Content recommendation method, device, electronic device and storage medium

Abstract

The present application provides a content recommendation method, device, electronic device and storage medium, which relate to the fields of cloud computing and Internet big data technology. The method includes: obtaining characteristic information of a user, determining the user's role type based on the characteristic information; and determining the first recommended content recommended to the user based on the data recall method corresponding to the role type. The content recommendation method provided by the present application determines the data recall method according to the user's role type to recommend content, which can make the recommended content more compatible with the user and improve the recommendation effect.

Description

Technical Field

[0001] The present application relates to the technical fields of cloud computing and Internet big data, and in particular to a content recommendation method, device, electronic device and storage medium. Background Art

[0002] With the development of network technology and mobile communication technology, people are more and more accustomed to obtaining various information through the Internet to meet the needs of daily life. As a result, many content publishing platforms have emerged.

[0003] Content publishing platforms recommend content to users by obtaining information that users may need. At present, general content publishing platforms mainly push content based on users' historical operation behaviors. However, for knowledge content platforms, there are few historical user operation behaviors that can be obtained, which leads to a low match between the content recommended to users and the users, affecting the effect of content recommendation. Summary of the invention

[0004] The present application provides a content recommendation method, device, electronic device and storage medium to improve the matching degree between recommended content and users, thereby improving the recommendation effect.

[0005] On the one hand, the present application provides a content recommendation method, comprising:

[0006] Obtain the user's characteristic information, and determine the user's role type based on the characteristic information;

[0007] Based on the data recall method corresponding to the role type, determine the first recommended content to be recommended to the user.

[0008] Optionally, the characteristic information includes the number of contents of historical operation contents within the first time range, and determining the role type of the user according to the characteristic information includes:

[0009] If the amount of content exceeds a preset amount threshold, the user is determined as a user of the first role type.

[0010] Optionally, determining the first recommended content to be recommended to the user based on the data recall method corresponding to the role type includes:

[0011] If the user is a user of the first role type, first recommended content is determined to be recommended to the user based on historical operation content within a first time range.

[0012] Optionally, determining first recommended content to be recommended to the user based on historical operation content within the first time range includes:

[0013] Determine content attributes of historical operation content within a first time range, where the content attributes include user interests or user needs;

[0014] Based on the content attributes, a first recommended content is determined to be recommended to the user.

[0015] Optionally, determining the content attribute of the historical operation content in the first time range includes:

[0016] If the historical operation content in the first time range meets the preset condition, determining the content attribute as user interest;

[0017] The preconditions include at least one of the following:

[0018] The historical operation content in the first time range is the user's subscription content;

[0019] The similarity between the historical operation content of the first time range and the historical operation content of the user in the second time range exceeds a first similarity threshold; wherein the second time range is determined based on the first time range;

[0020] The relevance of the historical operation content in the first time range to the role type corresponding to the user does not exceed the relevance threshold.

[0021] Optionally, determining the first recommended content to be recommended to the user based on the content attribute includes:

[0022] If the content attribute is user interest, determining first recommended content recommended to the user based on the historical operation content of the user in a third time range;

[0023] If the content attribute is user demand, determining first recommended content to be recommended to the user based on the historical operation content of the user in a fourth time range;

[0024] Among them, the third time range is greater than the fourth time range.

[0025] Optionally, determining the role type of the user according to the characteristic information includes:

[0026] If the amount of content does not exceed a preset amount threshold, the user is determined as a user of the second role type.

[0027] Optionally, determining the first recommended content to be recommended to the user based on the data recall method corresponding to the role type includes:

[0028] If the user is a user of the second role type, the first recommended content recommended to the user is determined based on the user tag of the user.

[0029] Optionally, the user tag has a corresponding relationship with the knowledge field to which the user belongs; or the user tag includes a user tag corresponding to a subordinate knowledge field of the knowledge field to which the user belongs.

[0030] Optionally, determining the first recommended content to be recommended to the user based on the data recall method corresponding to the role type includes:

[0031] Based on the data recall method corresponding to the role type, determine the basic data to be recommended corresponding to the role type;

[0032] The plurality of contents in the basic data to be recommended are sorted according to at least one of data category or data recall priority, and a first recommended content to be recommended to the user is determined from the plurality of contents according to the sorting result.

[0033] Optionally, the method further comprises:

[0034] Determining second recommended content recommended to the user based on the relevant information of the user;

[0035] The relevant information includes at least one of the following:

[0036] The user's subscription content, the user's similar users' historical operation content, and the user's corresponding industry information; similar users are users whose similarity to the user exceeds a second similarity threshold.

[0037] On the other hand, the present application provides a content recommendation device, comprising:

[0038] A type determination module is used to obtain characteristic information of the user and determine the role type of the user according to the characteristic information;

[0039] The content recommendation module is used to determine the first recommended content to be recommended to the user based on the data recall method corresponding to the role type.

[0040] On the other hand, the present application provides an electronic device, including:

[0041] at least one processor; and

[0042] a memory communicatively connected to the at least one processor; wherein,

[0043] 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 so that the at least one processor can execute the method in any embodiment of the present application.

[0044] On the other hand, the present application provides a non-transitory computer-readable storage medium storing computer instructions, which are used to enable a computer to execute the method in any embodiment of the present application.

[0045] The present application provides a content recommendation method, device, electronic device and storage medium, which determine the role type of the user according to the characteristic information, and determine the content recommended to the user based on the data recall method corresponding to the role type: for users of the first role type, since the number of historical operation contents is large, the recommended content can be determined based on the historical operation contents; for users of the second role type, since the number of historical operation contents is small, the recommended content is determined based on the user tags, which can avoid the problem of poor recommendation effect when the historical operation contents are small, thereby improving the matching degree between the recommended content and the user. Furthermore, when determining the recommended content based on the historical operation contents, since the user's interests are long-term stable and the user's needs change rapidly, therefore, if the content attribute is the user's interests, the recommended content is determined based on the user's longer historical operation contents; if the content attribute is the user's needs, the recommended content is determined based on the user's shorter historical operation contents, thereby further improving the matching degree between the recommended content and the user, thereby improving the recommendation effect.

[0046] 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 application, nor is it intended to limit the scope of the present application. Other features of the present application will become easily understood through the following description. BRIEF DESCRIPTION OF THE DRAWINGS

[0047] The accompanying drawings are used to better understand the present solution and do not constitute a limitation of the present application.

[0048] Figure 1 A schematic diagram of a system architecture of a content recommendation method in an embodiment of the present application;

[0049] Figure 2 This is a flowchart of a content recommendation method in an embodiment of the present application;

[0050] Figure 3 This is a flowchart of a content recommendation method in an embodiment of the present application;

[0051] Figure 4 A schematic diagram of a content recommendation method in an embodiment of the present application;

[0052] Figure 5 This is a schematic diagram of a content recommendation device in an embodiment of the present application;

[0053] Figure 6 A block diagram of an electronic device used to implement an embodiment of the present application. DETAILED DESCRIPTION

[0054] The following is a description of exemplary embodiments of the present application in conjunction with the accompanying drawings, including various details of the embodiments of the present application 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 can be made to the embodiments described herein without departing from the scope and spirit of the present application. Similarly, for the sake of clarity and conciseness, the description of well-known functions and structures is omitted in the following description.

[0055] In order to more clearly demonstrate the content recommendation method provided in the embodiments of the present application, the application scenario of the content recommendation method provided in the embodiments of the present application is first introduced.

[0056] The content recommendation method provided by this application can be applied to content recommendation on various content publishing platforms, such as news and information content publishing platforms, professional knowledge content publishing platforms, e-commerce platforms, social media platforms, and internal platforms for employees of various departments to obtain relevant information. The characteristic information of platform users can be obtained through the server of the content publishing platform, and the role type of the user can be determined based on the characteristic information. Based on the data recall method corresponding to the role type, the recommended content can be determined and recommended to the user.

[0057] Figure 1 Schematic diagram of the system architecture of the content recommendation method in one embodiment of the present application; Figure 1 As shown, the server of the content publishing platform obtains the characteristic information of the platform users, determines the role type of each user according to the characteristic information of each user, determines the content recommended to the corresponding user based on the data recall method corresponding to each role type, and sends it to the user terminal of each user.

[0058] The present application embodiment provides a content recommendation method. Figure 2 1 is a flowchart of a content recommendation method according to an embodiment of the present application. The method can be applied to a content recommendation device, which can be deployed in a server or other processing device. In some possible implementations, the method can also be implemented by a processor calling a computer-readable instruction stored in a memory. Figure 2 As shown, including:

[0059] Step S201, obtaining characteristic information of the user, and determining the role type of the user according to the characteristic information.

[0060] The embodiment of the present application is introduced with the server as the execution subject. The server obtains the characteristic information of the user, wherein the characteristic information can be information representing the unique attributes of the user, including but not limited to the basic information of the user, occupation category, job characteristics, and information related to the user's historical operation content.

[0061] The specific method of obtaining the user's characteristic information may be to receive the characteristic information sent by the user terminal, or to retrieve the user's characteristic information from a preset database based on the user's identification, or to obtain the characteristic information through statistical analysis based on the user's historical operation behavior. The method of obtaining the characteristic information may be determined based on the specific content of the characteristic information, and this application does not limit this.

[0062] The specific implementation method of determining the role type of the user according to the characteristic information may include: the server pre-establishes an association relationship between the characteristic information and the role type, and obtains the role type of the user according to the characteristic information and the association relationship. In addition, the server may pre-configure multiple role types, and after obtaining the characteristic information of the user, match and calculate the characteristic information with the multiple role types respectively, and determine the role type corresponding to the characteristic information according to the matching results.

[0063] The role type may be a type pre-classified according to the characteristic information of multiple users. For example, users may be classified into different role types such as R&D, sales, product, and operation according to the characteristics of their job positions.

[0064] Step S202: determining the first recommended content to be recommended to the user based on the data recall method corresponding to the role type.

[0065] The association between the role type and the data recall method is preconfigured. After the user's role type is determined, the corresponding data recall method can be obtained according to the role type and the association. The recall basis is obtained according to the data recall method, and the data similar to the recall basis is recalled as the first recommended content, or the data similar to the recall basis is screened to obtain the first recommended content and sent to the user terminal.

[0066] Among them, data recall can be based on the characteristics of existing data, recalling data similar to the existing data in a large amount of data, and recommending these similar data to users. The data recall method in the embodiment of the present application can be a specific implementation method for obtaining the recall basis. Different data recall methods can obtain different recall bases, and the recalled data can be recommended to users based on different recall bases.

[0067] In the related art, in the scenario where the recommended content is item information, the items that the user is interested in are determined based on the user's historical behavior on the items, and recommendations are made. However, this algorithm is not applicable in the recommendation scenario of knowledge content platforms such as documents and videos. First, for users in the knowledge content platform, for example, users related to technology research and development, their historical behavior may be very small, so the co-occurrence matrix of users and items will be very sparse, and subsequent algorithm steps cannot be performed. Secondly, the main purpose of users using the knowledge content platform is to consult certain product information and cases in the application field, so it is based on demand. Interests are relatively stable and change slowly, while the demand for information of knowledge content users changes very quickly. The accuracy of direct content recommendation based on rapidly changing user behavior is low. Another major type of recommendation algorithm is a deep learning algorithm based on user behavior sequences. This type of algorithm embeds the items in the user's historical click sequence into a fixed-dimensional vector space to obtain a fixed-dimensional vector, and then splices it with other features as a feature, and inputs it into the neural network to optimize the objective function. However, this type of algorithm is also based on user interests and requires more user clicks. For knowledge content platforms, the scenarios are mainly biased towards user needs, and users have fewer clicks, which cannot provide sufficient training data for neural network training.

[0068] The content recommendation method provided in this application determines the user's role type according to feature information, and determines the content recommended to the user based on the data recall method corresponding to the role type. It is applicable to a variety of different types of content publishing platforms, and can make the recommended content more closely matched with the user, thereby improving the recommendation effect.

[0069] In a possible implementation, the characteristic information in step S201 includes the number of historical operation contents in the first time range, and determining the role type of the user according to the characteristic information includes:

[0070] If the amount of content exceeds a preset amount threshold, the user is determined as a user of the first role type.

[0071] In practical applications, the role type of the user may be determined according to the number of historical operation contents of the user in the first time range. If the number of contents exceeds a preset threshold, the user is a user of the first role type.

[0072] Among them, the historical operation content can be the content corresponding to the historical operation behavior of the user, and the operation behavior can include click, search, etc. The first time range can be a time period before the current time, for example, 1 month, 2 months, etc. before the current time. The number of contents can be the sum of the number of contents clicked and searched by the user, or the number of contents can be determined separately according to the number of clicked contents or the number of searched contents. For example, if the user clicked on 10 articles and searched for 20 contents within 1 month before the current time, the number of contents is 30; or, if the user clicked on 25 articles within 2 months before the current time, the number of contents is 25. The first time range and quantity threshold can be configured according to specific needs.

[0073] In one example, for a knowledge content platform, users with roles related to products and sales have rich historical operation behaviors such as searching and clicking, and thus users with roles related to products and sales are determined as users of the first role type.

[0074] It should be noted that, in addition to users with roles related to products and sales, users of the first role type may also be any user with a large amount of historical operation content, and this application does not limit this.

[0075] In the technical solution of the present application, after determining the role type of the user, different data recall methods are used for content recommendation for different role types, as shown in the following embodiments:

[0076] In a possible implementation, step S202, based on the data recall method corresponding to the role type, determines the first recommended content recommended to the user, including:

[0077] If the user is a user of the first role type, first recommended content is determined to be recommended to the user based on historical operation content within a first time range.

[0078] In actual applications, if the user is determined to be a user of the first role type based on the feature information, since the number of historical operation contents of the user of this role type is relatively large, the first recommended content is determined based on the historical operation contents. Optionally, the recall basis is obtained based on the historical operation contents of the first time range, and after obtaining the contents similar to the recall basis, the contents similar to the recall basis can be used as the first recommended content; or, a part of the contents similar to the recall basis is selected and recommended to the user as the first recommended content.

[0079] The specific implementation method of determining the first recommended content based on the historical operation content within the first time range is shown in the following embodiment:

[0080] In a possible implementation, determining first recommended content to be recommended to the user based on historical operation content within a first time range includes:

[0081] Determine content attributes of historical operation content within a first time range, where the content attributes include user interests or user needs;

[0082] Based on the content attributes, a first recommended content is determined to be recommended to the user.

[0083] In actual applications, if it is determined that the user is a user of the first role type, the content attribute of the historical operation content in the first time range is judged, wherein the content attribute is an attribute that characterizes the relationship between the historical operation content and the user, and the relationship between the historical operation content and the user can be the user's attention level or attention frequency for the historical operation content, etc. The content attribute includes user interests and user needs. After judging whether the historical operation content belongs to user interests or user needs, different methods are adopted to determine the first recommended content according to different content attributes.

[0084] For how to determine the content attributes of the historical operation content, see the following embodiment:

[0085] In a possible implementation, determining the content attribute of the historical operation content in the first time range includes:

[0086] If the historical operation content in the first time range meets the preset condition, determining the content attribute as user interest;

[0087] The preconditions include at least one of the following:

[0088] The historical operation content in the first time range is the user's subscription content;

[0089] The similarity between the historical operation content of the first time range and the historical operation content of the user in the second time range exceeds a first similarity threshold; wherein the second time range is determined based on the first time range;

[0090] The relevance of the historical operation content in the first time range to the role type corresponding to the user does not exceed the relevance threshold.

[0091] In actual application, whether the historical operation content in the first time range meets the preset conditions is judged to determine whether the content attribute is user interest or user demand. If the preset conditions are met, the historical operation content belongs to user interest, otherwise, the historical operation content belongs to user demand.

[0092] Among them, the historical operation content in the first time range is the user's subscribed content. For example, the article clicked by the user belongs to the column subscribed by the user. The column subscribed by the user includes column content such as user attention and user collection that the user has long-term attention needs, so it is more inclined to user interests.

[0093] The similarity between the historical operation content in the first time range and the historical operation content in the second time range of the user exceeds the first similarity threshold, for example, for the historical operation content, whether the user has clicked on similar content in a longer time interval. If the user has clicked on the same type of content in a longer period of time, this stable click behavior can indicate that the historical operation content belongs to the user's interest. The second time range is determined based on the first time range, and can be determined based on the click time of the historical click content in the first time range.

[0094] The relevance of the historical operation content in the first time range to the role type corresponding to the user does not exceed the relevance threshold. For example, if the user is a user of a product-related role, and the historical operation content in the first time range belongs to organizational culture or public content unrelated to the product, then the historical operation content belongs to the user's interest.

[0095] The content attributes of the historical operation contents in the first time range are different, and the manner of determining the first recommended content is different, as shown in the following embodiments:

[0096] In a possible implementation, determining the first recommended content to be recommended to the user based on the content attribute includes:

[0097] If the content attribute is user interest, determining first recommended content recommended to the user based on the historical operation content of the user in a third time range;

[0098] If the content attribute is user demand, determining first recommended content to be recommended to the user based on the historical operation content of the user in a fourth time range;

[0099] Among them, the third time range is greater than the fourth time range.

[0100] In actual applications, after determining the content attributes, for historical operation contents whose content attributes are user needs, since the user needs change frequently, when determining the first recommended content, historical operation contents with a shorter time range from the current time are selected as the basis for recall. For example, the content clicked in the last one or two days can be selected as the basis for recall. For historical operation contents whose content attributes are user interests, historical operation contents with a longer time range from the current time are selected as the basis for recall. After determining the basis for recall, similar contents are recalled and added to the candidate list for recommendation, and the content is selected from the candidate list as the first recommended content.

[0101] In the embodiment of the present application, the user's historical operation content is further subdivided into user needs and user interests. When determining the basis for recall, since user needs change very quickly, only the most recent historical operation content is retained as the basis for recall; while user interests are more long-term and stable, they will be retained and tracked for a long time as the basis for recall. By distinguishing between user needs and user interests, user interests can be retained and tracked for a long time, while quickly providing feedback on changes in user needs, thereby recommending more matching content to users.

[0102] In a possible implementation, in step S201, determining the role type of the user according to the characteristic information includes:

[0103] If the amount of content does not exceed a preset amount threshold, the user is determined as a user of the second role type.

[0104] In practical applications, the user's role type may be determined based on the number of historical operation contents of the user in the first time range. If the number of contents does not exceed a preset threshold, the user is a user of the second role type.

[0105] In one example, for a knowledge content platform, users related to technology research and development have very few search and click behavior records, so users with roles related to technology research and development are determined as users of the second role type.

[0106] It should be noted that, in addition to users with roles related to technology research and development, users of the second role type may also be any user with a relatively small amount of historical operation content, and this application does not limit this.

[0107] For the user of the second role type, the method of determining the first recommended content is as follows:

[0108] In a possible implementation, in step S202, determining the first recommended content to be recommended to the user based on the data recall method corresponding to the role type includes:

[0109] If the user is a user of the second role type, the first recommended content recommended to the user is determined based on the user tag of the user.

[0110] In actual applications, if the user is a user of the second role type, since the number of historical operation contents in the first time range is small, if recommendations are made based on the historical operation contents in the first time range, sufficient recall basis may not be obtained. Therefore, the first recommended content is determined based on the user tag. Optionally, based on the keywords in the user tag, content matching the user tag is queried, and all or part of the content is selected as the first recommended content.

[0111] In one example, for R&D-related users with few clicks, the recall method based on user tags will be mainly relied on. The corresponding user tags are set for the users in advance. The user tags can be product tags of the products that the users participate in R&D, so that the users are associated with the products, and the users are recommended content related to the products.

[0112] How to configure the user tag is described in detail in the following embodiments:

[0113] In a possible implementation manner, the user tag of the user has a corresponding relationship with the knowledge field to which the user belongs.

[0114] In actual applications, user tags can be configured according to the knowledge domain to which the user belongs. The knowledge domain can be a domain related to the user's role type, for example, the department to which the user's job position belongs. Establishing a correspondence between the user tag and the knowledge domain to which the user belongs, and recommending content based on the user tag can be a method of recommending content to the user based on the knowledge domain.

[0115] In one example, user tags are configured according to the department to which the user's job position belongs, and the same user tag is configured for users in the same department. This processing method can make user tags radiate to more users and improve the efficiency of tag configuration.

[0116] In a possible implementation, the user tag of the user includes a user tag corresponding to a subordinate knowledge field of the knowledge field to which the user belongs.

[0117] In practical applications, in addition to the user tags corresponding to the knowledge field to which the user belongs, user tags may also include user tags corresponding to the subordinate knowledge field of the knowledge field to which the user belongs. For example, the subordinate knowledge field of the knowledge field may be a subordinate department of the current department, and the user tags of the superior department include the user tags of the subordinate department. Through such nested processing, user tags will radiate more users, and one user may correspond to multiple user tags. Content recommendation based on multiple user tags can increase the matching degree between the recommended content and the user.

[0118] In a possible implementation, in step S202, determining the first recommended content recommended to the user based on the data recall method corresponding to the role type includes:

[0119] Based on the data recall method corresponding to the role type, determine the basic data to be recommended corresponding to the role type;

[0120] The plurality of contents in the basic data to be recommended are sorted according to at least one of data category or data recall priority, and a first recommended content to be recommended to the user is determined from the plurality of contents according to the sorting result.

[0121] In actual applications, a recall basis is obtained based on the data recall method corresponding to the role type, and content similar to the recall basis is obtained as the basic data to be recommended. The basic data to be recommended may include multiple categories of content, and the categories can be divided according to different similarities with the recall basis, according to different recall methods, or according to other characteristics of each content. This application does not limit this.

[0122] Among them, the data recall priority can be the priority corresponding to each recall method. It can be understood that the priorities of the recall basis obtained by different recall methods are different, and accordingly, the priorities of the content similar to the recall basis are different. After sorting the multiple contents in the basic data to be recommended, the first recommended content is determined from the multiple contents according to the sorting result, so that the content that better meets the user's needs can be recommended, thereby improving the effect of content recommendation.

[0123] In a possible implementation, the method further includes:

[0124] Determining second recommended content recommended to the user based on the relevant information of the user;

[0125] The relevant information includes at least one of the following:

[0126] The user's subscription content, the user's similar users' historical operation content, and the user's corresponding industry information; similar users are users whose similarity to the user exceeds a second similarity threshold.

[0127] In actual applications, in addition to determining the first recommended content recommended to the user based on the data recall method corresponding to the role type, the second recommended content can also be determined based on the user's relevant information to enrich the recommended content and meet the diverse needs of the user. Specifically, the content similar to the user's subscription content, the historical operation content of the user's similar users, and the industry information corresponding to the user are obtained, and at least one of them is recommended to the user as the second recommended content. Among them, the industry information may include: the latest hot information of the industry that has been deeply processed; the latest headlines including the development trend information of the industry; the latest intelligence information of the industry, etc.

[0128] Figure 3 Flow chart of a content recommendation method in one embodiment of the present application. Figure 3 As shown, the method includes:

[0129] Step S301: Acquire the quantity of historical operation contents of a user within a first time range.

[0130] Step S302: determine whether the content quantity exceeds a preset quantity threshold.

[0131] Step S303: If the amount of content exceeds a preset amount threshold, the user is determined as a user of the first role type.

[0132] Step S304: If the amount of content does not exceed a preset amount threshold, the user is determined as a user of the second role type.

[0133] Step S305: If the user is a user of the first role type, it is determined whether the content attribute of the historical operation content within the first time range is a user demand or a user interest.

[0134] Step S306: If the content attribute is user interest, determine the first recommended content recommended to the user based on the historical operation content of the user in the third time range.

[0135] Step S307: if the content attribute is user demand, determine the first recommended content recommended to the user based on the historical operation content of the user in the fourth time range. The third time range is greater than the fourth time range.

[0136] Step S308: If the user is a user of the second role type, determining the first recommended content recommended to the user based on the user tag of the user.

[0137] The content recommendation method provided in the present application determines the role type of the user according to the content quantity of the historical operation content within the first time range, adopts different data recall methods for different role types, and determines the content recommended to the user, which can make the recommended content more compatible with the user and improve the recommendation effect.

[0138] Figure 4 This is a schematic diagram of the content recommendation method in an embodiment of the present application. In the embodiment of the present application, the content recommendation method is applied to the knowledge management platform of enterprises and institutions. The knowledge management platform stores a huge amount of product information, product dynamics, sales cases, organizational culture and other content. In order to allow employee users to obtain the information they need and are interested in, a content (which can be a feed stream) recommendation function is provided. At the same time, since users come from different positions such as research and development, sales, products, and operations, they have different needs and access methods. Sales-related users have relatively rich historical behavior records such as searches and clicks. However, since sales-related users, such as pre-sales architects, solution architects, etc., will change the fields and products involved, their demand for content is changing rapidly. For development-related users, there are few historical behavior records such as searches and clicks, but since these users are responsible for the research and development of certain products, they are associated with these products, and accurate product labels are configured for these users.

[0139] like Figure 4As shown, a database of data to be recommended is pre-built, and the database includes data such as historical click content, user subscription information, user identity information, product information, industry information, industry headlines, and industry intelligence of multiple users. The user's role type is determined according to the number of historical click content of the user. For sales-related users, there are more historical behavior records, so the historical click records are classified, that is, the content attribute of the historical click content is determined to be user demand or user interest, and the data recall method is determined according to the content attribute. Recall from the database based on historical clicks. If the content attribute is user interest, the long-term historical click content is used as the recall basis to recall similar content from the database; if the content attribute is user demand, the short-term historical click content is used as the recall basis to recall similar content from the database. For development-related users, there are fewer historical behavior records, so user tags are generated, and similar content is recalled from the database based on user tags. In addition, whether it is a development-related user or a sales-related user, data can be recalled from the database based on similar user recall, user subscription recall, latest popular information recall, latest headline recall, and latest intelligence recall. The data recalled through various methods are used as the basic data to be recommended. The basic data to be recommended is sorted and rearranged according to the document category, or sorted according to the recall priority, and content recommendation is performed based on the sorting results. When sorting and rearranging, for sales-related users, due to the large number of historical behavior records, the content recalled based on the user's historical behavior will be ranked at the front. For development-related users, due to their lack of historical click behavior but with more user tags, the content recalled based on user tags will be ranked at the front during the sorting and rearrangement stage. In this way, the recommended content will be more closely matched with the user and the recommendation effect will be better.

[0140] Figure 5 Schematic diagram of a content recommendation device in one embodiment of the present application. Figure 5 As shown, the content recommendation device may include:

[0141] The type determination module 501 is used to obtain the characteristic information of the user and determine the role type of the user according to the characteristic information;

[0142] The content recommendation module 502 is used to determine the first recommended content to be recommended to the user based on the data recall method corresponding to the role type.

[0143] The content recommendation device provided in the present application determines the role type of the user according to the characteristic information, and determines the content recommended to the user based on the data recall method corresponding to the role type, so that the recommended content can be more closely matched with the user and the recommendation effect can be improved.

[0144] In a possible implementation, the characteristic information includes the amount of content of historical operation content within a first time range, and the type determination module 501 includes a type determination unit configured to:

[0145] If the amount of content exceeds a preset amount threshold, the user is determined as a user of the first role type.

[0146] In a possible implementation, the content recommendation module 502 is used to:

[0147] If the user is a user of the first role type, first recommended content is determined to be recommended to the user based on historical operation content within a first time range.

[0148] In a possible implementation, the content recommendation module 502 includes an attribute determination unit and a content recommendation unit;

[0149] An attribute determination unit, configured to determine content attributes of historical operation content within a first time range, wherein the content attributes include user interests or user needs;

[0150] The content recommendation unit is used to determine the first recommended content recommended to the user based on the content attribute.

[0151] In a possible implementation manner, the attribute determination unit is specifically configured to:

[0152] If the historical operation content in the first time range meets the preset condition, determining the content attribute as user interest;

[0153] The preconditions include at least one of the following:

[0154] The historical operation content in the first time range is the user's subscription content;

[0155] The similarity between the historical operation content of the first time range and the historical operation content of the user in the second time range exceeds a first similarity threshold; wherein the second time range is determined based on the first time range;

[0156] The relevance of the historical operation content in the first time range to the role type corresponding to the user does not exceed the relevance threshold.

[0157] In a possible implementation, the content recommendation unit is specifically configured to:

[0158] If the content attribute is user interest, determining first recommended content recommended to the user based on the historical operation content of the user in a third time range;

[0159] If the content attribute is user demand, determining first recommended content to be recommended to the user based on the historical operation content of the user in a fourth time range;

[0160] Among them, the third time range is greater than the fourth time range.

[0161] In a possible implementation manner, the type determination unit is configured to:

[0162] If the amount of content does not exceed a preset amount threshold, the user is determined as a user of the second role type.

[0163] In a possible implementation, the content recommendation module 502 is used to:

[0164] If the user is a user of the second role type, the first recommended content recommended to the user is determined based on the user tag of the user.

[0165] In a possible implementation, the user tag has a corresponding relationship with the knowledge field to which the user belongs; or

[0166] The user tags include user tags corresponding to the subordinate knowledge fields of the knowledge field to which the user belongs.

[0167] In a possible implementation, the content recommendation module 502 is used to:

[0168] Based on the data recall method corresponding to the role type, determine the basic data to be recommended corresponding to the role type;

[0169] The plurality of contents in the basic data to be recommended are sorted according to at least one of data category or data recall priority, and a first recommended content to be recommended to the user is determined from the plurality of contents according to the sorting result.

[0170] In a possible implementation, the content recommendation module 502 is further configured to:

[0171] Determining second recommended content recommended to the user based on the relevant information of the user;

[0172] The relevant information includes at least one of the following:

[0173] The user's subscription content, the user's similar users' historical operation content, and the user's corresponding industry information; similar users are users whose similarity to the user exceeds a second similarity threshold.

[0174] The functions of each unit, module or sub-module in each device in the embodiments of the present application can refer to the corresponding description in the above-mentioned content recommendation method embodiment, and will not be repeated here.

[0175] According to another aspect of the present application, there is provided an electronic device, including:

[0176] at least one processor; and

[0177] a memory communicatively connected to the at least one processor; wherein,

[0178] 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 so that the at least one processor can execute the method in any embodiment of the present application.

[0179] According to another aspect of the present application, a non-transitory computer-readable storage medium storing computer instructions is provided, wherein the computer instructions are used to enable a computer to execute the method in any embodiment of the present application.

[0180] Figure 6 FIG. 1 is a block diagram of an electronic device used to implement an embodiment of the present application. Figure 6 As shown, the electronic device includes: a memory 610 and a processor 620. The memory 610 stores a computer program that can be run on the processor 620. When the processor 620 executes the computer program, the method in the above embodiment is implemented. The number of the memory 610 and the processor 620 can be one or more.

[0181] The electronic device also includes:

[0182] The communication interface 630 is used to communicate with external devices and perform data exchange transmission.

[0183] If the memory 610, the processor 620 and the communication interface 630 are implemented independently, the memory 610, the processor 620 and the communication interface 630 can be connected to each other through a bus and communicate with each other. The bus can be an Industry Standard Architecture (ISA) bus, a Peripheral Component Interconnect (PCI) bus or an Extended Industry Standard Architecture (EISA) bus. The bus can be divided into an address bus, a data bus, a control bus, etc. For ease of representation, Figure 6 Only one thick line is used in the diagram, but this does not mean that there is only one bus or only one type of bus.

[0184] Optionally, in a specific implementation, if the memory 610, the processor 620 and the communication interface 630 are integrated on a chip, the memory 610, the processor 620 and the communication interface 630 can communicate with each other through an internal interface.

[0185] An embodiment of the present application provides a computer-readable storage medium storing a computer program, which implements the method provided in the embodiment of the present application when the program is executed by a processor.

[0186] An embodiment of the present application also provides a chip, which includes a processor for calling and executing instructions stored in the memory from the memory, so that a communication device equipped with the chip executes the method provided by the embodiment of the present application.

[0187] An embodiment of the present application also provides a chip, including: an input interface, an output interface, a processor and a memory, wherein the input interface, the output interface, the processor and the memory are connected via an internal connection path, and the processor is used to execute the code in the memory. When the code is executed, the processor is used to execute the method provided in the embodiment of the application.

[0188] It should be understood that the processor may be a central processing unit (CPU), or other general-purpose processors, digital signal processors (DSP), application specific integrated circuits (ASIC), field programmable gate arrays (FPGA) or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. The general-purpose processor may be a microprocessor or any conventional processor, etc. It is worth noting that the processor may be a processor supporting the advanced RISC machines (ARM) architecture.

[0189] Further, optionally, the above-mentioned memory may include a read-only memory and a random access memory, and may also include a non-volatile random access memory. The memory may be a volatile memory or a non-volatile memory, or may include both volatile and non-volatile memories. Among them, the non-volatile memory may include a read-only memory (ROM), a programmable read-only memory (PROM), an erasable programmable read-only memory (EPROM), an electrically erasable programmable read-only memory (EEPROM), or a flash memory. The volatile memory may include a random access memory (RAM), which is used as an external cache. By way of example but not limitation, many forms of RAM are available. For example, static RAM (SRAM), dynamic random access memory (DRAM), synchronous DRAM (SDRAM), double data rate synchronous dynamic random access memory (DDR SDRAM), enhanced synchronous dynamic random access memory (ESDRAM), synchronous link dynamic random access memory (SLDRAM) and direct rambus RAM (DR RAM).

[0190] In the above embodiments, it can be implemented in whole or in part by software, hardware, firmware or any combination thereof. When implemented using software, it can be implemented in whole or in part in the form of a computer program product. The computer program product includes one or more computer instructions. When the computer program instructions are loaded and executed on a computer, the process or function according to the present application is generated in whole or in part. The computer can be a general-purpose computer, a special-purpose computer, a computer network, or other programmable devices. The computer instructions can be stored in a computer-readable storage medium, or transmitted from one computer-readable storage medium to another computer-readable storage medium.

[0191] In the description of this specification, the description with reference to the terms "one embodiment", "some embodiments", "example", "specific example", or "some examples" means that the specific features, structures, materials or characteristics described in conjunction with the embodiment or example are included in at least one embodiment or example of the present application. Moreover, the specific features, structures, materials or characteristics described may be combined in any one or more embodiments or examples in a suitable manner. In addition, those skilled in the art may combine and combine different embodiments or examples described in this specification and the features of different embodiments or examples, unless they are contradictory.

[0192] In addition, the terms "first" and "second" are used for descriptive purposes only and should not be understood as indicating or implying relative importance or implicitly indicating the number of technical features indicated. Therefore, a feature defined as "first" or "second" may explicitly or implicitly include at least one of the features. In the description of this application, the meaning of "plurality" is two or more, unless otherwise clearly and specifically defined.

[0193] Any process or method description in the flow chart or otherwise described herein can be understood to represent a module, fragment or portion of a code including one or more executable instructions for implementing the steps of a specific logical function or process. And the scope of the preferred embodiment of the present application includes other implementations, in which the functions may not be performed in the order shown or discussed, including in a substantially simultaneous manner or in a reverse order according to the functions involved.

[0194] The logic and / or steps represented in the flowchart or otherwise described herein, for example, can be considered as an ordered list of executable instructions for implementing logical functions, which can be embodied in any computer-readable medium for use by an instruction execution system, apparatus or device (such as a computer-based system, a system including a processor or other system that can fetch instructions from an instruction execution system, apparatus or device and execute instructions), or used in combination with these instruction execution systems, apparatuses or devices.

[0195] It should be understood that the various parts of the present application can be implemented with hardware, software, firmware or a combination thereof. In the above embodiments, multiple steps or methods can be implemented with software or firmware stored in a memory and executed by a suitable instruction execution system. All or part of the steps of the above embodiment method can be completed by instructing the relevant hardware through a program, which can be stored in a computer-readable storage medium, and when the program is executed, it includes one of the steps of the method embodiment or a combination thereof.

[0196] In addition, each functional unit in each embodiment of the present application can be integrated into a processing module, or each unit can exist physically separately, or two or more units can be integrated into one module. The above-mentioned integrated module can be implemented in the form of hardware or in the form of a software functional module. If the above-mentioned integrated module is implemented in the form of a software functional module and sold or used as an independent product, it can also be stored in a computer-readable storage medium. The storage medium can be a read-only memory, a disk or an optical disk, etc.

[0197] The above is only a specific implementation of the present application, but the protection scope of the present application is not limited thereto. Any technician familiar with the technical field can easily think of various changes or substitutions within the technical scope disclosed in the present application, which should be included in the protection scope of the present application. Therefore, the protection scope of the present application should be based on the protection scope of the claims.

Claims

1. A content recommendation method, characterized in that: The method comprises: Acquire characteristic information of a user, and determine the role type of the user according to the characteristic information; the characteristic information includes the number of contents of historical operation contents within a first time range; if the number of contents exceeds a preset number threshold, determine the user as a user of the first role type; Based on the data recall method corresponding to the role type, if the user is a user of the first role type, based on the historical operation content within the first time range, determining first recommended content recommended to the user; determining content attributes of the historical operation content within the first time range, the content attributes including user interests or user needs; and determining first recommended content recommended to the user based on the content attributes; Among them, if the content attribute is user interest, the first recommended content recommended to the user is determined based on the historical operation content of the user within a third time range; if the content attribute is user demand, the first recommended content recommended to the user is determined based on the historical operation content of the user within a fourth time range; wherein the third time range is larger than the fourth time range.

2. The method according to claim 1, characterized in that in, The determining of the content attributes of the historical operation content within the first time range includes: If the historical operation content within the first time range meets a preset condition, determining that the content attribute is user interest; The preset condition includes at least one of the following: The historical operation content within the first time range is the subscription content of the user; The similarity between the historical operation content of the first time range and the historical operation content of the user in the second time range exceeds a first similarity threshold; wherein the second time range is determined based on the first time range; The relevance between the historical operation content within the first time range and the role type corresponding to the user does not exceed a relevance threshold.

3. The method according to claim 1, characterized in that in, The determining the role type of the user according to the characteristic information includes: If the content quantity does not exceed a preset quantity threshold, the user is determined as a user of the second role type.

4. The method according to claim 3, characterized in that in, The determining, based on the data recall method corresponding to the role type, the first recommended content recommended to the user includes: If the user is a user of the second role type, a first recommended content recommended to the user is determined based on the user tag of the user.

5. The method according to claim 4, characterized in that in, The user tag has a corresponding relationship with the knowledge field to which the user belongs; or The user tags include user tags corresponding to subordinate knowledge fields of the knowledge field to which the user belongs.

6. The method according to claim 1, characterized in that in, The method further comprises: Determining second recommended content recommended to the user based on the relevant information of the user; The relevant information includes at least one of the following: The user's subscription content, the user's similar user's historical operation content, and the user's corresponding industry information; the similar user is a user whose similarity with the user exceeds a second similarity threshold.

7. A content recommendation device, characterized in that: The device comprises: A type determination module is used to obtain characteristic information of a user and determine the role type of the user according to the characteristic information; the characteristic information includes the number of contents of historical operation contents within a first time range; if the number of contents exceeds a preset number threshold, the user is determined as a user of the first role type; a content recommendation module, configured to determine, based on the data recall method corresponding to the role type, if the user is a user of the first role type, first recommended content recommended to the user based on historical operation content within the first time range; determine content attributes of the historical operation content within the first time range, the content attributes including user interests or user needs; and determine first recommended content recommended to the user based on the content attributes; Among them, if the content attribute is user interest, the first recommended content recommended to the user is determined based on the historical operation content of the user within a third time range; if the content attribute is user demand, the first recommended content recommended to the user is determined based on the historical operation content of the user within a fourth time range; wherein the third time range is larger than the fourth time range.

8. An electronic device, characterized in that: include: 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 6.

Images