Method and device for evaluating importance of article, electronic equipment and storage medium

Abstract

Embodiments of the present application provide an article importance evaluation method and device, electronic equipment and storage medium, and relate to the fields of big data, cloud technology, medicine and retrieval technology. The method comprises: obtaining a directed graph, each node in the graph representing an article to be evaluated or a reference article, and the direction of the directed edge being from the node corresponding to the cited article to the node corresponding to the citing article; for any target article in the set of articles to be evaluated, determining each citation path of the target article corresponding to each reference article according to the graph, the source node of the citation path representing the reference article and the destination node representing the target article; determining the importance of each citation path according to the citation relationship of each node on the citation path except the destination node; and determining the importance of the target article according to the importance of each citation path. Based on the method, more objective and comprehensive evaluation of the article can be achieved, and when the method is applied to retrieval, the retrieval result can be effectively improved.

Description

Technical Field

[0001] This application relates to the fields of big data, cloud technology, medicine, and retrieval technology. Specifically, this application relates to a method, apparatus, electronic device, and storage medium for assessing the importance of an article. Background Technology

[0002] With the rapid development of information technology, people have more and more ways to acquire data, and the total amount of data is also growing explosively. The emergence of massive amounts of data in various fields has led to a continuous increase in the demand for data retrieval optimization. How to help users find the articles they want faster and better is one of the important issues that relevant technical personnel have been researching.

[0003] For example, in scientific research fields (such as life sciences and biomedicine), with the increasing emphasis placed on scientific research by various countries, the number of scientific papers is growing rapidly every year. Researchers, professionals in related fields, and even the general public frequently have a need to search for these papers. To meet this need, various search / retrieval systems have emerged in the current technology. Since search systems typically retrieve papers with high importance (high ranking scores) from databases based on search keywords, the reasonableness and accuracy of the assessment of paper importance will have a significant impact on the search results.

[0004] Currently, the importance of an article is mostly assessed based on relevant information about the article itself, such as the number of times the article has been cited, the author of the article, the keywords in the article, the number of downloads, etc. While these assessment methods can meet basic needs to a certain extent, it has been found in practical applications that the existing assessment results of article importance are not objective enough, and the assessment methods still need to be improved. Summary of the Invention

[0005] The purpose of this application is to provide a method, apparatus, electronic device, and storage medium for evaluating the importance of articles that can better meet the needs of practical applications. To achieve the above objective, the technical solutions provided by this application are as follows:

[0006] On the one hand, embodiments of this application provide a method for evaluating the importance of an article, the method comprising:

[0007] Obtain the directed graph corresponding to the set of articles to be evaluated, wherein the directed graph is constructed based on the citation relationship between the articles in the set of articles to be evaluated and the reference set of articles. Each node in the directed graph represents an article to be evaluated in the set of articles to be evaluated or a reference article in the reference set of articles. The direction of the directed edges in the directed graph is from the node corresponding to the citation article to the node corresponding to the cited article.

[0008] For any target article in the set of articles to be evaluated, the citation paths of the target article corresponding to each reference article are determined according to the directed graph. The source node of each citation path is a node corresponding to a reference article, and the destination node is a node corresponding to the target article.

[0009] For each of the aforementioned reference paths, the importance of the reference path is determined based on the reference relationships between the nodes on the reference path other than the destination node.

[0010] The importance of the target article is determined based on the importance of each of the cited paths.

[0011] On the other hand, embodiments of this application provide an apparatus for evaluating the importance of an article, the apparatus comprising:

[0012] The data acquisition module is used to acquire the directed graph corresponding to the set of articles to be evaluated. The directed graph is constructed based on the citation relationship between the articles in the set of articles to be evaluated and the reference set. Each node in the directed graph represents an article to be evaluated in the set of articles to be evaluated or a reference article in the reference set. The direction of the directed edges in the directed graph is from the node corresponding to the citation article to the node corresponding to the cited article.

[0013] The citation path determination module is used to determine, for any target article in the set of articles to be evaluated, each citation path corresponding to each reference article of the target article according to the directed graph, wherein the source node of each citation path is a node corresponding to a reference article, and the destination node is a node corresponding to the target article.

[0014] The importance determination module is used to determine the importance of each reference path based on the reference relationships between the nodes on the reference path other than the target node; and to determine the importance of the target article based on the importance of each reference path.

[0015] Optionally, when determining the importance of the target article based on the importance of each of the aforementioned citation paths, the importance determination module can be used to:

[0016] For each source node appearing in each of the aforementioned reference paths, the importance of the target article corresponding to the source node is determined based on the importance of each path containing that source node in each of the aforementioned reference paths; the importance of the target article is determined based on the importance of the target article corresponding to each of the aforementioned source nodes.

[0017] Optionally, for each source node, the importance determination module, when determining the importance of the target article corresponding to the reference article corresponding to that source node, can be used to:

[0018] The sum of the importance of each path that contains the source node in each of the aforementioned reference paths is taken as the importance of the target article corresponding to the reference article corresponding to the source node.

[0019] When determining the importance of a target article based on the importance of each of the source nodes, the importance determination module can be used for:

[0020] The importance of the target article is determined by ranking the importance of each source node from highest to lowest, where k ≥ 1.

[0021] Optionally, for each of the aforementioned reference paths, when determining the importance of the reference path based on the reference relationships between the nodes along the reference path other than the destination node, the importance determination module may be used to:

[0022] For a source node on the citation path, the importance of the source node on the citation path is determined based on the relevant information of the articles that have a citation relationship with the reference article corresponding to the source node.

[0023] For each node on the citation path other than the source node and the destination node, the importance of the node on the citation path is determined based on the relevant information of the articles that have a citation relationship with the corresponding article and the importance of the nodes that precede the node on the citation path.

[0024] The importance of the reference path is determined based on the importance of each node on the reference path other than the destination node.

[0025] Optionally, when determining the importance of a node in a citation path based on information about articles that have a citation relationship with the corresponding article and the importance of the nodes preceding that node in the citation path, the importance determination module can be used for:

[0026] Based on the relevant information of the articles that have a citation relationship with the corresponding article, determine the first importance of the node; based on the importance of each node preceding the node on the citation path, determine the second importance of the node on the citation path; based on the first and second importance of the node, obtain the total importance of the node on the citation path.

[0027] Optionally, for each reference path, the importance determination module may, when determining the importance of the reference path, use the sum of the importance of each node on the reference path other than the destination node as the importance of the reference path.

[0028] Optionally, for any node, the relevant information of the articles that have a citation relationship with the article corresponding to that node includes at least one of the following:

[0029] The first number of times the first article cites the second article, where the first article is the article corresponding to this node, and the second article is the article cited by the first article;

[0030] The second number of third articles citing the first article;

[0031] The first article's citation frequency or number of times;

[0032] The position of the first article in each of the third articles.

[0033] Optionally, the relevant information of the articles that have a citation relationship with the article corresponding to the node includes the first number corresponding to the node; when the importance determination module determines the first importance of the node based on the relevant information of the articles that have a citation relationship with the article corresponding to the node, it can be used to:

[0034] The primary importance of a node is determined based on the primary quantity corresponding to that node.

[0035] Wherein, the first importance of a node is negatively correlated with the first quantity corresponding to that node; or,

[0036] If the first quantity corresponding to the node is less than or equal to the preset quantity, the first importance of the node is positively correlated with the first quantity corresponding to the node; if the first quantity corresponding to the node is greater than the preset quantity, the first importance of the node is negatively correlated with the first quantity corresponding to the node.

[0037] Optionally, the importance of the target article is determined by the following expression:

[0038] R(p)=e -D(p)

[0039]

[0040] D(p) = Min k (D1(p),D2(p),…D i (p))

[0041] Among them, D i (C i ) = 0

[0042]

[0043] Where p represents the target article, R(p) is the importance of the target article, and i represents the reference article C. i D i(p) represents p corresponding to C i The importance of D i The larger (p) is, the less important the representation; Min k (D1(p),D2(p),…D i (p) represents D1(p), D2(p), ..., D i The kth minimum value in (p), k≥1, q represents the minimum value in C. i The corresponding node is the article corresponding to any node on each of the reference paths of the source node, excluding the destination node. This indicates that the node corresponding to q is in the range of C. i The corresponding node represents the importance of a reference path to the source node, q. out This represents the number of articles cited by q, and d is a preset value.

[0044] Optionally, the importance determination module can also be used to: update the directed graph based on the citation relationship corresponding to the new article when a new article is obtained, wherein the new article is a new article to be evaluated or a new reference article; and redetermine the importance of the articles to be evaluated in the set of articles to be evaluated based on the updated directed graph.

[0045] On the other hand, embodiments of this application also provide an article retrieval method, which includes:

[0046] Obtain a search request; based on the search terms in the search request, determine at least one candidate article from the search database that matches the search request;

[0047] Based on the importance of each candidate article, the target article corresponding to the search request is determined from each candidate article, wherein the importance of each article in the search database is determined using the article importance evaluation method provided in the embodiments of this application;

[0048] The target article is sent as a search result to the requesting end of the search request.

[0049] On the other hand, embodiments of this application also provide an article retrieval device, which includes:

[0050] The retrieval request acquisition module is used to acquire retrieval requests;

[0051] The retrieval module is used to determine at least one candidate article matching the retrieval request from the retrieval database based on the retrieval terms in the retrieval request, and to determine the target article corresponding to the retrieval request from the retrieval request based on the importance of each candidate article. The importance of each article in the retrieval database is determined using the article importance assessment method provided in the embodiments of this application.

[0052] The search result feedback module is used to send the target article as a search result to the requesting end of the search request.

[0053] On the other hand, embodiments of this application also provide an electronic device, which includes a memory and a processor. The memory stores a computer program, and the processor executes the computer program to implement the method provided in any optional embodiment of this application.

[0054] On the other hand, embodiments of this application also provide a computer-readable storage medium storing a computer program that, when executed by a processor, implements the method provided in any optional embodiment of this application.

[0055] On the other hand, embodiments of this application also provide a computer program product, which includes a computer program that, when executed by a processor, implements the methods provided in any optional embodiment of this application.

[0056] The beneficial effects of the technical solution provided in this application are as follows:

[0057] The article importance assessment method provided in this application does not merely consider the article's own citation information (such as the total number of citations) when determining the importance of the article to be assessed. Instead, it uses multiple reference articles in a reference dataset as references, constructs a directed graph of citation relationships based on the citation relationships between these reference articles and the article to be assessed, and then uses this directed graph, starting from the nodes corresponding to each reference article, to find each citation path of the article to be assessed. By evaluating the importance of each citation path, the importance of the article to be assessed is realized. Since the citation paths all originate from the reference articles and can directly or indirectly link to the article to be assessed, using the importance of each citation path determined based on the citation relationships corresponding to each node on the citation path to assess the importance of the article to be assessed makes the assessment results more objective and comprehensive, and can better meet the needs of practical applications. When this assessment method is applied to article retrieval scenarios, it can effectively improve retrieval results. Attached Figure Description

[0058] To more clearly illustrate the technical solutions in the embodiments of this application, the accompanying drawings used in the description of the embodiments of this application will be briefly introduced below.

[0059] Figure 1 This is a flowchart illustrating a method for assessing the importance of an article as provided in an embodiment of this application.

[0060] Figure 2 This is a schematic diagram of a directed graph provided in one example of this application;

[0061] Figure 3 This is a flowchart illustrating an article retrieval method provided in an embodiment of this application;

[0062] Figure 4 This is a schematic diagram of the structure of a retrieval system provided in an embodiment of this application;

[0063] Figure 5 This is a schematic diagram of the overall architecture of a scheme for determining the importance of an article provided in the embodiments of this application;

[0064] Figure 6 This is a schematic diagram illustrating the implementation process of a method for determining article ranking scores provided in this application embodiment;

[0065] Figure 7 This is a schematic diagram of the structure of an article importance assessment device provided in the embodiments of this application;

[0066] Figure 8 This is a schematic diagram of the structure of a retrieval device provided in an embodiment of this application;

[0067] Figure 9 This is a schematic diagram of the structure of an electronic device to which this application applies. Detailed Implementation

[0068] The embodiments of this application are described below with reference to the accompanying drawings. It should be understood that the embodiments described below with reference to the accompanying drawings are exemplary descriptions for explaining the technical solutions of the embodiments of this application, and do not constitute a limitation on the technical solutions of the embodiments of this application.

[0069] Those skilled in the art will understand that, unless specifically stated otherwise, the singular forms “a,” “an,” “the,” and “the” used herein may also include the plural forms. It should be further understood that the terms “comprising” and “including” as used in embodiments of this application mean that the corresponding feature can be implemented as the presented feature, information, data, step, operation, element, and / or component, but do not exclude implementation as other features, information, data, step, operation, element, component, and / or combinations thereof supported by the art. It should be understood that when we say that an element is “connected” or “coupled” to another element, the one element can be directly connected or coupled to the other element, or it can mean that the one element and the other element establish a connection relationship through an intermediate element. Furthermore, “connected” or “coupled” as used herein can include wireless connection or wireless coupling. The term “and / or” as used herein indicates at least one of the items defined by the term; for example, “A and / or B” can be implemented as “A,” or as “B,” or as “A and B.” When describing multiple (two or more) items, if the relationship between the multiple items is not explicitly defined, the multiple items can refer to one, several or all of the multiple items. For example, the description of "parameter A includes A1, A2, A3" can be implemented as parameter A includes A1 or A2 or A3, or it can be implemented as parameter A includes at least two of the three items A1, A2 and A3.

[0070] This application provides a method for evaluating the importance of an article and a method for retrieving an article. Both the evaluation and retrieval methods can be executed by any computer device. The devices executing the evaluation and retrieval methods can be the same or different. Optionally, either the evaluation or retrieval method can be executed by a terminal device or a server; for example, both the evaluation and retrieval methods can be executed by a retrieval server.

[0071] The server can be a standalone physical server, a server cluster or distributed system consisting of multiple physical servers, or a cloud server providing cloud computing services. The terminal (also called a user terminal or terminal device) can be a smartphone, tablet, laptop, desktop computer, smart voice interaction device (e.g., smart speaker), wearable electronic device (e.g., smartwatch), in-vehicle terminal, smart home appliance (e.g., smart TV), AR / VR device, etc., but is not limited thereto. The terminal and server can be directly or indirectly connected via wired or wireless communication, and this application does not impose any restrictions on this connection.

[0072] Optionally, the data processing involved in the methods provided in this application embodiment can be implemented based on cloud technology. For example, the calculation operations involved in the method for evaluating the importance of articles can be implemented using cloud computing technology, and the storage of the data involved (such as the storage of the reference article set and the article set to be evaluated, as well as the storage of the determined article importance, ranking scores, etc.) can be carried out using cloud storage.

[0073] Cloud computing is a computing model that distributes computing tasks across a resource pool composed of a large number of computers, enabling various application systems to obtain computing power, storage space, and information services as needed. The network providing these resources is called the "cloud." From the user's perspective, resources in the "cloud" are infinitely scalable, readily available, on-demand, expandable, and pay-as-you-go. Cloud storage is a new concept that extends and develops from cloud computing. A distributed cloud storage system (hereinafter referred to as a storage system) refers to a storage system that uses cluster applications, grid technology, and distributed storage file systems to aggregate a large number of various types of storage devices (also called storage nodes) in a network through application software or application interfaces to work collaboratively and provide data storage and business access functions.

[0074] The solution provided in this application can be applied to the processing of big data. For example, the collection of articles to be evaluated may include papers with a large amount of data. Based on the solution provided in this application, an objective assessment of the importance of a large number of papers can be achieved, providing a foundation for the processing or application of these papers.

[0075] Big data refers to data sets that cannot be captured, managed, and processed within a certain timeframe using conventional software tools. It represents massive, rapidly growing, and diverse information assets that require new processing models to achieve stronger decision-making, insightful discovery, and process optimization capabilities. With the advent of the cloud era, big data has attracted increasing attention. Big data requires specialized technologies to effectively process large amounts of data within a tolerable timeframe. Technologies suitable for big data include massively parallel processing databases, data mining, distributed file systems, distributed databases, cloud computing platforms, the internet, and scalable storage systems.

[0076] To better understand the methods provided in the embodiments of this application, some related technologies or technical terms involved in the embodiments of this application will be introduced and explained below.

[0077] Ranking score: The ranking score given to each article.

[0078] Shortest distance: The shortest distance refers to the distance between each article and any core article.

[0079] Citation-linked graph-based ranking system: A scoring system that uses a graph (directed graph) constructed based on the relationship between citations and cited articles to calculate the importance ranking of articles.

[0080] The technical solutions of this application and their effects are described below through several exemplary embodiments. It should be noted that the following embodiments can be referenced, borrowed from, or combined with each other. Identical terms, similar features, and similar implementation steps in different embodiments will not be repeated.

[0081] Figure 1 This illustration shows a flowchart of a method for evaluating the importance of an article according to an embodiment of this application. Optionally, this method can be executed by a server, such as... Figure 1 As shown, the method may include steps S110 to S140.

[0082] Step S110: Obtain the directed graph corresponding to the set of articles to be evaluated. The directed graph is constructed based on the citation relationship between the articles in the set of articles to be evaluated and the reference set. Each node in the directed graph represents an article to be evaluated in the set of articles to be evaluated or a reference article in the reference set. The direction of the directed edges in the directed graph is from the node corresponding to the citation article to the node corresponding to the cited article.

[0083] The set of articles to be evaluated includes multiple articles, i.e., articles whose importance needs to be assessed. The set of reference articles, also known as the core set or seed set, includes multiple reference / core articles. In this embodiment, reference articles are used as benchmarks for assessing the importance of the articles to be evaluated. This embodiment does not limit the specific method of obtaining reference articles; optionally, authoritative and valuable articles in the technical field to which the articles in the set of articles to be evaluated belong can be obtained as reference articles.

[0084] The solution provided in this application can be applied to the importance assessment of relevant articles (literature, papers, etc.) in any field. Optionally, the article to be assessed can be a paper or journal article in a related scientific field such as life sciences, biomedicine, and healthcare. Reference articles can be selected from high-value and influential papers or journal articles in these related fields. For example, articles from authoritative national or international biomedical literature databases can be selected, such as the MEDLINE database or other authoritative databases in this field. The MEDLINE database is one of the most trusted databases internationally, broadly covering reference articles in the biomedical and health fields, especially biomedicine. Using articles from authoritative databases as reference articles can effectively ensure the quality and quantity of reference articles. Of course, reference articles can also be manually selected by experts in the field, or multiple experts can formulate screening rules for reference articles, according to which relevant personnel or electronic devices can screen from a large number of candidate articles.

[0085] The article importance assessment method provided in this application is based on a reference set of articles and the citation relationships between articles. To facilitate a convenient, quick, and visual understanding of the citation relationships between articles, after obtaining the reference set and the set of articles to be evaluated, a directed graph can be constructed based on the citation relationships (article citation relationships) of each article in these two sets. The directed graph includes multiple nodes and directed edges between nodes. In this application embodiment, each article in the reference set and the set of articles to be evaluated corresponds to a node in the directed graph. Which nodes are connected by edges and the direction of the edges are determined based on the citation relationships between the nodes. For example, if node a and node b represent article A and article B respectively, and article A cites article B, then the directed graph includes a directed edge from node a to node b.

[0086] Optionally, the directed graph described above can be constructed in the following way:

[0087] Each article in the set of articles to be evaluated and the set of reference articles is assigned as a node;

[0088] For each reference article in the reference article set, based on the citation relationship of the reference article, a directed edge is created between the node corresponding to the reference article and the node corresponding to the article cited by the reference article. The direction of the directed edge is from the reference article to the article it cites.

[0089] Based on the citation relationships of each article to be evaluated, directed edges are created between the nodes corresponding to each article to be evaluated. These directed edges also point from the node corresponding to the cited article to the node corresponding to the cited article.

[0090] In this alternative approach, even if the article to be evaluated cites a reference article, nodes pointing from the node corresponding to the article to the reference article do not need to be created in the directed graph. This is because the article to be evaluated is the one whose importance needs to be determined, while the reference article is merely a reference, and its importance does not need to be evaluated. Therefore, the information about the reference article being cited can be disregarded, reducing the space occupied by the directed graph's data storage and the amount of data that needs to be processed subsequently. Of course, if the article to be evaluated cites a reference article, directed edges can be created between the corresponding nodes in the directed graph without affecting the subsequent evaluation results.

[0091] When constructing a directed graph, the mutual reference relationships between reference articles can be considered or not. That is, if a reference article C1 references a reference article C2, there can be a directed edge in the directed graph from the node corresponding to article C1 to the node corresponding to article C2, or there can be no such edge.

[0092] As an example, Figure 2 The diagram illustrates a partial structure of a directed graph, showing the directed graph structure corresponding to 4 reference articles and 6 articles to be evaluated. Nodes C1, C2, C3, and C4 represent reference articles C1 to C4, respectively, while nodes p and q1 to q5 represent articles p and q1 to q5, respectively, to be evaluated. Figure 2 As can be seen from the diagram, reference article C2 cites articles q3, q4, and q5 to be evaluated; article q4 cites article q1 to be evaluated; and article q1 cites article p to be evaluated. The citation relationships between the articles are clearly visible through the directed graph. This diagram also considers the citation relationships between reference articles; reference article C1 cites reference article C2. If the citation relationships between reference articles were not considered... Figure 2 The text may not contain directed edges between nodes corresponding to reference articles.

[0093] Step S120: For any target article in the set of articles to be evaluated, determine the citation paths of the target article corresponding to each reference article according to the directed graph. The source node of each citation path is the node corresponding to a reference article, and the destination node is the node corresponding to the target article.

[0094] Step S130: For each reference path, determine the importance of the reference path based on the reference relationships of each node on the reference path other than the destination node.

[0095] Step S140: Determine the importance of the target article based on the importance of each citation path.

[0096] The target article can be any article in the set of articles to be evaluated. Based on the method provided in this application, the importance of any article in the set of articles to be evaluated can be assessed. Of course, in practical applications, one, some, or all articles in the set of articles to be evaluated can also be specified for evaluation according to application requirements. For any target article to be evaluated, the above-mentioned citation path refers to a directed path in a directed graph with the node corresponding to the reference article as the source node (starting point) and the node corresponding to the target article as the destination node. Each citation path contains at least two nodes, that is, the citation path may only include the source node and the destination node, or it may also include intermediate nodes between the source node and the destination node.

[0097] Still with Figure 2 Taking the directed graph structure shown as an example, assuming the target article is article p, according to... Figure 2 As can be seen from the directed graph, for reference article C2, there are three citation paths to article p: C2→q4→q1→p, C2→q3→p, and C2→q5→q2→p. The source node in each of these three citation paths is the node C2 corresponding to reference article C2, and the destination node is the node p of the target article. All three paths contain intermediate nodes; for example, the intermediate nodes of the path C2→q4→q1→p include nodes q4 and q1.

[0098] It is understandable that if the nodes corresponding to the referenced articles are also connected by directed edges, some reference paths may contain nodes corresponding to multiple referenced articles. However, the source node mentioned above in this application refers to the starting point of the path, such as... Figure 2 The reference path shown is a link from node C1 to node p. This path includes nodes corresponding to the two reference articles C1 and C2. The source node of this path is C1.

[0099] Depend on Figure 2 As can be seen, the citation path of the target article represents the path that can be directly or indirectly linked to the target article from the reference articles. Therefore, the citation path can also be called a link chain. Since the citation chain of the target article ultimately links to the target article from the reference articles, and the reference articles are reference points, such as core articles in the same or related fields as the target article, the importance of the target article can be objectively assessed based on these citation chains.

[0100] Since the nodes (i.e., articles) and their corresponding citation relationships are different on different citation paths, the importance of different paths (which can also be understood as the value of the path, i.e., its contribution to the importance of the target article) is also different. Therefore, after determining each citation path of the target article based on the directed graph, the importance of each citation path can be evaluated based on the citation relationships of each node on each citation path other than the target node. Then, the importance of the target article can be determined based on the importance of each citation path of the target article.

[0101] In this context, nodes along a referencing path, excluding the target node, include the source node of the referencing path and intermediate nodes between the source and target nodes. For ease of description, the nodes along a referencing path, excluding the target node, are referred to as path nodes in the following description. In this embodiment, the referencing relationship corresponding to a node may include at least one of the following: information about the articles referenced by the node (i.e., the articles corresponding to the node) or information about other articles referencing the node. The specific content of the article's relevant information is not uniquely limited in this embodiment; theoretically, any information related to the article corresponding to the node and relating to its value or influence is acceptable. Optionally, this relevant information may include, but is not limited to, the number of articles, the reference location, and the reference frequency.

[0102] For a target article, its citation paths, which link to other articles (including references), reflect how the article has been directly or indirectly cited. The citation relationships between each node on the citation path reflect the value of that node. Therefore, the importance of a citation path can be determined based on the citation relationships between its nodes. After determining the importance of each citation path, the overall importance of the target article can be determined by synthesizing the importance of all citation paths.

[0103] In this application, the specific method for determining the importance of a reference path based on the importance of the reference relationships corresponding to each path node on that path is not limited to a single embodiment. Optionally, for each path node on a reference path, the initial importance of the path node (i.e., the article corresponding to the node) can be determined based on the reference relationships corresponding to the path node and using existing methods for evaluating article importance. After determining the initial importance of each path node on the path, the importance of the reference path can be obtained by fusing the initial importance of each path node. For example, the sum or average of the initial importance of each path node on the reference path can be used as the importance of the reference path. Different weights can also be assigned to different nodes, and the initial importance of each path node can be weighted and summed according to the weight of each path node. The result of the weighted summation can be used as the importance of the reference path. For example, the source nodes (i.e., the nodes corresponding to the reference articles) in each path node can be assigned a larger weight than non-source nodes.

[0104] As an alternative approach, for each reference path, the importance of the reference path is determined based on the reference relationships between the nodes along the path, excluding the destination node. This can include:

[0105] For a source node on the citation path, the importance of the source node on the citation path is determined based on the relevant information of the articles that have a citation relationship with the reference article corresponding to the source node.

[0106] For each node on the citation path other than the source node and the destination node, the importance of the node on the citation path is determined based on the relevant information of the articles that have a citation relationship with the corresponding article and the importance of the nodes that precede the node on the citation path.

[0107] The importance of a reference path is determined based on the importance of each node on the reference path, excluding the destination node.

[0108] In this scheme, for each reference path, the importance of each path node on the reference path is determined iteratively, starting from the source node and proceeding along the path direction. The source node, being the starting point of the reference path, can have its importance determined directly based on information from other articles that reference it. For each path node other than the source node, since one or more other path nodes precede it on the reference path and these other path nodes directly or indirectly reference it, the importance of each path node on the reference path is considered not only in terms of information from other articles that reference it, but also in terms of the importance of the path nodes preceding it, in order to more objectively evaluate the importance of each node on the reference path.

[0109] Optionally, for each path node on a citation path other than the source node, the importance of that node on the citation path is determined based on information about articles that have a citation relationship with the corresponding article, and the importance of the nodes preceding that node on the citation path, including:

[0110] Determine the primary importance of the node based on information about articles that have a citation relationship with the corresponding article;

[0111] The second importance of a node is determined based on the importance of the nodes preceding it on the reference path.

[0112] The importance of a node on the reference path is determined by its first and second importance.

[0113] In other words, we can first determine the importance of a node in one dimension based on its own relevant reference information (i.e., the relevant information that has a reference relationship with the node mentioned above), and then use the importance of each path node before the node to evaluate the importance of the node in another dimension. By combining the importance of the two dimensions, we can obtain the importance of the node in the current reference path.

[0114] For a given node, articles that have a citation relationship with the article corresponding to that node can include at least one of the following: articles cited by the article corresponding to that node, or articles that cite that node (i.e., the article corresponding to that node). The information related to articles that have a citation relationship with the article corresponding to that node can be one or more pieces of information, theoretically including various information that can be used to assess the value or importance of an article, such as the number of articles cited by the article corresponding to that node, the author of the article corresponding to that node, the number of relevant keywords in the article corresponding to that node, etc. Optionally, for a given node, the information related to articles that have a citation relationship with the article corresponding to that node includes at least one of the following:

[0115] The first number of times the first article cites the second article, where the first article is the article corresponding to this node, and the second article is the article cited by the first article;

[0116] The second number of third articles citing the first article;

[0117] The first article's citation frequency or number of times;

[0118] The position of the first article in each of the third articles.

[0119] For ease of description, for a given node, the article corresponding to that node is called the first article, the article cited by the first article is called the second article, and the article that cited the first article is called the third article. For example, if a node corresponds to article a, article a cites article b, and article c cites article a, then article a is the first article, article b is the second article, and article c is the third article. It's understandable that in practical applications, an article may cite multiple articles, or be cited by multiple other articles.

[0120] In this context, for a second article, the number of times the first article is cited can simply be the number of times it is cited by all articles in the reference set and the set of articles to be evaluated. The citation frequency of the first article can be the ratio of its citation count to the sum of the citation counts of all the aforementioned articles. Since the number of citations and citation frequency of the first article can reflect its importance to other articles, the citation frequency or number of citations can be used as a factor in assessing its importance; for example, a higher citation frequency or a greater number of citations may indicate greater relative importance.

[0121] Furthermore, the position of cited articles within a single article often reflects their importance or reference value. For instance, a cited article appearing in the abstract is generally more valuable than one in the background section. Therefore, the position of the first article in other articles can be considered a factor in assessing its importance. Optionally, a pre-configured correspondence between different citation positions and the primary importance of cited articles can be established. This correspondence can then determine the impact of the first article's position in various third articles on its primary importance.

[0122] Regarding the specific method for determining the first importance of a node based on information related to articles that have a citation relationship with the corresponding article, this application embodiment does not impose a unique limitation. Optionally, existing methods for determining article importance based on citation relationships can be used to determine the aforementioned first importance. Optionally, for the second importance corresponding to a path node, the sum or average of the importance of all nodes preceding the node on the citation path can be used to determine the second importance of the node. Alternatively, different weights can be assigned to each node preceding the node based on the distance between the node and the nodes preceding it on the citation path. For example, nodes closer to the node on the citation path have greater weights. Nodes directly connected to the node via directed edges on the path have the highest relative weight. The second importance of the node can be obtained by weighting and summing the importance of each node preceding it on the citation path based on the weights of the nodes preceding it.

[0123] After obtaining the first and second importance of a node, the sum of the first and second importance can be used as the node's importance on the reference path. Alternatively, the weights corresponding to the first and second importance can be obtained, and the first and second importance can be weighted and summed to obtain the node's importance on the reference path. For example, the weight of the first importance can be greater than the weight of the second importance, and the sum of the two importances equals 1. In other words, for a node, compared to the importance of other nodes that directly or indirectly reference it, the first importance determined based on the node's own relevant reference information can account for a larger proportion of the node's importance on the reference path.

[0124] As an optional solution provided in this application, for each path node, the relevant information about the articles that have a citation relationship with the article corresponding to that node may include the first quantity mentioned above, that is, the number of other articles cited by that node; correspondingly, determining the first importance of the node based on the relevant information about the articles that have a citation relationship with the article corresponding to that node includes:

[0125] The primary importance of a node is determined based on the primary quantity corresponding to that node.

[0126] Wherein, the first importance of a node is negatively correlated with the first quantity corresponding to that node; or,

[0127] If the first quantity corresponding to the node is less than or equal to the preset quantity, the first importance of the node is positively correlated with the first quantity corresponding to the node; if the first quantity corresponding to the node is greater than the preset quantity, the first importance of the node is negatively correlated with the first quantity corresponding to the node.

[0128] For an article, if it cites too many articles, its value usually decreases. Considering this factor, the solution provided in this application uses the number of articles cited by the article corresponding to the path node to weight the first importance of that path node. The solution provides two methods, which can be selected according to the actual application scenario or needs. Optionally, the reciprocal of the total number of articles cited by the article corresponding to the path node can be used as the first importance of the path node, or the reciprocal can be used to weight a preset importance value. That is, the more articles cited by the path node, the lower the first importance caused by this factor. Optionally, a threshold can be set. If the number of articles cited by a path node exceeds a preset number, the more articles cited by the node, the smaller its contribution to the importance of the article. If the number of articles cited by a path node is relatively small, the number of article citations can be controlled to be positively correlated with the first importance. For example, the preset number can be configured based on experience or experimental values, and a first correspondence between the number of citations less than or equal to the preset number and the importance value can be pre-configured, as well as a second correspondence between the number of citations greater than the preset number and the importance value. For each path node, after determining the total number of articles cited by the article corresponding to the path node, the first or second correspondence can be selected based on the relationship between the total number and the preset number to determine the importance value corresponding to the total number.

[0129] It is understandable that, in the above alternatives, for a source node on a reference path, the importance of that source node on that reference path is its primary importance.

[0130] by Figure 2Taking the citation path C2→q3→p of article p as an example, the path nodes include the source node C2 and the intermediate node q3. For the source node C2, its importance on the citation path can be determined directly based on the total number of articles it cites (which can be determined from the citation relationship of the reference articles; the other articles cited by the reference articles may include articles to be evaluated in the evaluation set, reference articles in the reference set, or other articles). For the intermediate node q3, its primary importance can be determined first based on the total number of articles it cites. The primary importance and the importance of the source node C2 on the citation path can be added together to obtain the importance of the intermediate node q3 on the citation path.

[0131] For a target article, after determining the importance of each citation path, the overall importance of the target article can be derived based on the importance of each citation path. Optionally, in step S140 above, determining the importance of the target article based on the importance of each citation path may include:

[0132] For each source node appearing in each citation path, the importance of the target article corresponding to the reference article of that source node is determined based on the importance of each path containing that source node in each citation path;

[0133] The importance of the target article is determined based on its importance to each source node.

[0134] For a node corresponding to a reference article, the target article may have multiple citation paths that all originate from that node, such as... Figure 2 In the example shown, for the node C2 corresponding to the reference article C2, the source node of the three reference paths of article P is node C2. That is, starting from the reference article C2, there are multiple different ways to link to the target article. In the optional solution provided by this application, if the source node of multiple paths of the target article is the same node, then the importance of the target article when the reference article corresponding to the source node is taken as a reference can be evaluated based on the importance of each path with the source node as the source node. That is, the importance of the target article to the reference article corresponding to the source node.

[0135] Since the target article's citation path may contain multiple source nodes, meaning it can be linked from different reference articles, once the importance of the target article relative to each source node is determined, its overall importance can be calculated based on that relative importance. For example, the maximum importance value among the target node's relationships to each source node (the more important the article, the higher the importance value) can be used as the target article's overall importance.

[0136] Understandably, in practical applications, if a reference article's corresponding node does not link to the target article's path, the target article's importance relative to that reference article can be set to a preset value, such as 0.

[0137] Optionally, for a source node, the sum of the importance of each path that includes the source node in each citation path of the target article can be used as the importance of the target article corresponding to the reference article of the source node.

[0138] After determining the importance of the target article to each source node, the sum or average of the importance of the target article to each source node can be used as the importance of the target article. Alternatively, the importance can be determined by sorting the importance of the target article to each source node from largest to smallest, and determining the kth importance of the target article to each source node as the importance of the target article, where k≥1.

[0139] The specific value of k can be determined based on experimental or empirical values. As an alternative, k can be set to 3, 4, or 5. The reason for choosing the k-th largest importance value rather than the maximum value in this approach is to avoid situations where an article is cited by one or two important articles but not frequently cited within the field. For example, if the target article's citation path only has a path with a node corresponding to a reference article as its source node, then the target article's importance relative to that reference article will be significantly greater than its importance relative to other reference articles (if the target article's citation path does not have a path with a node corresponding to a reference article as its source node, the target article's importance relative to that reference article can be set to a very small preset value, such as 0). In this case, directly selecting the maximum importance value among the various source nodes would lead to a biased evaluation result that does not reflect reality, essentially treating a low-probability event as the final result. By selecting the kth (k>1) importance of the target article relative to each reference article in descending order of importance, the aforementioned problems can be effectively reduced. This can minimize the occurrence of situations where the importance of the target article relative to each reference article is highly left-skewed or contains several extremely large outliers.

[0140] In an optional embodiment of this application, the importance of the target article can be determined by the following expression:

[0141] R(p)=e -D(p)

[0142]

[0143] D(p) = Min k (D1(p),D2(p),…D i (p))

[0144] Among them, D i (C i ) = 0

[0145]

[0146] Where p represents the target article, R(p) is the importance of the target article, and i represents the reference article C. i D i (p) The target article p corresponds to the reference article C. i The importance can be understood as the distance from the target article (p) to the reference article (c). i The distance, D iThe larger (p) is, the smaller the importance of the characterization is; d is a decay / attenuation factor, which is a preset value that can be determined based on empirical values or experimental values, wherein 0<d<1, q represents C in each citation path of target article p i the article corresponding to any node (i.e., each path node) other than the destination node on each citation path with the corresponding node as the source node, q out represents the number of articles cited by q, and q→p represents that the direction of the citation path is a path ending with article p.

[0147] represents the first importance of article q, D i (q) represents the importance of article q corresponding to reference article C i (determined based on the importance of each node before the node corresponding to article q on the citation path), represents the importance of the node corresponding to q on the citation path. It can be seen that in this optional solution, the importance of target article p relative to reference article C i is the sum of the importance on the citation path of each node other than the node corresponding to article p on each citation path with the node corresponding to C i as the source node, Min k (D1(p), D2(p), … D i (p)) represents the k-th minimum value among D1(p), D2(p), … D i (p), that is, the k-th maximum value among the importance of target article p relative to each reference article.

[0148] wherein, the above D i (C i )=0 indicates that the distance between core articles is 0, for example, in Figure 2 on the basis, there is a directed edge directly from C2 to p, then for article p, the distance of the citation path C2→p is for reference article C1, a citation path of article p corresponding to C1 is C1→C2→p, and the distance of this citation path is

[0149] Assuming there are i reference articles, the distance between the target article and each of the i reference articles can be calculated using the above expression. Substituting the kth minimum of these k distances into the first expression yields the final importance assessment result for the target article. It should be noted that in practical applications, if the target article does not exist in a reference path with the node corresponding to a certain reference article as its source node, the importance of the target article corresponding to that reference article can be directly set to 0, or the above distance can be set to a maximum value. For example, if the node corresponding to the 3rd reference article has no path pointing to the node corresponding to the target article, D3(p) can be set to infinity.

[0150] The article importance assessment method provided in this application can start with a set of core papers that can be evaluated by domain experts and are considered to be the most valuable and prestigious papers in the field, or high-value and highly authoritative articles obtained through other means (i.e., reference articles). Then, these core papers are linked with other papers (articles to be evaluated) through citations. This realizes an article importance assessment method based on the citation path starting from reference articles, which can achieve a more comprehensive, objective and accurate assessment of articles.

[0151] Optionally, the method provided in this application embodiment can allocate the link length from each article to be evaluated to the reference articles (i.e., the first importance of the node corresponding to the aforementioned article) through the reference path, and calculate the distance D from each article to the set of reference articles (i.e., the distance D between each reference article and the reference article set). i The k-th shortest distance (p) is used to determine the ranking score (i.e., the importance of the article to be evaluated) based on the calculated shortest distance.

[0152] The method provided in this application embodiment can theoretically be applied to any field that requires the assessment of article importance. In implementation, it only requires obtaining a set of reference articles in the corresponding field. Based on this set, the method provided in this application embodiment can be used to assess the importance of the articles to be evaluated in that field. For example, it can be applied to the assessment of papers or journal articles in the fields of biology and biomedicine, and the assessment results can be applied to paper search engines in that field. Using the method provided in this application embodiment can effectively improve the performance of academic literature search engine results in that field, better meeting the needs of relevant users. For example, for researchers in biomedicine, practitioners in the medical and health field, and the general public or patients, when they need to find academic articles published by specific titles or authors (such as articles about a certain disease), the method provided in this application embodiment can provide these users with the latest and most cutting-edge research-related professional information, and can also provide authoritative and reliable information. It is especially helpful for experts or ordinary people in other fields who do not focus on a particular field for a long time.

[0153] The method provided in this application is an article importance assessment method based on a citation link graph (i.e., the directed graph mentioned above). This assessment method can be used to rank the articles in the set of articles to be assessed. In other words, the solution provided in this application can realize an article ranking system. This system can be constructed based on a set of core articles in the field to which the article to be assessed belongs. Optionally, these core articles can be ideally defined by experts from the field through manual sorting. The ranking score is calculated based on the shortest distance between the paper (the article to be assessed) and the core paper set, thereby allowing relatively new papers to rank higher when they are cited by one of the core papers or when they themselves belong to the core paper set, i.e., the shortest distance is equal to 0.

[0154] By using a collection of reference articles in the relevant field as a foundation, the method provided in this application has a more reliable evaluation basis. This is because the importance of an article is not simply determined by the number of citations, but by its overall quality. This avoids the possibility of manipulated evaluation results due to manipulation of the number of citations. For example, if the ranking is simply based on the number of citations, a paper cited once by a low-quality paper might be ranked the same as a paper cited once by an internationally authoritative paper, resulting in an unreasonable evaluation result. The method provided in this application can effectively avoid these problems.

[0155] In an optional embodiment of this application, the method may further include:

[0156] When a new article is obtained, the directed graph is updated based on the citation relationship corresponding to the new article, where the new article is either a new article to be evaluated or a new reference article.

[0157] The importance of the articles to be evaluated in the set of articles to be evaluated is reassessed based on the updated directed graph.

[0158] Based on the solution provided in this application, when a new article to be evaluated appears, the evaluation of that new article can be carried out conveniently and quickly, and the importance of existing articles can be updated in real time. This is because the appearance of a new article will cause changes in the directed graph, that is, changes in the citation relationships between articles. Similarly, when a new article is a new reference article, the addition of a new article will also cause changes in the directed graph.

[0159] Specifically, nodes corresponding to new articles can be added to the directed graph, and new directed edges can be added to the directed graph based on the citation relationships between the new articles and existing articles in the set of articles to be evaluated, as well as reference articles in the set of reference articles. After generating the new directed graph, the importance of articles to be evaluated whose citation paths have been updated can be recalculated based on the new directed graph. That is, articles whose paths have been updated can be used as target articles, and the importance of the target articles can be calculated by re-executing the method provided in the embodiments of this application based on the new directed graph.

[0160] Based on the article importance assessment method provided in the embodiments of this application, the embodiments of this application also provide an article retrieval method, which is executed by a retrieval server, such as... Figure 3 As shown, the method may include:

[0161] Step S310: Obtain a search request, which includes one or more search terms;

[0162] Step S320: Based on the search terms in the search request, determine at least one candidate article from the search database that matches the search request;

[0163] Step S330: Based on the importance of each candidate article, determine the target article corresponding to the search request from each candidate article, wherein the importance of each article in the above-mentioned search database is determined by the article importance evaluation method provided in the embodiments of this application;

[0164] Step S340: Send the target article as a search result to the requesting end of the search request.

[0165] In the retrieval scenario, the specific format of the keywords (i.e., search terms) in the aforementioned retrieval request is not limited in this embodiment. They can be a single character, a word, a sentence, or a paragraph, etc. After receiving the retrieval request, the retrieval server can search the database for matching candidate articles based on the keywords contained therein. The strategy for determining whether an article matches the keywords can be configured according to actual application needs; this embodiment is not limited. For example, an article containing all or only some of the keywords may be considered a match. After finding several matching candidate articles in the database, the final retrieval results can be obtained by ranking the candidate articles based on their importance. Articles with importance greater than a set threshold or those ranked from highest to lowest importance are selected. These retrieval results can then be provided to the requester.

[0166] Optionally, the search database can contain article repositories from multiple fields. If the search request includes a specified field, the search can be conducted from the article repository within that specified field based on the search terms. If no field is specified, the search can be conducted from various article repositories. In practical applications, the search service can provide users with field selection options. For example, the user interface of the search client can display field selection options. If the user clicks on this option, various selectable fields can be displayed, from which the user can choose, or the user can directly enter a field (such as a field abbreviation or keywords).

[0167] It is understood that the article importance assessment method provided in this application embodiment can be applied to, but is not limited to, retrieval application scenarios. Theoretically, this method can be applied to any scenario requiring the assessment of article value. To better understand and illustrate the method provided in this application embodiment, the following describes in detail an optional implementation of this application using the example of generating ranking scores for journal articles and papers in the life sciences and biomedicine fields, and retrieving technical literature in this field based on the ranking scores.

[0168] Figure 4 The diagram shows a structural schematic of a retrieval system applicable to an embodiment of this application, as shown below. Figure 4As shown, the retrieval system may include a ranking server 10, a cloud storage server 20, a retrieval server 30, and a terminal device 40. The retrieval server 30 is the server for an application providing article retrieval services. The terminal device can be a user terminal for the application user, who can initiate a retrieval request through the application's user interface. The cloud storage server 20 is a retrieval database storing various journal articles and papers (articles to be evaluated) in the life sciences and biomedicine fields, or it may also store a core collection of articles (reference collection) in this field. The ranking server 10 can calculate the importance of each article to be evaluated in the cloud storage server 20 by executing the article importance evaluation method provided in this embodiment, and rank these articles using the article importance as a ranking score. The ranking results can then be submitted to the cloud storage server, which can store each article to be evaluated in association with its ranking order. After receiving a retrieval request from the terminal device 40, the retrieval server 30 can query matching articles in the cloud storage server 20 and send articles with ranking scores exceeding a threshold as retrieval results to the terminal device 40.

[0169] Figure 5 This diagram illustrates an overall framework of an embodiment of this application. In the aforementioned application scenario, C1, C2, ..., Cn represent core articles in the life sciences and biomedicine fields, p and q represent articles to be evaluated, i.e., articles in the retrieval database that need to be ranked. The directional lines in the diagram represent citation paths between articles. It should be understood that the diagram only schematically illustrates the citation relationships between some articles. For example, with... Figure 5 Taking the core article C3 and article p (the article corresponding to R3(p) in the diagram) as an example, we can calculate the sum of the ranking values ​​of all articles (the importance of the node corresponding to the article on each reference path) on each reference path between article p and core article C3 (referred to as the path sum). Finally, by comparing the above path sums from article p to each core article, we take the kth maximum value as the ranking value of article p.

[0170] Figure 6 The diagram illustrates an optional implementation process in one embodiment of this application. The following is a detailed description in conjunction with... Figures 4 to 6 This scenario embodiment will be described. In this scenario embodiment, Figure 6 The method shown can be derived from Figure 4 The ranking server in the middle performs the following, such as Figure 6 As shown, the implementation process may include the following steps:

[0171] Steps S61 and S62: Obtain a set of articles A that have citation relationships, and obtain a set of core articles B that cite articles in the aforementioned set of articles A.

[0172] In this scenario, group A of articles is the set of articles to be evaluated, and at least some of the articles in this set have citation relationships (as described above). Group B of core articles is the set of reference articles in this scenario, wherein at least one of the cited articles of each core article in this set is the article to be evaluated, meaning that each core article cites at least one other core article. It is understood that steps S61 and S62 may be performed in any order.

[0173] Step S63: Assign length to the segments on the reference path based on the attributes of the reference path and the attributes of the articles attached to that path.

[0174] Step S64: Based on the length of the referenced road segment, calculate the shortest distance from each article in B to each article in A.

[0175] Step S65: For each article in A, determine its ranking score based on the shortest distance between the article and each core article in B.

[0176] Step S66: Generate rankings based on ranking scores.

[0177] After obtaining the aforementioned set of articles A and the set of core articles B, a corresponding directed graph can be constructed based on the citation relationships between these articles.

[0178] This step is used to calculate and determine the length of each directed edge in each directed graph (i.e., the length of the aforementioned road segment), which is the first importance corresponding to the starting point of each directed edge. The longer the road segment, the lower the relative importance of the starting point of the directed edge. In this scenario embodiment, the shortest distance between each article to be evaluated and each core article in A can characterize the importance of the article to be evaluated relative to each core article; the greater the distance, the lower the importance.

[0179] by Figure 2 Taking the example in the diagram, for the article p to be evaluated, node C2 and nodes q1 to q5 are all path nodes on the reference path from the core article C2 (i.e., node C2 of the core article in the directed graph) to article p. That is, all nodes except the destination node p are referenced by... Figure 2As can be seen from the diagram, the width of the directed edge (i.e., link) pointing from each path node to the next node in the referenced path is inconsistent. Different widths can represent road segments with different length values, which in turn represent the primary importance of the starting node of that road segment. For example, the length of the road segment pointing from q4 to q1 can represent the importance of one aspect of node q4. The length of each road segment can be affected by the attributes of the referenced path or multiple attributes of the nodes attached to the referenced path. Optionally, the length of a path can be determined solely by the total number of articles referenced by the starting node of that road segment (i.e., q in the previous text). out The impact of ).

[0180] After determining the length of each reference segment in the directed graph, for each article in A (the article to be evaluated), the distance from each core article to that article via each reference path (i.e., the article's reference path) can be calculated. If a core article has multiple reference paths to it, the sum of these distances can be used as the shortest distance between the article and the core article. Finally, the article's ranking score can be calculated based on the k-th smallest shortest distance among the shortest distances between the article and all core articles. After calculating the ranking score for each article in A, the ranking among all articles in A can be given, for example, sorted by score in ascending order, with higher-ranked articles having higher relative importance.

[0181] The following section uses specific examples to illustrate the calculation process for the ranking score of each article in A.

[0182] As an alternative approach, the ranking score of each article in A corresponding to each core article in B can be calculated iteratively using the following formula (1):

[0183]

[0184] Among them, R i (p) indicates the article to be evaluated from p to the core article C. i The ranking score of article P on the path, that is, the ranking score of article P corresponding to core article C. i Importance, similarly, R i (q) indicates the path from article q to core article C. i The ranking score of article q on the path, which is the core article C. i The ranking score of any article q can be obtained by weighting the reciprocal of the total number of articles it cites, i.e., R. (The node corresponding to the reference path of article p is considered as an indicator of the importance of any path node on that reference path.) i (q) / q out R i(p) can be derived from p corresponding to C. i The ranking score of each node on each reference path is summed to obtain the result. This indicates that article p is not a core article, but rather an article in A that needs to be sorted. Corresponding to the calculation method of expression (1), in practical applications, R... i (C i ) can be preset, for example, R i (C i ) = 1.

[0185] After calculating the ranking score of the article p to be evaluated on each core article's path, the k-th largest (Max) can be selected by sorting the scores from largest to smallest. k The ranking score of article p is taken as the ranking score of article p, as shown in the following expression (2):

[0186] R(p) = Max k (R1(p),R2(p),…R i (p)) (2)

[0187] As an alternative, expressions (1) and (2) above can be transformed into expressions (3) to (5) below. In expressions (3) to (5), the sorting scores in expressions (1) and (2) above are converted by the path length, as follows:

[0188] R(p)=e -D(p) (3)

[0189]

[0190] D(p) = Min k (D1(p),D2(p),…D i (p)) (5)

[0191] Where R(p) represents the ranking score calculated using formulas (1) and (2), and D(p) represents the final shortest distance between article p and the core article. i (p) indicates that article p corresponds to core article C. i The final distance (characterizing the importance of article p corresponding to the core article).

[0192] Formula (3) can be used to convert formula (1) into formula (4). Accordingly, D can be set. i (C i=0, indicating that the distance between the two core articles is 0. The final shortest distance D(p) from any article p to be evaluated to a core article can be defined as the kth minimum value among the shortest distances from article p to each core article, i.e., expression (5), which corresponds to the kth maximum value in expression (2). In other words, a higher ranking score is equivalent to a shorter distance, and the shorter the distance, the more important the paper is, and the higher the ranking score.

[0193] The following is based on Figure 2 Using the directed graph shown as an example, the calculation process of the ranking score D2(p) of article p relative to the core article C2 is explained.

[0194] like Figure 2 As shown, there are three citation paths from core article C2 to article p: path C2→q4→q1→p, path C2→q3→p, and path C2→q5→q2→P. Based on the above expression (4), we can obtain:

[0195]

[0196] Where D2(C2) = 0;

[0197]

[0198]

[0199]

[0200]

[0201]

[0202] It can be seen that D2(p) is the sum of the distances (distances represent importance, with larger distances indicating lower importance) of each node along the three citation paths from core article C2 to article p within their respective citation paths. For example, the distance for the path C2→q4→q1→p is... in, Right now The distance of the source node C2 on this path. This is the distance of intermediate node q4 on this path, which is the sum of the distance to source node C2 (corresponding to second importance) and the distance of q4 itself (corresponding to first importance). Similarly, the distance of intermediate node q1 on this path is... That is, the sum of the distance between the node q4, which is before q1, and the source node C2, and the distance of q1 itself.

[0203] Using the above method, the shortest distance D between article p and each core article can be calculated. i Then, the kth smallest distance D(p) is used as the shortest distance between article p and the core article. This distance is substituted into expression (3) to obtain the ranking score of article p.

[0204] After calculating the ranking score for each article in A, ranking server 10 can send the ranking results and scores of these articles to cloud storage server 20. Cloud storage server 20 can store this data and provide it to retrieval server 30. Retrieval server 30 can provide the best retrieval results to the requester based on the search terms in the search request and the ranking of each article in A. Optionally, retrieval server 30 can sort the ranking scores of the retrieved candidate articles from highest to lowest according to the search keywords, and send the top-ranked articles to terminal device 40. Optionally, the terminal device 40 can display the names, authors, abstract information, etc. of these returned articles, and can also display the search keywords contained in the text to the user in a high-key or other prompt manner.

[0205] For example, an oncologist might discover that late-stage cancer patients experience significant energy loss and weight reduction, and that a considerable number of cancer patients die not from the tumor itself, but from unexplained energy loss (cancer). He might need to find research on lipid metabolism, and he could enter relevant search terms on the search page of his terminal device 40 to retrieve valuable articles in that field from the search server 30.

[0206] For example, olumiant (baricitinib), one of the earliest drugs used to treat severe COVID-19, was originally intended for rheumatoid arthritis, an autoimmune disease. Using this drug undoubtedly requires physicians specializing in fever or respiratory medicine to have knowledge of immunology. Users in need can access authoritative and up-to-date articles in this field through online searches. Similarly, members of the public who suspect they have a disease, or patients needing to understand their symptoms, may wish to find professional research articles to help them understand their possible condition and seek medical attention early. The method provided in this application can guide them to the most authoritative literature reviews and important research progress on the relevant disease.

[0207] The graph-based (directed graph) ranking system provided in this application, by utilizing the citation structure of academic literature, can provide a ranking score for each article. This score represents a measure of the article's importance, and the importance of different citations (articles citing the article) of an article (the aforementioned q) is also considered. outThe weights of each citation are different, and each citation can have a different weight, which can better reflect the importance of each citation. Therefore, the ranking score of the article to be evaluated does not mean the total number of citations of the article, but the weighted sum of all citations. Therefore, the solution provided by the embodiments of this application can achieve a more comprehensive and objective evaluation of the article.

[0208] Furthermore, testing revealed that when the article importance assessment method provided in this application is applied to a retrieval system, compared to some existing retrieval methods, this application...

[0209] It should be noted that the article importance evaluation scheme provided in this application, regarding the method of determining link weight, is defined in some of the previous embodiments as the reciprocal of the total number of cited articles of the article at the starting point of the link (i.e., the article corresponding to the starting point of each directed edge in the directed graph) (i.e., 1 / q in expression (4)). out However, in practice, the method of determining link weight is not limited to this one. More relevant information about the article can be considered when calculating link weight. For example, more features related to the article and its corresponding citation relationship can be considered, such as the citation position and citation frequency of the article.

[0210] To demonstrate the advantages of the method provided in this application, we applied the solution to a retrieval system and compared its performance with that of existing retrieval systems. We used two retrieval systems based on different principles to search for the same keywords and retrieved the returned articles. Then, we compared the retrieval performance of the two systems from several different perspectives: the relevance of the returned articles to the searched keywords, the authority of the returned article ranking, and the category of the returned articles (this is a descriptive rather than quantitative comparison experiment). The comparison revealed that the retrieval system based on the method provided in this application returns articles that are more authoritative overall. Furthermore, during retrieval, this system allows for refined searching by specific sub-fields, or after obtaining the search results, users can filter relevant articles from specific sub-fields based on classification options.

[0211] Based on the same principle as the article importance assessment method provided in the embodiments of this application, the embodiments of this application also provide an article importance assessment device, such as... Figure 7 As shown, the evaluation device 100 may include a data acquisition module 110, a reference path determination module 120, and an importance determination module 130.

[0212] The data acquisition module 110 is used to acquire the directed graph corresponding to the set of articles to be evaluated. The directed graph is constructed based on the citation relationship between the articles in the set of articles to be evaluated and the reference set. Each node in the directed graph represents an article to be evaluated in the set of articles to be evaluated or a reference article in the reference set. The direction of the directed edge in the directed graph is from the node corresponding to the citation article to the node corresponding to the cited article.

[0213] The citation path determination module 120 is used to determine, for any target article in the set of articles to be evaluated, each citation path corresponding to each reference article of the target article according to the directed graph. The source node of each citation path is the node corresponding to a reference article, and the destination node is the node corresponding to the target article.

[0214] The importance determination module 130 is used to determine the importance of each reference path based on the reference relationships of each node on the reference path other than the target node; and to determine the importance of the target article based on the importance of each reference path.

[0215] Optionally, the importance determination module, when determining the importance of a target article based on the importance of each citation path, can be used for:

[0216] For each source node appearing in each citation path, the importance of the target article corresponding to the source node is determined based on the importance of each path containing that source node; the importance of the target article is determined based on the importance of the target article corresponding to each source node.

[0217] Optionally, for each source node, the importance determination module can be used to determine the importance of the target article corresponding to the reference article for that source node by:

[0218] The sum of the importance of each path that contains the source node is used as the importance of the target article in relation to the reference article corresponding to the source node.

[0219] The importance determination module, when determining the importance of a target article based on its importance to each source node, can be used for:

[0220] The importance of the target article is determined by ranking it from most important to least important, and then determining the kth most important value of the target article relative to each source node, where k ≥ 1.

[0221] Optionally, for each reference path, the importance determination module, when determining the importance of the reference path based on the reference relationships between the nodes along the path other than the destination node, can be used for:

[0222] For a source node on the citation path, the importance of the source node on the citation path is determined based on the relevant information of the articles that have a citation relationship with the reference article corresponding to the source node.

[0223] For each node on the citation path other than the source node and the destination node, the importance of the node on the citation path is determined based on the relevant information of the articles that have a citation relationship with the corresponding article and the importance of the nodes that precede the node on the citation path.

[0224] The importance of a reference path is determined based on the importance of each node on the reference path, excluding the destination node.

[0225] Optionally, when determining the importance of a node in a citation path based on information about articles that have a citation relationship with the corresponding article and the importance of the nodes preceding that node in the citation path, the importance determination module can be used for:

[0226] Based on the relevant information of the articles that have a citation relationship with the corresponding article, determine the first importance of the node; based on the importance of each node preceding the node on the citation path, determine the second importance of the node on the citation path; based on the first and second importance of the node, obtain the total importance of the node on the citation path.

[0227] Optionally, for each reference path, the importance determination module can, when determining the importance of the reference path, use the sum of the importance of all nodes on the reference path other than the destination node as the importance of the reference path.

[0228] Optionally, for any node, the relevant information of articles that have a citation relationship with the article corresponding to that node includes at least one of the following:

[0229] The first number of times the first article cites the second article, where the first article is the article corresponding to this node, and the second article is the article cited by the first article;

[0230] The second number of third articles citing the first article;

[0231] The first article's citation frequency or number of times;

[0232] The position of the first article in each of the third articles.

[0233] Optionally, the relevant information of articles that have a citation relationship with the article corresponding to the node includes a first number; when determining the first importance of the node based on the relevant information of articles that have a citation relationship with the article corresponding to the node, the importance determination module can be used to:

[0234] The primary importance of a node is determined based on the primary quantity corresponding to that node.

[0235] Wherein, the first importance of a node is negatively correlated with the first quantity corresponding to that node; or,

[0236] If the first quantity corresponding to the node is less than or equal to the preset quantity, the first importance of the node is positively correlated with the first quantity corresponding to the node; if the first quantity corresponding to the node is greater than the preset quantity, the first importance of the node is negatively correlated with the first quantity corresponding to the node.

[0237] Optionally, the importance of the target article is determined by the following expression:

[0238] R(p)=e -D(p)

[0239]

[0240] D(p) = Min k (D1(p),D2(p),…D i (p))

[0241] Among them, D i (C i ) = 0

[0242]

[0243] Where p represents the target article, R(p) is the importance of the target article, and i represents the reference article C. i D i (p) represents p corresponding to C i The importance of D i The larger (p) is, the less important the representation; Min k (D1(p),D2(p),…D i (p) represents D1(p), D2(p), ..., D i The kth minimum value in (p), k≥1, q represents the minimum value in C. i The corresponding node is the article corresponding to any node on each of the reference paths of the source node, excluding the destination node. This indicates that the node corresponding to q is in the range of C. i The corresponding node represents the importance of a reference path to the source node, q. out This represents the number of articles cited by q, and d is a preset value.

[0244] Optionally, the importance determination module can also be used to: update the directed graph based on the citation relationships corresponding to the new article when a new article is obtained, where the new article is a new article to be evaluated or a new reference article; and redetermine the importance of the articles to be evaluated in the set of articles to be evaluated based on the updated directed graph.

[0245] Based on the same principle as the article retrieval method provided in the embodiments of this application, the embodiments of this application also provide an article retrieval device, such as... Figure 8 As shown, the retrieval device 200 may include a retrieval request acquisition module 210, a retrieval module 220, and a retrieval result feedback module 230.

[0246] The retrieval request acquisition module 210 is used to acquire retrieval requests;

[0247] The retrieval module 220 is configured to determine at least one candidate article matching the retrieval request from the retrieval database based on the retrieval terms in the retrieval request, and to determine the target article corresponding to the retrieval request from the retrieval request based on the importance of each candidate article. The importance of each article in the retrieval database is determined using the article importance assessment method provided in the embodiments of this application.

[0248] The search result feedback module 230 is used to send the target article as a search result to the requesting end of the search request.

[0249] The apparatus in this application embodiment can execute the method provided in this application embodiment, and the implementation principle is similar. The actions performed by each module in the apparatus of each embodiment of this application correspond to the steps in the method of each embodiment of this application. For detailed functional descriptions of each module of the apparatus, please refer to the descriptions in the corresponding methods shown above, which will not be repeated here.

[0250] Based on the methods provided in the embodiments of this application, an electronic device is also provided in the embodiments of this application. The electronic device includes a memory, a processor, and a computer program stored in the memory. When the processor executes the computer program stored in the memory, it can implement the methods in any optional embodiments of this application.

[0251] Optional, Figure 9 The diagram shows a structural schematic of an electronic device to which an embodiment of the present invention applies, such as... Figure 9 As shown, the electronic device can be a server or a user terminal, and it can be used to implement the methods provided in any embodiment of the present invention.

[0252] like Figure 9 As shown, the electronic device 2000 may primarily include at least one processor 2001. Figure 9The diagram shows components such as a memory 2002, a communication module 2003, and an input / output interface 2004. Optionally, these components can be connected and communicate with each other via a bus 2005. It should be noted that... Figure 9 The structure of the electronic device 2000 shown is merely illustrative and does not constitute a limitation on the electronic devices to which the methods provided in the embodiments of this application are applicable.

[0253] The memory 2002 can be used to store operating systems and applications, etc. The applications can include computer programs that implement the methods shown in the embodiments of the present invention when invoked by the processor 2001, and can also include programs for implementing other functions or services. The memory 2002 can be ROM (Read Only Memory) or other types of static storage devices that can store static information and instructions, RAM (Random Access Memory) or other types of dynamic storage devices that can store information and computer programs, or it can be EEPROM (Electrically Erasable Programmable Read Only Memory), CD-ROM (Compact Disc Read Only Memory) or other optical disc storage, optical disk storage (including compressed optical discs, laser discs, optical discs, digital universal optical discs, Blu-ray discs, etc.), magnetic disk storage media or other magnetic storage devices, or any other medium that can be used to carry or store desired program code in the form of instructions or data structures and that can be accessed by a computer, but is not limited thereto.

[0254] Processor 2001 is connected to memory 2002 via bus 2005, and implements corresponding functions by calling application programs stored in memory 2002. Processor 2001 can be a CPU (Central Processing Unit), a general-purpose processor, a DSP (Digital Signal Processor), an ASIC (Application Specific Integrated Circuit), an FPGA (Field Programmable Gate Array), or other programmable logic devices, transistor logic devices, hardware components, or any combination thereof. It can implement or execute the various exemplary logic blocks, modules, and circuits described in conjunction with the disclosure of this invention. Processor 2001 can also be a combination that implements computing functions, such as a combination of one or more microprocessors, a combination of a DSP and a microprocessor, etc.

[0255] Electronic device 2000 can connect to a network via communication module 2003 (which may include, but is not limited to, components such as a network interface) to communicate with other devices (such as user terminals or servers) through the network and achieve data interaction, such as sending data to or receiving data from other devices. Communication module 2003 may include wired network interfaces and / or wireless network interfaces, meaning the communication module may include at least one of wired or wireless communication modules.

[0256] Electronic device 2000 can connect to required input / output devices, such as keyboards and display devices, via input / output interface 2004. Electronic device 2000 itself may have a display device, and other display devices can also be connected externally via interface 2004. Optionally, storage devices, such as hard drives, can also be connected via interface 2004 to store data from electronic device 2000, retrieve data from storage devices, or store data from storage devices into memory 2002. It is understood that input / output interface 2004 can be a wired interface or a wireless interface. Depending on the actual application scenario, the device connected to input / output interface 2004 can be a component of electronic device 2000 or an external device connected to electronic device 2000 when needed.

[0257] The bus 2005 used to connect the various components may include a pathway for transmitting information between these components. The bus 2005 may be a PCI (Peripheral Component Interconnect) bus or an EISA (Extended Industry Standard Architecture) bus, etc. Depending on its function, the bus 2005 can be divided into address bus, data bus, control bus, etc.

[0258] Optionally, for the solution provided in the embodiments of the present invention, the memory 2002 can be used to store a computer program that executes the solution of the present invention, and the processor 2001 runs the computer program. When the processor 2001 runs the computer program, it implements the operation of the method or apparatus provided in the embodiments of the present invention.

[0259] Based on the same principle as the method provided in the embodiments of this application, the embodiments of this application provide a computer-readable storage medium storing a computer program, which, when executed by a processor, can implement the corresponding content of the aforementioned method embodiments.

[0260] This application also provides a computer program product, which includes a computer program that, when executed by a processor, can implement the corresponding content of the aforementioned method embodiments.

[0261] It should be noted that the terms "first," "second," "third," "fourth," "1," "2," etc. (if present) in the specification, claims, and accompanying drawings of this application are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence. It should be understood that such data can be interchanged where appropriate so that the embodiments of this application described herein can be implemented in a sequence other than that shown in the figures or text.

[0262] It should be understood that although arrows indicate various operation steps in the flowcharts of this application's embodiments, the order in which these steps are implemented is not limited to the order indicated by the arrows. Unless explicitly stated herein, in some implementation scenarios of this application's embodiments, the implementation steps in each flowchart can be executed in other orders as required. Furthermore, some or all steps in each flowchart, based on the actual implementation scenario, may include multiple sub-steps or multiple stages. Some or all of these sub-steps or stages can be executed at the same time, and each sub-step or stage can also be executed at different times. In scenarios where execution times differ, the execution order of these sub-steps or stages can be flexibly configured according to requirements, and this application's embodiments do not limit this.

[0263] The above description is only an optional implementation method for some implementation scenarios of this application. It should be noted that for those skilled in the art, other similar implementation methods based on the technical concept of this application without departing from the technical concept of this application also fall within the protection scope of the embodiments of this application.

Claims

1. A method for assessing the importance of an article, characterized in that, The method includes: Obtain a directed graph corresponding to the set of articles to be evaluated. The directed graph is constructed based on the citation relationships between the articles in the set of articles to be evaluated and the reference set. Each node in the directed graph represents an article to be evaluated in the set of articles to be evaluated or a reference article in the reference set. The direction of the directed edges in the directed graph is from the node corresponding to the citation article to the node corresponding to the cited article. The reference set is the core set of articles in the technical field to which the set of articles to be evaluated belongs. For any target article in the set of articles to be evaluated, the citation paths of the target article corresponding to each of the reference articles are determined according to the directed graph. The source node of each citation path is a node corresponding to a reference article, and the destination node is a node corresponding to the target article. For each source node in the aforementioned reference path, the importance of the source node in the reference path is determined based on the relevant information of the articles that have a citation relationship with the reference article corresponding to the source node; for each node in the reference path other than the source node and the destination node, the importance of the node in the reference path is determined based on the relevant information of the articles that have a citation relationship with the article corresponding to the node, as well as the importance of each node in the reference path preceding the node; the importance of the reference path is determined based on the importance of each node in the reference path other than the destination node in the reference path. The importance of the target article is determined based on the importance of each of the cited paths.

2. The method according to claim 1, characterized in that, Determining the importance of the target article based on the importance of each of the aforementioned citation paths includes: For each source node appearing in each of the aforementioned reference paths, the importance of the target article corresponding to the reference article of that source node is determined based on the importance of each path containing that source node in each of the aforementioned reference paths; The importance of the target article is determined based on the importance of the target article corresponding to each of the source nodes.

3. The method according to claim 2, characterized in that, The step of determining the importance of the reference article corresponding to the source node for the target article based on the importance of each path containing the source node in each of the aforementioned reference paths includes: The sum of the importance of each path that contains the source node in each of the aforementioned reference paths is taken as the importance of the target article corresponding to the reference article corresponding to the source node. The step of determining the importance of the target article based on its importance to each of the source nodes includes: The importance of the target article is determined by ranking the target article from highest to lowest importance among the importance of each source node, where k ≥ 1.

4. The method according to claim 1, characterized in that, The step of determining the importance of a node in a citation path based on information about articles that have a citation relationship with the corresponding article, and the importance of the nodes preceding that node in the citation path, includes: Determine the primary importance of the node based on information about articles that have a citation relationship with the corresponding article; The second importance of a node is determined based on the importance of the nodes preceding it on the reference path. The importance of a node on the reference path is determined by its first and second importance.

5. The method according to claim 1, characterized in that, The relevant information of the articles that have a citation relationship with the article corresponding to this node includes at least one of the following: The first number of second articles cited by the first article, wherein the first article is the article corresponding to the node, and the second article is the article cited by the first article; The second number of third articles citing the first article; The citation frequency or number of times the first article is cited; The location where the first article is cited in each of the aforementioned third articles.

6. The method according to claim 5, characterized in that, The relevant information of the articles that have a citation relationship with the article corresponding to the node includes the first number corresponding to the node; The step of determining the primary importance of a node based on relevant information from articles that have a citation relationship with the corresponding article includes: The primary importance of a node is determined based on the primary quantity corresponding to that node. Wherein, the first importance of a node is negatively correlated with the first quantity corresponding to that node; or, If the first quantity corresponding to the node is less than or equal to the preset quantity, the first importance of the node is positively correlated with the first quantity corresponding to the node; if the first quantity corresponding to the node is greater than the preset quantity, the first importance of the node is negatively correlated with the first quantity corresponding to the node.

7. The method according to any one of claims 1 to 6, characterized in that, The method further includes: When a new article is obtained, the directed graph is updated based on the citation relationship corresponding to the new article, wherein the new article is a new article to be evaluated or a new reference article; The importance of the articles to be evaluated in the set of articles to be evaluated is reassessed based on the updated directed graph.

8. An article retrieval method, characterized in that, The method includes: Obtain the search request; Based on the search terms in the search request, at least one candidate article matching the search request is determined from the search database; Based on the importance of each candidate article, the target article corresponding to the search request is determined from each candidate article, wherein the importance of each article in the search database is determined by the method described in any one of claims 1 to 7; The target article is sent as a search result to the requesting end of the search request.

9. A device for assessing the importance of an article, characterized in that, The device includes: The data acquisition module is used to acquire a directed graph corresponding to the set of articles to be evaluated. The directed graph is constructed based on the citation relationships between the articles in the set of articles to be evaluated and the reference set. Each node in the directed graph represents an article to be evaluated in the set of articles to be evaluated or a reference article in the reference set. The direction of the directed edges in the directed graph is from the node corresponding to the citation article to the node corresponding to the cited article. The reference set is a core set of articles in the technical field to which the set of articles to be evaluated belongs. The citation path determination module is used to determine, for any target article in the set of articles to be evaluated, each citation path corresponding to each of the reference articles for the target article according to the directed graph, wherein the source node of each citation path is a node corresponding to a reference article, and the destination node is a node corresponding to the target article; The importance determination module is used to determine the importance of each source node on each reference path based on relevant information of articles that have a citation relationship with the source node; for each node on the reference path other than the source node and the destination node, the importance of that node on the reference path is determined based on relevant information of articles that have a citation relationship with the node and the importance of each node preceding that node on the reference path; the importance of the reference path is determined based on the importance of each node on the reference path other than the destination node; and the importance of the target article is determined based on the importance of each reference path.

10. The apparatus according to claim 9, characterized in that, The importance determination module is used for: For each source node appearing in each of the aforementioned reference paths, the importance of the target article corresponding to the reference article of that source node is determined based on the importance of each path containing that source node in each of the aforementioned reference paths; The importance of the target article is determined based on the importance of the target article corresponding to each of the source nodes.

11. The apparatus according to claim 10, characterized in that, The importance determination module is used for: The sum of the importance of each path that contains the source node in each of the aforementioned reference paths is taken as the importance of the target article corresponding to the reference article corresponding to the source node. The importance of the target article is determined by ranking the target article from highest to lowest importance among the importance of each source node, where k ≥ 1.

12. The apparatus according to claim 9, characterized in that, The importance determination module is used for: Determine the primary importance of the node based on information about articles that have a citation relationship with the corresponding article; The second importance of a node is determined based on the importance of the nodes preceding it on the reference path. The importance of a node on the reference path is determined by its first and second importance.

13. The apparatus according to claim 9, characterized in that, The relevant information of the articles that have a citation relationship with the article corresponding to this node includes at least one of the following: The first number of second articles cited by the first article, wherein the first article is the article corresponding to the node, and the second article is the article cited by the first article; The second number of third articles citing the first article; The citation frequency or number of times the first article is cited; The location where the first article is cited in each of the aforementioned third articles.

14. The apparatus according to claim 13, characterized in that, The relevant information of the articles that have a citation relationship with the article corresponding to the node includes the first number corresponding to the node; The importance determination module is used for: The primary importance of a node is determined based on the primary quantity corresponding to that node. Wherein, the first importance of a node is negatively correlated with the first quantity corresponding to that node; or, If the first quantity corresponding to the node is less than or equal to the preset quantity, the first importance of the node is positively correlated with the first quantity corresponding to the node; if the first quantity corresponding to the node is greater than the preset quantity, the first importance of the node is negatively correlated with the first quantity corresponding to the node.

15. The apparatus according to any one of claims 9 to 14, characterized in that, The importance determination module is also used for: When a new article is obtained, the directed graph is updated based on the citation relationship corresponding to the new article, wherein the new article is a new article to be evaluated or a new reference article; The importance of the articles to be evaluated in the set of articles to be evaluated is reassessed based on the updated directed graph.

16. An article retrieval device, characterized in that, The device includes: The retrieval request acquisition module is used to acquire retrieval requests; A retrieval module is configured to determine at least one candidate article matching the retrieval request from a retrieval database based on the retrieval terms in the retrieval request, and to determine the target article corresponding to the retrieval request from the retrieval request based on the importance of each candidate article, wherein the importance of each article in the retrieval database is determined by the method described in any one of claims 1 to 7; The search result feedback module is used to send the target article as a search result to the requesting end of the search request.

17. An electronic device, characterized in that, The electronic device includes a memory and a processor, the memory storing a computer program, and the processor executing the computer program to implement the method according to any one of claims 1 to 8.

18. A computer-readable storage medium, characterized in that, The storage medium stores a computer program, which, when executed by a processor, implements the method described in any one of claims 1 to 8.

Images