflect the user's current areas of interest. According to the updated interest model and the device type of the terminal device currently used by the user, the content to be recommended is determined and pushed to the user. This allows the user to view content that is of interest and adapted to the terminal device, improves the accuracy and flexibility of content recommendations, and enhances the user experience.

[0137] Figure 7 A flowchart of a method for content recommendation based on user interests provided in an embodiment of the present disclosure.

[0138] This embodiment is based on the above embodiment. Figure 7 As shown, the method comprises the following steps:

[0139] S701. Obtain current geographic location, current time, device type, and historical behavior data.

[0140] Exemplarily, the current geographic location may be the location coordinates obtained by GPS, the current time may be the current date and time point, the device type refers to the device type of the terminal device currently used by the user, and the historical behavior data may include the user's long-term and short-term behavior data. The long-term and short-term behavior data may include the user's historical behavior data in a first time period and the user's historical behavior data in a second time period. The first time period may be greater than the second time period, and the second time period may be included in the first time period. For example, the first time period is the past year, and the second time period is the past month.

[0141] S702: Generate current scene information according to the current geographical location and current time.

[0142] Exemplarily, the spatial information can be determined according to the current geographical location, and the time information can be determined according to the current time. The current scene information can be determined according to the current spatial information and time information. That is, the current scene information can include spatial information and time information.

[0143] S703: Update the keywords and weights of the interest model according to the current scene information.

[0144] Exemplarily, the scene category that the user is currently in is determined based on the current scene information. Different scene categories may correspond to different preference areas, indicating that the user has different areas of interest in different scenes. The current interest model is updated based on the current preference area.

[0145] The interest model includes multiple keywords, each with its own weight. You can update the keywords in the interest model, and you can also update the weights of the keywords. This improves the matching accuracy of recommendations for users at different times and locations, and can recommend appropriate content based on the environment.

[0146] S704: Update the keywords and weights of the interest model based on the historical behavior data.

[0147] For example, the keywords and / or weights in the interest model can be updated based on the user's long-term and short-term historical behavior data. This enables the detection of short-term popular content and mixes it with long-term content to ensure that the recommended content meets current interests while not missing long-term needs.

[0148] S705: Determine the updated interest model.

[0149] Exemplarily, an updated interest model is obtained based on current scene information and historical behavior data.

[0150] S706: Determine content to be recommended based on the updated interest model and device type.

[0151] Exemplarily, a large amount of content may be screened based on the updated interest model. For example, each content may be scored and content with a high score may be used as content to be recommended.

[0152] By detecting the type of device the user is using, targeted content recommendation strategies can be triggered accordingly. For example, the device resolution, device ID, and user behavior can be directly obtained through the system API. User behavior can be the number of times a user views short videos or the time they stay on long texts on different devices. According to the device type, the form of recommended content can be adjusted and optimized.

[0153] S707: Display the content to be recommended on the terminal device.

[0154] Exemplarily, the display form of the content to be recommended is adapted to the terminal device, so that the content to be recommended is displayed on the terminal device.

[0155] S708: Obtain the user's operation behavior on the recommended content.

[0156] Exemplarily, once all the content to be recommended is selected by the user, the user's operation behavior on the recommended content can be obtained.

[0157] S709: Update the interest model according to the user's operational behavior towards the recommended content.

[0158] For example, the interest model can be further optimized based on the user's operation behavior. The user's operation behavior is captured in real time as feedback on the user's attitude towards the recommended content. The user's operation behavior may include clicking, sliding, skipping, exiting, liking, commenting, sharing, etc. After the content to be recommended is pushed to the front-end interface, the user's operation behavior is captured and recorded in real time. These behaviors reflect the user's attention to the recommended content and affect the adjustment of the interest model.

[0159] The recommended content can be scored based on these user operations. For example, if the user forwards and likes the recommended content, the score of the recommended content can be higher. The higher the score, the more interested the user is in the field to which the recommended content belongs. The weight of the keywords in the field in the interest model can be increased to adjust the interest model. Based on this closed-loop feedback optimization method, the interest model can be continuously improved to maintain the interactive characteristics of "learning and adaptation", avoid outdated or biased recommendations, and improve user experience.

[0160] For example, a user browsed some technical articles on a PC on Monday morning, and then went camping the following weekend and used a mobile phone to search for camping-related content. The system can intelligently capture the user's behavioral changes and push personalized content to the user.

[0161] Specifically, the GPS positioning module detects that the user has come to a scenic spot outside the city on the weekend, and automatically enters the "weekend scene, outdoor scene". The user's device changes are captured, and the device switches from the PC device on weekdays to the mobile device. According to the specific device characteristics, the lengthy recommendations can be automatically reduced, and some short and practical videos or articles can be mainly promoted. Combined with similar travel or equipment articles that the user has viewed in the past few weeks, relevant content can be recommended to the user. The user clicks on the recommendations of multiple camping equipment guides and stays on these pages for a long time. He also searches and views related camping equipment content and adds some outdoor products to the favorites. Based on these user feedback behaviors, the LSTM algorithm and user-specific behavior analysis are used to re-update the interest model of the weekend day, adjust the priority of the next batch of camping-related content in real time, and finally ensure that the recommended content matches the real-time scene and personal needs. The next time the user goes to another camping location, different types of equipment reviews, related experience sharing videos and other content can be pushed first according to the time point and historical records.

[0162] Figure 8 The structure block diagram of a content recommendation device based on user interests provided by an embodiment of the present disclosure. For the convenience of explanation, only the parts related to the embodiment of the present disclosure are shown. Figure 8 The content recommendation device 800 based on user interests includes: an acquisition unit 801, an updating unit 802 and a recommendation unit 803.

[0163] The acquisition unit 801 is used to acquire current scene information and a pre-stored interest model; wherein the current scene information represents the spatiotemporal scene in which the user is currently located, and the interest model is a user portrait model representing the field of the user's interested content;

[0164] An updating unit 802, configured to update the interest model according to the current scene information to obtain an updated interest model;

[0165] The recommendation unit 803 is used to determine the content to be recommended according to the updated interest model and the device type of the terminal device currently used by the user, and push the content to be recommended to the user.

[0166] Fig. 9 A structural block diagram of a content recommendation device based on user interests provided by an embodiment of the present disclosure, such as Fig. 9 As shown, the content recommendation device 900 based on user interests includes an acquisition unit 901 , an updating unit 902 and a recommendation unit 903 , wherein the updating unit 902 includes a domain prediction module 9021 and a model updating module 9022 .

[0167] A domain prediction module 9021 is used to predict the user's preferred domain based on the current scene information; wherein the preferred domain represents the predicted domain of interest to the user;

[0168] The model updating module 9022 is used to update the interest model according to the preference domain to obtain an updated interest model.

[0169] In one example, the current scene information includes time information and space information; the domain prediction module 9021 includes:

[0170] A category determination submodule, used to determine the category of the scene in which the user is currently located according to the time information and the space information; wherein the scene category represents the schedule of the user in the current scene;

[0171] The domain prediction submodule is used to predict the user's preferred domain according to the scene category in which the user is currently located.

[0172] In one example, the category determination submodule is specifically used to:

[0173] Determine the current time category of the user according to the time information; wherein the time category represents the state of the user at the current time;

[0174] Determine the space category where the user is currently located according to the space information; wherein the space category represents the state of the user in the current location;

[0175] The scene category in which the user is currently located is determined according to the time category and the space category.

[0176] In one example, the domain prediction submodule is specifically used to:

[0177] According to a preset first association relationship, a domain corresponding to the scene category in which the user is currently located is determined as the preferred domain; wherein the preset first association relationship represents an association relationship between the scene category and the domain.

[0178] In one example, the interest model includes multiple keywords and weights corresponding to the keywords, the keywords represent the fields, and the weights represent the user's interest level; the model updating module 9022 includes:

[0179] A target determination submodule is used to determine the keyword corresponding to the preferred field as the target word according to a preset second association relationship; wherein the preset second association relationship represents the association relationship between the field and the keyword;

[0180] The model updating submodule is used to update the interest model according to the target word to obtain an updated interest model.

[0181] In one example, the model update submodule is specifically used to:

[0182] If the target word exists in the interest model, the weight corresponding to the target word in the interest model is adjusted to obtain the updated interest model.

[0183] In one example, the model update submodule is specifically used to:

[0184] If the target word does not exist in the interest model, the target word is added to the interest model, and a preset initial weight is assigned to the target word to obtain the updated interest model.

[0185] In one example, the recommendation unit 903 includes:

[0186] A candidate determination module, configured to determine candidate content from a preset content database according to the updated interest model; wherein the candidate content is content that the user is interested in;

[0187] The content recommendation module is used to determine the device type of the terminal device currently used by the user, and determine the content to be recommended from the candidate content according to the device type of the terminal device.

[0188] In one example, the content recommendation module includes:

[0189] The format determination submodule is used to determine the content format corresponding to the device type of the terminal device as the target format according to a preset third association relationship; wherein the preset third association relationship represents the association relationship between the device type and the content format;

[0190] The content determination submodule is used to determine the candidate content in the target format as the content to be recommended.

[0191] In one example, it also includes:

[0192] A data acquisition unit, configured to acquire historical behavior data of the user; wherein the historical behavior data includes behavior data of the user in a preset first time period and behavior data of the user in a preset second time period, the preset first time period is greater than the preset second time period, and the behavior data represents the operation behavior and operation content of the user;

[0193] The interest model updating unit is used to update the interest model according to the historical behavior data to obtain an updated interest model.

[0194] In one example, the interest model includes a plurality of keywords and weights corresponding to the keywords, the keywords represent the fields, and the weights represent the user's interest level; the interest model updating unit includes:

[0195] A historical word extraction module, used to perform keyword extraction processing on the historical behavior data to obtain keywords corresponding to the historical behavior data, which are historical words;

[0196] The historical word updating module is used to update the interest model according to the historical words to obtain an updated interest model.

[0197] In one example, the historical word update module includes:

[0198] The weight adjustment submodule is used to adjust the weight corresponding to the historical word in the interest model if the historical word exists in the interest model, so as to obtain the updated interest model.

[0199] In one example, the historical word update module includes:

[0200] The historical word adding submodule is used to add the historical word to the interest model if the historical word does not exist in the interest model, and assign a preset initial weight to the historical word to obtain the updated interest model.

[0201] In one example, it also includes:

[0202] An operation acquisition unit, used to acquire the user's operation behavior on the content to be recommended, and determine the keywords corresponding to the content to be recommended;

[0203] A model adjustment unit, configured to adjust the weights of the keywords corresponding to the content to be recommended in the updated interest model according to the user's operation behavior on the content to be recommended, so as to obtain an adjusted model; wherein the adjusted model represents the updated interest model after the weights are adjusted;

[0204] The re-recommendation unit is used to determine new content to be recommended according to the adjusted model and the device type of the terminal device currently used by the user, and push the new content to be recommended to the user.

[0205] According to an embodiment of the present disclosure, the present disclosure also provides an electronic device.

[0206] Fig.10 A structural block diagram of an electronic device provided in an embodiment of the present disclosure, such as Fig.10As shown, the electronic device 1000 includes: at least one processor 1002; and a memory 1001 communicatively connected to the at least one processor 1002; wherein the memory stores instructions executable by the at least one processor 1002, and the instructions are executed by the at least one processor 1002 so that the at least one processor 1002 can execute the user interest-based content recommendation method disclosed in the present invention.

[0207] The electronic device 1000 further includes a receiver 1003 and a transmitter 1004. The receiver 1003 is used to receive instructions and data sent by other devices, and the transmitter 1004 is used to send instructions and data to external devices.

[0208] According to an embodiment of the present disclosure, the present disclosure also provides an electronic device, a readable storage medium and a computer program product.

[0209] According to an embodiment of the present disclosure, the present disclosure also provides a computer program product, which includes: a computer program, the computer program is stored in a readable storage medium, at least one processor of an electronic device can read the computer program from the readable storage medium, and at least one processor executes the computer program so that the electronic device executes the solution provided by any of the above embodiments.

[0210] Fig.11 A schematic block diagram of an example electronic device 1100 that can be used to implement an embodiment of the present disclosure is shown. The electronic device is intended to represent various forms of digital computers, such as laptop computers, desktop computers, workstations, personal digital assistants, servers, blade servers, mainframe computers, and other suitable computers. The electronic device can also represent various forms of mobile devices, such as personal digital processing, cellular phones, smart phones, wearable devices, and other similar computing devices. The components shown herein, their connections and relationships, and their functions are merely examples and are not intended to limit the implementation of the present disclosure described and / or required herein.

[0211] like Fig.11 As shown, the device 1100 includes a computing unit 1101, which can perform various appropriate actions and processes according to a computer program stored in a read-only memory (ROM) 1102 or a computer program loaded from a storage unit 1108 into a random access memory (RAM) 1103. In the RAM 1103, various programs and data required for the operation of the device 1100 can also be stored. The computing unit 1101, the ROM 1102, and the RAM 1103 are connected to each other via a bus 1104. An input / output (I / O) interface 1105 is also connected to the bus 1104.

[0212] A number of components in the device 1100 are connected to the I / O interface 1105, including: an input unit 1106, such as a keyboard, a mouse, etc.; an output unit 1107, such as various types of displays, speakers, etc.; a storage unit 1108, such as a disk, an optical disk, etc.; and a communication unit 1109, such as a network card, a modem, a wireless communication transceiver, etc. The communication unit 1109 allows the device 1100 to exchange information / data with other devices through a computer network such as the Internet and / or various telecommunication networks.

[0213] The computing unit 1101 may be a variety of general and / or special processing components with processing and computing capabilities. Some examples of the computing unit 1101 include, but are not limited to, a central processing unit (CPU), a graphics processing unit (GPU), various dedicated artificial intelligence (AI) computing chips, various computing units running machine learning model algorithms, digital signal processors (DSPs), and any appropriate processors, controllers, microcontrollers, etc. The computing unit 1101 performs the various methods and processes described above, such as a content recommendation method based on user interests. For example, in some embodiments, the content recommendation method based on user interests may be implemented as a computer software program, which is tangibly included in a machine-readable medium, such as a storage unit 1108. In some embodiments, part or all of the computer program may be loaded and / or installed on the device 1100 via the ROM 1102 and / or the communication unit 1109. When the computer program is loaded into the RAM 1103 and executed by the computing unit 1101, one or more steps of the content recommendation method based on user interests described above may be performed. Alternatively, in other embodiments, the computing unit 1101 may be configured to execute the content recommendation method based on user interests in any other appropriate manner (eg, by means of firmware).

[0214] Various implementations of the systems and techniques described above herein can be implemented in digital electronic circuit systems, integrated circuit systems, field programmable gate arrays (FPGAs), application specific integrated circuits (ASICs), application specific standard products (ASSPs), systems on chips (SOCs), complex programmable logic devices (CPLDs), computer hardware, firmware, software, and / or combinations thereof. These various implementations can include: being implemented in one or more computer programs that can be executed and / or interpreted on a programmable system including at least one programmable processor, which can be a special purpose or general purpose programmable processor that can receive data and instructions from a storage system, at least one input device, and at least one output device, and transmit data and instructions to the storage system, the at least one input device, and the at least one output device.

[0215] The program code for implementing the method of the present disclosure may be written in any combination of one or more programming languages. These program codes may be provided to a processor or controller of a general-purpose computer, a special-purpose computer, or other programmable data processing device, so that the program code, when executed by the processor or controller, enables the functions / operations specified in the flow chart and / or block diagram to be implemented. The program code may be executed entirely on the machine, partially on the machine, partially on the machine and partially on a remote machine as a stand-alone software package, or entirely on a remote machine or server.

[0216] In the context of the present disclosure, a machine-readable medium may be a tangible medium that may contain or store a program for use by or in conjunction with an instruction execution system, device, or equipment. A machine-readable medium may be a machine-readable signal medium or a machine-readable storage medium. A machine-readable medium may include, but is not limited to, an electronic, magnetic, optical, electromagnetic, infrared, or semiconductor system, device, or equipment, or any suitable combination of the foregoing. A more specific example of a machine-readable storage medium may include an electrical connection based on one or more lines, a portable computer disk, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disk read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the foregoing.

[0217] To provide interaction with a user, the systems and techniques described herein can be implemented on a computer having: a display device (e.g., a CRT (cathode ray tube) or LCD (liquid crystal display) monitor) for displaying information to the user; and a keyboard and pointing device (e.g., a mouse or trackball) through which the user can provide input to the computer. Other types of devices can also be used to provide interaction with the user; for example, the feedback provided to the user can be any form of sensory feedback (e.g., visual feedback, auditory feedback, or tactile feedback); and input from the user can be received in any form (including acoustic input, voice input, or tactile input).

[0218] The systems and techniques described herein may be implemented in a computing system that includes back-end components (e.g., as a data server), or a computing system that includes middleware components (e.g., an application server), or a computing system that includes front-end components (e.g., a user computer with a graphical user interface or a web browser through which a user can interact with implementations of the systems and techniques described herein), or a computing system that includes any combination of such back-end components, middleware components, or front-end components. The components of the system may be interconnected by any form or medium of digital data communication (e.g., a communication network). Examples of communication networks include: a local area network (LAN), a wide area network (WAN), and the Internet.

[0219] A computer system may include a client and a server. The client and the server are generally remote from each other and usually interact through a communication network. The relationship between the client and the server is generated by computer programs running on the corresponding computers and having a client-server relationship with each other. The server may be a cloud server, also known as a cloud computing server or cloud host, which is a host product in the cloud computing service system to solve the defects of difficult management and weak business scalability in traditional physical hosts and VPS services ("Virtual Private Server", or "VPS" for short). The server may also be a server of a distributed system, or a server combined with a blockchain.

[0220] It should be understood that the various forms of processes shown above can be used to reorder, add or delete steps. For example, the steps recorded in this disclosure can be executed in parallel, sequentially or in different orders, as long as the desired results of the technical solutions disclosed in this disclosure can be achieved, and this document does not limit this.

[0221] The above specific implementations do not constitute a limitation on the protection scope of the present disclosure. It should be understood by those skilled in the art that various modifications, combinations, sub-combinations and substitutions can be made according to design requirements and other factors. Any modification, equivalent substitution and improvement made within the spirit and principle of the present disclosure shall be included in the protection scope of the present disclosure.

Claims

1. A content recommendation method based on user interests, comprising: Acquire current scene information and a pre-stored interest model; wherein the current scene information represents the spatiotemporal scene in which the user is currently located, and the interest model is a user portrait model representing the field of content that the user is interested in; According to the current scene information, the interest model is updated to obtain an updated interest model; According to the updated interest model and the device type of the terminal device currently used by the user, the content to be recommended is determined, and the content to be recommended is pushed to the user.

2. The method according to claim 1, wherein: The updating of the interest model according to the current scene information to obtain an updated interest model includes: Predicting the user's preferred areas based on the current scene information; wherein the preferred areas represent the predicted areas of interest to the user; The interest model is updated according to the preference domain to obtain an updated interest model.

3. The method according to claim 2, wherein: The current scene information includes time information and space information; and predicting the user's preferred area based on the current scene information includes: Determine the scene category in which the user is currently located according to the time information and the space information; wherein the scene category represents the user's schedule in the current scene; Predict the user's preferred area based on the scene category in which the user is currently located.

4. The method according to claim 3, wherein: The determining, according to the time information and the space information, the scene category in which the user is currently located includes: Determine the current time category of the user according to the time information; wherein the time category represents the state of the user at the current time; Determine the space category where the user is currently located according to the space information; wherein the space category represents the state of the user in the current location; The scene category in which the user is currently located is determined according to the time category and the space category.

5. The method according to claim 3 or 4, wherein: The predicting the user's preferred area according to the scene category in which the user is currently located includes: According to a preset first association relationship, a domain corresponding to the scene category in which the user is currently located is determined as the preferred domain; wherein the preset first association relationship represents an association relationship between the scene category and the domain.

6. The method according to any one of claims 2 to 5, wherein: The interest model includes a plurality of keywords and weights corresponding to the keywords, the keywords represent fields, and the weights represent the degree of interest of the user; the updating of the interest model according to the preferred fields to obtain the updated interest model includes: According to a preset second association relationship, a keyword corresponding to the preferred field is determined as a target word; wherein the preset second association relationship represents an association relationship between the field and the keyword; The interest model is updated according to the target word to obtain an updated interest model.

7. The method according to claim 6, wherein: The updating of the interest model according to the target word to obtain an updated interest model includes: If the target word exists in the interest model, adjusting the weight corresponding to the target word in the interest model to obtain the updated interest model; If the target word does not exist in the interest model, the target word is added to the interest model, and a preset initial weight is assigned to the target word to obtain the updated interest model.

8. The method according to any one of claims 1 to 7, wherein: The determining the content to be recommended according to the updated interest model and the device type of the terminal device currently used by the user includes: Determining candidate content from a preset content database according to the updated interest model; wherein the candidate content is content that the user is interested in; The device type of the terminal device currently used by the user is determined, and the content to be recommended is determined from the candidate content according to the device type of the terminal device.

9. The method according to claim 8, wherein: The determining the content to be recommended from the candidate content according to the device type of the terminal device includes: According to a preset third association relationship, determining the content format corresponding to the device type of the terminal device as the target format; wherein the preset third association relationship represents the association relationship between the device type and the content format; The candidate content in the target format is determined as the content to be recommended.

10. The method according to any one of claims 1 to 9, further comprising: Acquire historical behavior data of the user; wherein the historical behavior data includes behavior data of the user in a preset first time period and behavior data of the user in a preset second time period, the preset first time period is greater than the preset second time period, and the behavior data represents the operation behavior and operation content of the user; The interest model is updated according to the historical behavior data to obtain an updated interest model.

11. The method according to claim 10, wherein: The interest model includes a plurality of keywords and weights corresponding to the keywords, wherein the keywords represent the fields and the weights represent the user's interest level; The updating of the interest model according to the historical behavior data to obtain an updated interest model includes: Perform keyword extraction processing on the historical behavior data to obtain keywords corresponding to the historical behavior data, which are historical words; The interest model is updated according to the historical words to obtain an updated interest model.

12. The method according to claim 11, wherein: The updating of the interest model according to the historical words to obtain an updated interest model includes: If the historical word exists in the interest model, adjusting the weight corresponding to the historical word in the interest model to obtain the updated interest model; If the historical word does not exist in the interest model, the historical word is added to the interest model, and a preset initial weight is assigned to the historical word to obtain the updated interest model.

13. The method according to any one of claims 1 to 12, further comprising: Acquire the user's operation behavior on the content to be recommended, and determine the keywords corresponding to the content to be recommended; According to the user's operation behavior on the content to be recommended, the weight of the keyword corresponding to the content to be recommended in the updated interest model is adjusted to obtain an adjusted model; wherein the adjusted model represents the updated interest model after the weight is adjusted; New content to be recommended is determined according to the adjusted model and the device type of the terminal device currently used by the user, and the new content to be recommended is pushed to the user.

14. A content recommendation device based on user interests, comprising: An acquisition unit, used to acquire current scene information and a pre-stored interest model; wherein the current scene information represents the spatiotemporal scene in which the user is currently located, and the interest model is a user portrait model representing the field of the user's interested content; An updating unit, configured to update the interest model according to the current scene information to obtain an updated interest model; The recommendation unit is used to determine the content to be recommended according to the updated interest model and the device type of the terminal device currently used by the user, and push the content to be recommended to the user.

15. An electronic device, comprising: at least one processor; as well as a memory communicatively connected to the at least one processor; wherein, The memory stores instructions that can be executed by the at least one processor, and the instructions are executed by the at least one processor to enable the at least one processor to perform the method according to any one of claims 1 to 13.

16. A non-transitory computer-readable storage medium storing computer instructions, wherein: The computer instructions are used to cause the computer to execute the method according to any one of claims 1-13.

17. A computer program product, wherein: The invention comprises a computer program, which implements the steps of the method according to any one of claims 1 to 13 when being executed by a processor.

Images