Knowledge recommendation method, device, electronic device and storage medium

Abstract

The present invention discloses a knowledge recommendation method, device, electronic device and storage medium. The method includes: obtaining a knowledge set and behavioral information of a target user, wherein the knowledge set includes multiple knowledge documents; determining candidate knowledge documents and predicting the value of the candidate knowledge documents to the target user, wherein the candidate knowledge documents are knowledge documents that are not covered by the target user in the knowledge set; constructing the original behavior matrix of the target user based on the behavioral information; determining the target behavior matrix of the target user based on the value of the candidate knowledge documents to the target user and the original behavior matrix; determining a knowledge recommendation list based on the target behavior matrix, and recommending the knowledge recommendation list to the target user, wherein the knowledge recommendation list includes at least one knowledge document. The solution provided by the present invention can solve the problem of inaccurate knowledge recommendation results caused by the inability of a single user's reading behavior to effectively cover the entire amount of knowledge, thereby realizing the recommendation of diverse knowledge.

Description

Technical Field

[0001] The present invention relates to the field of database management technology, and in particular to a knowledge recommendation method, device, electronic device and storage medium. Background Art

[0002] With the rapid development of information technology in the big data era, the amount of information generated by enterprises and other institutions is exponentially increasing. The focus of enterprise knowledge base development has gradually shifted from transforming physical data into information and knowledge, to avoiding information overload and improving knowledge output capabilities. The ability to fully utilize existing resources and recommend the right knowledge to the right people at the right time, thereby helping users make the right decisions, has become a key criterion for evaluating the success of enterprise knowledge base development. Compared to other businesses, commercial banks have complex user institution restrictions, weak knowledge relevance, and a high demand for knowledge timeliness. Therefore, knowledge recommendation must balance accuracy, timeliness, and diversity.

[0003] Existing knowledge recommendation methods typically recommend knowledge to target users based on their historical click data. However, in enterprise-level knowledge bases storing massive amounts of knowledge, individual reading behavior rarely effectively covers the entire knowledge base. Consequently, knowledge recommendations are limited to the knowledge or catalogs that individuals have read, easily leading to a knowledge prison. Summary of the Invention

[0004] The present invention provides a knowledge recommendation method, device, electronic device and storage medium, which can solve the problem of inaccurate knowledge recommendation results caused by the inability of a single user's reading behavior to effectively cover the entire amount of knowledge, and realize the recommendation of diverse knowledge.

[0005] According to one aspect of the present invention, a knowledge recommendation method is provided, comprising:

[0006] Acquire a knowledge set and target user behavior information, wherein the knowledge set includes multiple knowledge documents;

[0007] Determine candidate knowledge documents and predict the value of candidate knowledge documents to target users, where candidate knowledge documents are knowledge documents that are not covered by the target users in the knowledge set;

[0008] Based on the behavioral information, construct the original behavior matrix of the target user;

[0009] Determine the target behavior matrix of the target user based on the value of the candidate knowledge document to the target user and the original behavior matrix;

[0010] A knowledge recommendation list is determined according to the target behavior matrix, and the knowledge recommendation list is recommended to the target user, wherein the knowledge recommendation list includes at least one knowledge document.

[0011] Optionally, each knowledge document has a directory attribute, and knowledge documents with the same directory attribute belong to the same directory;

[0012] Predict the value of candidate knowledge documents to target users, including:

[0013] Determining the contribution of other users to the directory to which the candidate knowledge document belongs, wherein other users are users who participated in creating, revising, or editing the candidate knowledge document;

[0014] The value of the candidate knowledge document to the target user is predicted based on the contribution of other users to the directory to which the candidate knowledge document belongs and the value of the candidate knowledge document to other users.

[0015] Optionally, the behavior information includes the start time and end time of reading the knowledge document, evaluation information, and recommendation result feedback information;

[0016] Based on the behavioral information, construct the original behavior matrix of the target user, including:

[0017] Generate a user behavior sequence based on the target user's start and end time of reading the knowledge document;

[0018] Expand the user behavior sequence based on the target user's feedback on the recommended results of the knowledge document;

[0019] According to the expanded user behavior sequence and the target user's evaluation information on the knowledge document, the original behavior matrix is ​​constructed.

[0020] Optionally, based on the value of the candidate knowledge document to the target user and the original behavior matrix, a target behavior matrix for the target user is determined, including:

[0021] According to the value of the candidate knowledge documents to the target users, the original behavior matrix is ​​reduced in dimension;

[0022] The target behavior matrix is ​​determined based on the original behavior matrix after dimensionality reduction and the time decay factor.

[0023] Optionally, determine a knowledge recommendation list based on the target behavior matrix, including:

[0024] Identify similar users to the target user based on the target behavior matrix;

[0025] Obtain a list of similar users’ preference knowledge;

[0026] Determine the knowledge recommendation list based on the preferred knowledge list.

[0027] Optionally, determine similar users to the target user based on the target behavior matrix, including:

[0028] According to the target behavior matrix, determine the cosine value of the knowledge value vector between the target user and other users;

[0029] Determine the similarity between the target user and other users based on the cosine value of the knowledge value vector and the correction factor;

[0030] If the similarity is greater than or equal to the preset threshold, other users are considered as similar users.

[0031] Optionally, based on the preferred knowledge list, determine a knowledge recommendation list, including:

[0032] Filter knowledge documents from the preferred knowledge list based on catalog relevance;

[0033] Make timeliness corrections to the screened knowledge documents and determine the knowledge recommendation list.

[0034] According to another aspect of the present invention, there is provided a knowledge recommendation device comprising: an acquisition module, a value prediction module, a matrix determination module and a recommendation module;

[0035] An acquisition module, configured to acquire a knowledge set and target user behavior information, wherein the knowledge set includes a plurality of knowledge documents;

[0036] A value prediction module is used to determine candidate knowledge documents and predict the value of the candidate knowledge documents to the target user, wherein the candidate knowledge documents are knowledge documents that the target user has not yet encountered in the knowledge set;

[0037] A matrix determination module is used to construct an original behavior matrix of the target user based on the behavior information; and to determine a target behavior matrix of the target user based on the value of the candidate knowledge document to the target user and the original behavior matrix;

[0038] The recommendation module is used to determine a knowledge recommendation list according to the target behavior matrix and recommend the knowledge recommendation list to the target user, wherein the knowledge recommendation list includes at least one knowledge document.

[0039] According to another aspect of the present invention, an electronic device is provided, the electronic device comprising:

[0040] at least one processor; and

[0041] a memory communicatively connected to at least one processor; wherein,

[0042] The memory stores a computer program that can be executed by at least one processor. The computer program is executed by the at least one processor so that the at least one processor can execute the knowledge recommendation method of any embodiment of the present invention.

[0043] According to another aspect of the present invention, a computer-readable storage medium is provided. The computer-readable storage medium stores computer instructions, and the computer instructions are used to enable a processor to implement the knowledge recommendation method of any embodiment of the present invention when executed.

[0044] The technical solution of the embodiment of the present invention obtains information about the knowledge set and the behavior of the target user, predicts the value of knowledge documents in the knowledge set that the target user has not yet covered, and constructs the target user's original behavior matrix. Based on the value of the candidate knowledge documents to the target user and the original behavior matrix, the target user's target behavior matrix is ​​determined, and then a knowledge recommendation list is determined and recommended to the target user. In this way, the complex professional requirements and cross-professional professional knowledge requirements of users in the commercial banking field are fully considered. After the target user logs in, the target user can be recommended knowledge that may need to be consulted based on the usage behavior of similar users. This solves the problem of inaccurate knowledge recommendation results in knowledge bases storing massive amounts of knowledge, which is caused by the inability of a single user's reading behavior to effectively cover the entire knowledge base, thereby achieving diverse knowledge recommendation. At the same time, the introduction of a time decay factor in the process of determining the target behavior matrix can effectively distinguish user behavior in the time dimension, ensuring the timeliness of knowledge recommendation. In addition, the user behavior information obtained by the present invention includes user feedback on the recommended results of knowledge documents, which provides a direct feedback channel for users and provides data support for empirical parameter tuning.

[0045] 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 invention, nor is it intended to limit the scope of the present invention. Other features of the present invention will become readily understood through the following description. BRIEF DESCRIPTION OF THE DRAWINGS

[0046] In order to more clearly illustrate the technical solutions in the embodiments of the present invention, the following briefly introduces the drawings required for use in the description of the embodiments. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work.

[0047] Figure 1 This is a flow chart of a knowledge recommendation method provided in the first embodiment of the present invention;

[0048] Figure 2 This is a flow chart of a knowledge recommendation method provided in the second embodiment of the present invention;

[0049] Figure 3 This is a schematic diagram of a dimensionality reduction process for an original behavior matrix provided by the second embodiment of the present invention;

[0050] Figure 4 This is a schematic diagram of obtaining directed associations between different knowledge directories provided in the second embodiment of the present invention;

[0051] Figure 5 This is a schematic diagram of the structure of a knowledge recommendation device provided in the third embodiment of the present invention;

[0052] Figure 6 This is a structural diagram of an electronic device provided in Example 4 of the present invention. DETAILED DESCRIPTION

[0053] In order to enable those skilled in the art to better understand the solutions of the present invention, the technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the embodiments described are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative efforts should fall within the scope of protection of the present invention.

[0054] It should be noted that the terms "original", "candidate", "target", etc. in the specification and claims of the present invention and the above-mentioned drawings are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence. It should be understood that the data used in this way can be interchangeable where appropriate, so that the embodiments of the present invention described herein can be implemented in an order other than those illustrated or described herein. In addition, the terms "including" and "having" and any variations thereof are intended to cover non-exclusive inclusions, for example, a process, method, system, product or device that includes a series of steps or units is not necessarily limited to those steps or units clearly listed, but may include other steps or units that are not clearly listed or inherent to these processes, methods, products or devices.

[0055] Example 1

[0056] Figure 1 This is a flow chart of a knowledge recommendation method provided by the first embodiment of the present invention. This embodiment is applicable to the case where knowledge is recommended to users based on a knowledge set. The method can be executed by a knowledge recommendation device, which can be implemented in the form of hardware and / or software. The knowledge recommendation device can be configured in an electronic device (such as a computer or server). Figure 1 As shown, the method includes:

[0057] S110: Acquire a knowledge set and behavior information of a target user, wherein the knowledge set includes a plurality of knowledge documents.

[0058] A knowledge set can also be understood as a knowledge base. The knowledge set mentioned in this invention can be the knowledge set of a specific enterprise or an open, shared knowledge set for all enterprises. A knowledge set stores multiple knowledge documents, each of which has a directory attribute. Knowledge documents with the same directory attribute are grouped into the same directory.

[0059] The target user is the user currently logged into the system (e.g., a commercial bank system) and is the user for whom knowledge recommendations are to be made. The target user's behavior information includes at least the target user's start and end time of reading the knowledge document, evaluation information, and recommendation result feedback. Specifically, the target user's behavior information can be extracted from the target user's user log.

[0060] S120: Determine candidate knowledge documents and predict the value of the candidate knowledge documents to the target user, wherein the candidate knowledge documents are knowledge documents that are not involved in the target user in the knowledge set.

[0061] Because a knowledge set includes multiple knowledge documents, a single user is unlikely to fully access (including at least one of browsing, creating, revising, and editing) all knowledge documents in the knowledge set. Furthermore, different users may access different knowledge documents. Therefore, when recommending knowledge to a target user, it is necessary to identify knowledge documents from the knowledge set that the target user has not accessed, i.e., candidate knowledge documents, and predict the value of the candidate knowledge documents to the target user.

[0062] Specifically, the method for predicting the value of a candidate knowledge document to a target user may include: determining the contribution of other users to the directory to which the candidate knowledge document belongs, wherein other users are users who participate in creating, revising, or editing the candidate knowledge document; and predicting the value of the candidate knowledge document to the target user based on the contribution of other users to the directory to which the candidate knowledge document belongs and the value of the candidate knowledge document to other users.

[0063] S130: Construct an original behavior matrix of the target user based on the behavior information.

[0064] The original behavior matrix may be a three-dimensional matrix, reflecting the value of the knowledge documents involved by the target users to the target users.

[0065] Specifically, the original behavior matrix of the target user can be constructed based on the target user's reading start time, reading end time, evaluation information and recommendation result feedback information of the knowledge document, with the target user's identification (such as ID), knowledge document identification and user behavior sequence as three dimensions.

[0066] It should be noted that there is no specific order in which steps S120 and S130 must be executed. That is, step S120 may be executed first, followed by step S130; step S130 may be executed first, followed by step S120; or steps S120 and S130 may be executed simultaneously. This is not specifically limited in the present embodiment.

[0067] S140: Determine a target behavior matrix for the target user based on the value of the candidate knowledge document to the target user and the original behavior matrix.

[0068] The target behavior matrix may be a two-dimensional matrix, which can quantify the user's behavior toward the knowledge document into a value score of the knowledge document to the user.

[0069] The value of the candidate knowledge document to the target user can be used to reduce the dimensionality of the original behavior matrix, converting the original three-dimensional matrix into a two-dimensional matrix, thereby improving the matrix over-sparseness problem in the process of converting the three-dimensional matrix into a two-dimensional matrix.

[0070] S150 . Determine a knowledge recommendation list according to the target behavior matrix, and recommend the knowledge recommendation list to the target user, wherein the knowledge recommendation list includes at least one knowledge document.

[0071] Specifically, the method for determining a knowledge recommendation list based on a target behavior matrix may include: determining similar users of a target user based on the target behavior matrix; obtaining a preferred knowledge list of the similar users; and determining a knowledge recommendation list based on the preferred knowledge list. Similar users of the target user are mined through the target behavior matrix, and a knowledge recommendation list is generated for the target user based on the preferred knowledge lists of the similar users, thereby implementing knowledge recommendation.

[0072] In this way, the complex professional limitations and cross-limit professional knowledge requirements of users in the commercial banking field can be fully considered. After the target user logs in, the knowledge that may need to be consulted can be recommended to the target user based on the usage behavior of similar users. This solves the problem of inaccurate knowledge recommendation results in a knowledge base that stores massive knowledge because the reading behavior of a single user cannot effectively cover the entire amount of knowledge, and realizes the diversity of knowledge recommendation.

[0073] Example 2

[0074] Figure 2 This is a flow chart of a knowledge recommendation method provided by the second embodiment of the present invention. This embodiment provides a detailed knowledge recommendation method based on the above-mentioned first embodiment. Figure 2 As shown, the method includes:

[0075] S210: Acquire a knowledge set and behavior information of a target user, wherein the knowledge set includes a plurality of knowledge documents.

[0076] A knowledge set can also be understood as a knowledge base. The knowledge set mentioned in this invention can be the knowledge set of a specific enterprise or an open, shared knowledge set for all enterprises. A knowledge set stores multiple knowledge documents, each of which has a directory attribute. Knowledge documents with the same directory attribute are grouped into the same directory.

[0077] The target user is the user currently logged into the system (e.g., a commercial bank system) and is the user for whom knowledge recommendations are to be made. The target user's behavior information includes at least the target user's start and end time of reading the knowledge document, evaluation information, and recommendation result feedback. Specifically, the target user's behavior information can be extracted from the target user's user log.

[0078] S221. Determine candidate knowledge documents, where the candidate knowledge documents are knowledge documents that are not involved in the target user in the knowledge set.

[0079] Because a knowledge set includes multiple knowledge documents, a single user likely cannot fully access (including browsing, creating, revising, editing, and performing at least one of the following operations) all knowledge documents in the knowledge set. Furthermore, different users may access different knowledge documents. Therefore, it is necessary to first identify knowledge documents from the knowledge set that the target user has not accessed, i.e., candidate knowledge documents.

[0080] S222: Determine the contribution of other users to the directory to which the candidate knowledge document belongs, wherein other users are users who participate in creating, revising, or editing the candidate knowledge document.

[0081] Contribution is used to quantify the influence of other users on the creation, revision, or editing of the directory to which the candidate knowledge document belongs. If a user has a high contribution to a directory, it can be roughly considered that the user's behavior on the knowledge in that directory is more valuable for reference, and thus has a higher weight in the knowledge recommendation process.

[0082] Specifically, for any other user, all knowledge documents created, revised or edited by the other user are first merged into a single knowledge document, and the merged knowledge document is processed by irony and stop word removal; then, according to the Term Frequency-Inverse Document Frequency (TF-IDF) algorithm, the weight of each element in the word set after word segmentation is calculated, and combined with the search log of the other user and the user's own organization, limit and other labels, the contribution of the other user to a certain directory is determined.

[0083] Among them, the TF-IDF algorithm is a statistical method that is often used to evaluate the importance of a word (character) to a document in a knowledge document collection. It is a weighting technology commonly used in the fields of information retrieval and data mining.

[0084] S223: Predict the value of the candidate knowledge document to the target user based on the contribution of other users to the directory to which the candidate knowledge document belongs and the value of the candidate knowledge document to other users.

[0085] Optionally, when predicting the value of a candidate knowledge document to a target user, the average value of other knowledge documents to the target user and the influence of similarities between other users and the target user may be considered to comprehensively predict the value of the candidate knowledge document to the target user.

[0086] In this way, the problem of user tendency quantification error caused by the small coverage area of ​​a single user behavior when the knowledge volume is too large is solved.

[0087] S231. Generate a user behavior sequence based on the target user's reading start time and reading end time of the knowledge document.

[0088] Based on the target user's start and end times for reading a knowledge document, we can determine the target user's reading duration for the knowledge document. Using the target user's reading duration, start and end times, we can determine a user behavior sequence. This user behavior sequence reflects the target user's position in the sequence for each knowledge document.

[0089] S232. Expand the user behavior sequence based on the target user's feedback information on the recommendation result of the knowledge document.

[0090] The target user's recommendation results for knowledge documents can be divided into positive feedback and negative feedback. Combined with pre-customized evaluation distinction criteria, the value of the knowledge document recommended by the system to the user after the behavior sequence node is quantified. Based on the target user's feedback information on the recommendation results of the knowledge document, the user behavior sequence can be expanded to improve the completeness of the user behavior sequence.

[0091] S233. Construct an original behavior matrix based on the expanded user behavior sequence and the target user's evaluation information on the knowledge document.

[0092] The original behavior matrix may be a three-dimensional matrix, reflecting the value of the knowledge documents involved by the target users to the target users.

[0093] S241. Perform dimensionality reduction processing on the original behavior matrix based on the value of the candidate knowledge document to the target user.

[0094] S242. Determine a target behavior matrix based on the original behavior matrix after dimensionality reduction and the time decay factor.

[0095] The target behavior matrix may be a two-dimensional matrix, which can quantify the user's behavior toward the knowledge document into a value score of the knowledge document to the user.

[0096] Figure 3 This is a schematic diagram of a dimensionality reduction process for the original behavior matrix provided by the second embodiment of the present invention. Figure 3 As shown in the figure, U represents the user set, K represents the knowledge set, S represents the user behavior sequence, and L represents the length of the user behavior sequence. The value of the candidate knowledge document to the target user can be used to reduce the dimensionality of the original behavior matrix, converting the original three-dimensional matrix into a two-dimensional matrix. The difference between the length of the user behavior sequence and the node where a certain behavior occurs is used as the time decay factor to determine the target behavior matrix and quantify the user's behavior towards knowledge.

[0097] S251. Determine similar users of the target user based on the target behavior matrix.

[0098] Specifically, the method of "determining similar users of the target user according to the target behavior matrix" in step S251 may include the following three steps:

[0099] Step a1: Based on the target behavior matrix, determine the cosine value of the knowledge value vector between the target user and other users.

[0100] Optionally, before executing step a1, the directed associations between different knowledge directories may be obtained based on the hierarchical relationship of the knowledge directories and the reading order and reading time of the users for the knowledge in different directories.

[0101] Figure 4 This is a schematic diagram of obtaining the directed association between different knowledge directories provided by the second embodiment of the present invention. Figure 4 As shown, after analyzing the hierarchical relationship of the directories to which the knowledge belongs, the correlation between the two directories in the directory structure tree is determined; then the user's cross-directory reading behavior from directory A to directory B is extracted, and the directory correlation output by the knowledge directory counter is used as the weight, and the one-way correlation from directory A to B is obtained after summing and normalization.

[0102] According to the target behavior matrix and the directed associations between different knowledge catalogs (if any), the cosine value of the knowledge value vector between the target user and other users can be determined.

[0103] Step a2: Determine the similarity between the target user and other users based on the cosine value of the knowledge value vector and the correction factor.

[0104] Step a3: Determine whether the similarity is greater than or equal to a preset threshold. If yes, proceed to step a4; if no, proceed to step a5.

[0105] Step a4: Use other users as similar users to the target user.

[0106] Step a5: Determine other users who are not similar users to the target user.

[0107] S252: Obtain a preference knowledge list of similar users.

[0108] S253. Determine a knowledge recommendation list based on the preferred knowledge list.

[0109] Specifically, the method of "determining a knowledge recommendation list based on the preferred knowledge list" in step S253 may include the following two steps:

[0110] Step b1: Based on catalog relevance, filter knowledge documents from the preferred knowledge list.

[0111] Step b2: Correct the timeliness of the screened knowledge documents and determine the knowledge recommendation list.

[0112] It should also be noted that the knowledge recommendation method provided by the present invention can be used in scenarios where the user has only logged into the system and has not yet browsed knowledge documents, or in scenarios where the user is currently browsing knowledge documents. When used in scenarios where the user has only logged into the system and has not yet browsed knowledge documents, the system makes proactive knowledge recommendations; when used in scenarios where the user is currently browsing knowledge documents, knowledge recommendations can be made based on the knowledge documents the user is currently browsing.

[0113] In this way, the complex professional limitations and cross-limit professional knowledge requirements of users in the commercial banking field can be fully considered. After the target user logs in, the knowledge that may need to be consulted can be recommended to the target user based on the usage behavior of similar users. This solves the problem of inaccurate knowledge recommendation results caused by the inability of a single user's reading behavior to effectively cover the entire amount of knowledge in a knowledge base that stores massive knowledge, and realizes the diversity of knowledge recommendations. At the same time, the introduction of a time decay factor in the process of determining the target behavior matrix can effectively distinguish user behavior in the time dimension, ensuring the timeliness of knowledge recommendations. In addition, the user behavior information obtained by the present invention includes feedback information on the user's recommendation results for knowledge documents, which provides a direct feedback channel for users and provides data support for experience parameter tuning.

[0114] An embodiment of the present invention provides a knowledge recommendation method, comprising: obtaining a knowledge set and behavioral information of a target user, wherein the knowledge set includes multiple knowledge documents; determining candidate knowledge documents and predicting the value of the candidate knowledge documents to the target user, wherein the candidate knowledge documents are knowledge documents that are not involved in the knowledge set by the target user; constructing an original behavior matrix of the target user based on the behavioral information; determining a target behavior matrix of the target user based on the value of the candidate knowledge documents to the target user and the original behavior matrix; determining a knowledge recommendation list based on the target behavior matrix, and recommending the knowledge recommendation list to the target user, wherein the knowledge recommendation list includes at least one knowledge document. By obtaining the knowledge set and the behavioral information of the target user, the value of the knowledge documents that are not involved in the knowledge set by the target user is predicted to the target user, and the original behavior matrix of the target user is constructed, and then the target behavior matrix of the target user is determined based on the value of the candidate knowledge documents to the target user and the original behavior matrix, and then the knowledge recommendation list is determined, and the knowledge recommendation list is recommended to the target user. In this way, the complex professional limitations and cross-limit professional knowledge requirements of users in the commercial banking field can be fully considered. After the target user logs in, the knowledge that may need to be consulted can be recommended to the target user based on the usage behavior of similar users. This solves the problem of inaccurate knowledge recommendation results caused by the inability of a single user's reading behavior to effectively cover the entire amount of knowledge in a knowledge base that stores massive knowledge, and realizes the diversity of knowledge recommendations. At the same time, the introduction of a time decay factor in the process of determining the target behavior matrix can effectively distinguish user behavior in the time dimension, ensuring the timeliness of knowledge recommendations. In addition, the user behavior information obtained by the present invention includes feedback information on the user's recommendation results for knowledge documents, which provides a direct feedback channel for users and provides data support for experience parameter tuning.

[0115] Example 3

[0116] Figure 5 This is a schematic diagram of the structure of a knowledge recommendation device provided by the third embodiment of the present invention. Figure 5 As shown, the device includes: an acquisition module 501, a value prediction module 502, a matrix determination module 503 and a recommendation module 504.

[0117] An acquisition module 501 is used to acquire a knowledge set and behavior information of a target user, wherein the knowledge set includes a plurality of knowledge documents;

[0118] The value prediction module 502 is used to determine candidate knowledge documents and predict the value of the candidate knowledge documents to the target user, wherein the candidate knowledge documents are knowledge documents that are not covered by the target user in the knowledge set;

[0119] The matrix determination module 503 is used to construct an original behavior matrix of the target user based on the behavior information; and determine a target behavior matrix of the target user based on the value of the candidate knowledge document to the target user and the original behavior matrix;

[0120] The recommendation module 504 is configured to determine a knowledge recommendation list according to the target behavior matrix and recommend the knowledge recommendation list to the target user, wherein the knowledge recommendation list includes at least one knowledge document.

[0121] Optionally, each knowledge document has a directory attribute, and knowledge documents with the same directory attribute belong to the same directory;

[0122] The value prediction module 502 is specifically used to determine the contribution of other users to the directory to which the candidate knowledge document belongs, where other users are users who participate in creating, revising or editing the candidate knowledge document; based on the contribution of other users to the directory to which the candidate knowledge document belongs and the value of the candidate knowledge document to other users, predict the value of the candidate knowledge document to the target user.

[0123] Optionally, the behavior information includes the start time and end time of reading the knowledge document, evaluation information, and recommendation result feedback information;

[0124] The matrix determination module 503 is specifically used to generate a user behavior sequence based on the target user's reading start time and reading end time of the knowledge document; expand the user behavior sequence based on the target user's recommendation result feedback information on the knowledge document; and construct the original behavior matrix based on the expanded user behavior sequence and the target user's evaluation information on the knowledge document.

[0125] Optionally, the matrix determination module 503 is specifically configured to perform dimensionality reduction processing on the original behavior matrix according to the value of the candidate knowledge document to the target user; and determine the target behavior matrix according to the reduced dimensionality original behavior matrix and the time decay factor.

[0126] Optionally, the recommendation module 504 is specifically configured to determine similar users of the target user according to the target behavior matrix; obtain a preference knowledge list of the similar users; and determine a knowledge recommendation list according to the preference knowledge list.

[0127] Optionally, the recommendation module 504 is specifically used to determine the cosine value of the knowledge value vector between the target user and other users based on the target behavior matrix; determine the similarity between the target user and other users based on the cosine value of the knowledge value vector and the correction factor; if the similarity is greater than or equal to a preset threshold, then the other users are regarded as similar users.

[0128] Optionally, the recommendation module 504 is specifically configured to filter knowledge documents from the preferred knowledge list based on catalog relevance; perform timeliness correction on the filtered knowledge documents, and determine a knowledge recommendation list.

[0129] The knowledge recommendation device provided in the embodiment of the present invention can execute the knowledge recommendation method provided in any embodiment of the present invention, and has the corresponding functional modules and beneficial effects of the execution method.

[0130] Example 4

[0131] Figure 6 A schematic diagram of the structure of an electronic device 10 that can be used to implement an embodiment of the present invention 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 (such as helmets, glasses, watches, etc.) 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 invention described and / or claimed herein.

[0132] like Figure 6 As shown, the electronic device 10 includes at least one processor 11 and a memory, such as a read-only memory (ROM) 12, a random access memory (RAM) 13, etc., which is communicatively connected to the at least one processor 11. The memory stores a computer program that can be executed by the at least one processor. The processor 11 can perform various appropriate actions and processes according to the computer program stored in the read-only memory (ROM) 12 or the computer program loaded from the storage unit 18 into the random access memory (RAM) 13. Various programs and data required for the operation of the electronic device 10 can also be stored in the RAM 13. The processor 11, ROM 12, and RAM 13 are connected to each other via a bus 14. An input / output (I / O) interface 15 is also connected to the bus 14.

[0133] Multiple components in the electronic device 10 are connected to the I / O interface 15, including an input unit 16, such as a keyboard, a mouse, etc.; an output unit 17, such as various types of displays, speakers, etc.; a storage unit 18, such as a magnetic disk, an optical disk, etc.; and a communication unit 19, such as a network card, a modem, a wireless communication transceiver, etc. The communication unit 19 allows the electronic device 10 to exchange information / data with other devices via a computer network such as the Internet and / or various telecommunication networks.

[0134] The processor 11 can be any general-purpose and / or specialized processing component with processing and computing capabilities. Some examples of the processor 11 include, but are not limited to, a central processing unit (CPU), a graphics processing unit (GPU), various specialized artificial intelligence (AI) computing chips, various processors that run machine learning model algorithms, a digital signal processor (DSP), and any appropriate processor, controller, microcontroller, etc. The processor 11 executes the various methods and processes described above, such as the knowledge recommendation method.

[0135] In some embodiments, the knowledge recommendation method can be implemented as a computer program, which is tangibly contained in a computer-readable storage medium, such as the storage unit 18. In some embodiments, part or all of the computer program can be loaded and / or installed on the electronic device 10 via the ROM 12 and / or the communication unit 19. When the computer program is loaded into the RAM 13 and executed by the processor 11, one or more steps of the knowledge recommendation method described above can be performed. Alternatively, in other embodiments, the processor 11 can be configured to perform the knowledge recommendation method in any other appropriate manner (for example, by means of firmware).

[0136] Various embodiments of the systems and techniques described 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), system-on-chip systems (SOCs), programmable logic devices (CPLDs), computer hardware, firmware, software, and / or combinations thereof. These various embodiments can include being implemented in one or more computer programs that are executable and / or interpreted on a programmable system that includes 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.

[0137] Computer programs for implementing the methods of the present invention may be written in any combination of one or more programming languages. These computer programs may be provided to a processor of a general-purpose computer, a special-purpose computer, or other programmable data processing device, such that when the computer program is executed by the processor, the functions / operations specified in the flowcharts and / or block diagrams are implemented. The computer program may be executed entirely on the machine, partially on the machine, as a stand-alone software package, partially on the machine and partially on a remote machine, or entirely on a remote machine or server.

[0138] In the context of the present invention, computer-readable storage media can be tangible media that can contain or store a computer program for use with an instruction execution system, device or equipment or used in combination with an instruction execution system, device or equipment. Computer-readable storage media can include but are not limited to electronic, magnetic, optical, electromagnetic, infrared or semiconductor systems, devices or equipment, or any suitable combination of the foregoing. Alternatively, computer-readable storage media can be machine-readable signal media. More specific examples of machine-readable storage media can include electrical connections based on one or more lines, portable computer disks, hard disks, random access memories (RAM), read-only memories (ROM), erasable programmable read-only memories (EPROM or flash memory), optical fibers, portable compact disk read-only memories (CD-ROM), optical storage devices, magnetic storage devices, or any suitable combination of the foregoing.

[0139] To provide interaction with a user, the systems and techniques described herein can be implemented on an electronic device 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 electronic device. 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).

[0140] The systems and techniques described herein can 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 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 can 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), a blockchain network, and the Internet.

[0141] A computing system may include clients and servers. The clients and servers are typically remote from each other and typically interact via a communication network. This client-server relationship arises through computer programs running on the respective computers, creating a client-server relationship. The server may be a cloud server, also known as a cloud computing server or cloud host. This server is a hosting product within the cloud computing service ecosystem that addresses the management difficulties and limited scalability of traditional physical hosting and VPS services.

[0142] It should be understood that the various forms of the processes shown above can be used to reorder, add, or delete steps. For example, the steps described in the present invention can be performed in parallel, sequentially, or in a different order, as long as the desired results of the technical solution of the present invention can be achieved. This is not limited herein.

[0143] The above specific embodiments do not limit the scope of protection of the present invention. Those skilled in the art will appreciate that various modifications, combinations, sub-combinations, and substitutions may be made based on design requirements and other factors. Any modifications, equivalent substitutions, and improvements made within the spirit and principles of the present invention are intended to be included within the scope of protection of the present invention.

Claims

1. A knowledge recommendation method, characterized in that: include: Acquire a knowledge set and behavior information of a target user, wherein the knowledge set includes a plurality of knowledge documents; Determining candidate knowledge documents and predicting the value of the candidate knowledge documents to the target user, wherein the candidate knowledge documents are knowledge documents in the knowledge set that are not covered by the target user; Constructing an original behavior matrix of the target user based on the behavior information; determining a target behavior matrix for the target user according to the value of the candidate knowledge document to the target user and the original behavior matrix; A knowledge recommendation list is determined according to the target behavior matrix, and the knowledge recommendation list is recommended to the target user, wherein the knowledge recommendation list includes at least one knowledge document.

2. The method according to claim 1, characterized in that Each of the knowledge documents has a directory attribute, and the knowledge documents with the same directory attribute belong to the same directory; The predicting the value of the candidate knowledge document to the target user includes: Determining the contribution of other users to the directory to which the candidate knowledge document belongs, wherein the other users are users who participated in creating, revising, or editing the candidate knowledge document; The value of the candidate knowledge document to the target user is predicted based on the contribution of the other users to the directory to which the candidate knowledge document belongs and the value of the candidate knowledge document to the other users.

3. The method according to claim 1 or 2, characterized in that The behavior information includes the start time of reading the knowledge document, the end time of reading, evaluation information and recommendation result feedback information; The constructing the original behavior matrix of the target user according to the behavior information includes: Generate a user behavior sequence according to the target user's reading start time and reading end time of the knowledge document; Expanding the user behavior sequence according to the target user's feedback information on the recommendation result of the knowledge document; The original behavior matrix is ​​constructed according to the expanded user behavior sequence and the target user's evaluation information on the knowledge document.

4. The method according to claim 3, characterized in that Determining the target behavior matrix of the target user according to the value of the candidate knowledge document to the target user and the original behavior matrix includes: Performing dimensionality reduction processing on the original behavior matrix according to the value of the candidate knowledge document to the target user; The target behavior matrix is ​​determined according to the original behavior matrix after dimensionality reduction and the time decay factor.

5. The method according to claim 1, wherein Determining a knowledge recommendation list according to the target behavior matrix includes: Determining similar users to the target user based on the target behavior matrix; Obtaining a preference knowledge list of the similar users; A knowledge recommendation list is determined based on the preferred knowledge list.

6. The method according to claim 5, characterized in that Determining similar users of the target user according to the target behavior matrix includes: Determining the cosine value of the knowledge value vector between the target user and other users according to the target behavior matrix; Determining the similarity between the target user and the other users based on the cosine value of the knowledge value vector and the correction factor; If the similarity is greater than or equal to a preset threshold, the other user is regarded as the similar user.

7. The method according to claim 5, characterized in that The step of determining a knowledge recommendation list based on the preferred knowledge list includes: Filtering knowledge documents from the preferred knowledge list based on catalog relevance; The filtered knowledge documents are corrected for timeliness to determine the knowledge recommendation list.

8. A knowledge recommendation device, characterized in that: include: Acquisition module, value prediction module, matrix determination module and recommendation module; The acquisition module is used to acquire a knowledge set and behavior information of a target user, wherein the knowledge set includes a plurality of knowledge documents; The value prediction module is used to determine candidate knowledge documents and predict the value of the candidate knowledge documents to the target user, wherein the candidate knowledge documents are knowledge documents that are not involved in the target user in the knowledge set; The matrix determination module is configured to construct an original behavior matrix of the target user based on the behavior information; and determine a target behavior matrix of the target user based on the value of the candidate knowledge document to the target user and the original behavior matrix; The recommendation module is used to determine a knowledge recommendation list according to the target behavior matrix, and recommend the knowledge recommendation list to the target user, wherein the knowledge recommendation list includes at least one knowledge document.

9. An electronic device, characterized in that: The electronic device comprises: at least one processor; and a memory communicatively connected to the at least one processor; wherein, The memory stores a computer program that can be executed by the at least one processor. The computer program is executed by the at least one processor to enable the at least one processor to perform the knowledge recommendation method according to any one of claims 1 to 7.

10. A computer-readable storage medium, characterized in that The computer-readable storage medium stores computer instructions, and the computer instructions are used to enable a processor to implement the knowledge recommendation method according to any one of claims 1 to 7 when executed.

Images