Information pushing method and device, storage medium and terminal

Abstract

An information pushing method and device, a storage medium and a terminal, the method comprising: collecting a target data set according to a first preset time step, the target data set comprising literature data of a candidate user; calculating feature data of the candidate user according to the literature data, inputting the feature data of the candidate user into a trained network model to obtain an evaluation value of the candidate user; calculating a rate of change of the evaluation value of the candidate user over time at the present collection to obtain a growth index of the candidate user at the present collection; and performing information pushing according to the growth index of the candidate user at the present collection. In the scheme, the feature data is input into the trained network model to obtain the evaluation value of each candidate user, the growth index is calculated according to the evaluation result, and information pushing is performed according to the growth index of each candidate user, so that the accuracy of information pushing can be improved.

Description

Technical Field

[0001] This invention relates to the field of data processing technology, and in particular to an information push method and apparatus, storage medium, and terminal. Background Technology

[0002] In existing technologies, the user information to be pushed is usually determined and pushed based on the user's paper and patent literature data. However, with the advent of the big data era and the rapid development of science and technology, there is an increasing amount of data such as papers, patents, and research projects. The user's literature data usually contains invalid literature data (for example, the difference between the publication time of the literature and the time of information push exceeds the preset time difference). Determining the push information based on these invalid literature data often leads to inaccurate information push.

[0003] Therefore, there is an urgent need for an information delivery method that can accurately deliver information based on literature data. Summary of the Invention

[0004] The technical problem solved by this invention is how to accurately push information based on users' academic information such as literature data.

[0005] To address the aforementioned technical problems, this invention provides an information push method, comprising: collecting a target dataset according to a first preset time step, the target dataset including literature data of candidate users; calculating feature data of the candidate users based on the literature data, inputting the feature data of the candidate users into a trained network model to obtain the evaluation value of the candidate users, wherein the trained network model is obtained by training a preset network model using feature data and evaluation values ​​of multiple user samples as training data; calculating and determining the rate of change of the evaluation value of the candidate users over time at the time of the current collection, the evaluation value of the candidate users at the time of the previous collection, and the first preset time step, to obtain the growth index of the candidate users at the time of the current collection; and pushing information based on the growth index of the candidate users at the time of the current collection.

[0006] Optionally, the feature data includes a first evaluation index for each document. The calculation method for the first evaluation index of a document includes: extracting the citation relationships between documents from the document data, and determining a citation network based on the citation relationships between documents. The citation network includes multiple nodes and connecting lines between nodes. The nodes represent documents, and the connecting lines represent citation relationships between documents. A central iteration node is determined. Citation nodes with citation relationships to the central iteration node are determined in the citation network, and the weight of the citation node is allocated to the weight of the central iteration node based on the number of nodes referenced by the citation node. Iteratively, citation nodes with citation relationships to the citation nodes determined in the previous iteration are determined in the citation network, and the weight of the citation node is allocated to the weight of the nodes it references based on the number of nodes it references in the current iteration. The weight change value of the citation node whose weight changed in the previous iteration is allocated to the weight of the nodes it references. When the number of iterations reaches a preset value, or the weight change value of the central iteration node falls within a preset value range, the weight of the central iteration node is the first evaluation index of the document corresponding to the central iteration node.

[0007] Optionally, the feature data includes a second evaluation index, and the calculation method of the second evaluation index includes: extracting the user's published literature data and cited literature data based on the literature data, wherein the user is the candidate user or the user sample, the published literature data includes the number of published literatures, and the cited literature data includes the number of cited literatures; and calculating and determining the user's second evaluation index based at least on the published literature data and the cited literature data.

[0008] Optionally, before calculating the feature data of the candidate users based on the document data, the method further includes: calculating the topic similarity between the document data of candidate users with the same name; if the topic similarity is less than a first preset threshold, determining the proportion of the document data of the candidate users with the same name that contains institutions with the same name within the same time period; if the proportion of the document data of the candidate users with the same name that contains institutions with the same name within the same time period is greater than a second preset threshold, determining that the document data of the candidate users with the same name has the same proportion of collaborating users within the same time period and within the same institution; if the proportion of the same collaborating users within the same time period and within the same institution is greater than a third preset threshold, determining that the candidate users with the same name are the same candidate user, otherwise determining that the candidate users with the same name are different candidate users.

[0009] Optionally, pushing information based on the growth index of the candidate users during this data collection includes: calculating and determining the comprehensive growth index of the candidate users during this data collection based on their growth index during each collection of the target dataset; and pushing information based on the comprehensive growth index of the candidate users during this data collection.

[0010] Optionally, the target dataset includes user identifiers of the candidate users, and the user identifiers correspond one-to-one with the candidate users. Before calculating the feature data based on the document data, the method further includes: sequentially searching for the user identifiers in a preset user dataset; if the user identifier is found in the user dataset, comparing the document data of the candidate user corresponding to the user identifier in the target dataset with the document data in the user dataset to determine whether the document data of the candidate user has been updated; for candidate users whose document data has been updated, updating their document data in the target dataset to the user dataset.

[0011] Optionally, the user dataset has an update marker, which is used to indicate candidate users whose literature data has been updated during this collection. Calculating the feature data based on the literature data and inputting the feature data of the candidate users into the trained network model to obtain the evaluation value of the candidate users includes: scanning the user dataset according to a second preset time step; when the update marker is detected, calculating the feature data of the candidate users indicated by the update marker based on the literature data; and inputting the feature data of the candidate users indicated by the update marker into the trained network model to obtain the evaluation value of the candidate users indicated by the update marker.

[0012] To address the aforementioned technical problems, this invention also provides an information push device, comprising: a collection module for collecting a target dataset according to a first preset time step, the target dataset including literature data of candidate users; a first calculation module for calculating feature data of the candidate users based on the literature data, inputting the feature data of the candidate users into a trained network model to obtain the evaluation value of the candidate users, the trained network model being trained by using feature data and evaluation values ​​of multiple user samples as training data; a second calculation module for calculating and determining the rate of change of the evaluation value of the candidate users over time during the current collection, the evaluation value of the candidate users during the previous collection, and the first preset time step, to obtain the growth index of the candidate users during the current collection; and a push module for pushing information based on the growth index of the candidate users during the current collection.

[0013] This invention also provides a storage medium storing a computer program, which, when run by a processor, executes the steps of the above-described information push method.

[0014] This invention also provides a terminal, including a memory and a processor. The memory stores a computer program that can run on the processor. When the processor runs the computer program, it executes the steps of the above-described information push method.

[0015] Compared with the prior art, the technical solution of the embodiments of the present invention has the following beneficial effects:

[0016] In the embodiment of the present invention, literature data is collected according to a first preset time step. For each collection of literature data, the characteristic data of each candidate user is calculated and determined, and the evaluation value at the time of collection is calculated based on the characteristic data. The growth index is calculated based on the evaluation value at the time of collection, the evaluation value at the time of collection, and the first preset time step. Since the growth index can represent the rate of change of the evaluation value over time, the obtained growth index can accurately represent the changes in the academic information such as literature data of the candidate users within the preset time range. Determining the information to be pushed based on this change can avoid the influence of outdated literature data on information push errors, thereby improving the accuracy of information push.

[0017] Furthermore, when calculating the evaluation value, the feature data used includes the first evaluation index of each document for each candidate user. In this embodiment of the invention, a citation network is constructed based on the document data of multiple candidate users, and the weight of each node is calculated according to the citation relationship of each node in the citation network. Changes in the weight of each node will cause corresponding changes in the weight of the nodes it cites, thereby eliminating invalid citations such as self-citations and mutual citations. The resulting first evaluation index (i.e., the weight of the node) can accurately represent the relative importance of each document. Then, the network model is trained using the feature data including the first evaluation index, and the evaluation values ​​of the candidate users obtained after using the trained network model are used. Since the first evaluation index can more accurately reflect the citation status of documents, the obtained evaluation values ​​and growth indices are also more accurate, improving the accuracy of the push.

[0018] Furthermore, when calculating the evaluation value, the feature data used also includes a second evaluation index for each candidate user. Since the second evaluation index is calculated using published literature data and cited literature data, the second evaluation index can simultaneously reflect the candidate user's publication volume and citation frequency, avoiding the distortion of results caused by using a single parameter. This makes the evaluation value obtained based on the feature data more realistic and complete, and the accuracy of the growth index calculated subsequently is also higher, thus improving the accuracy of the pushed information.

[0019] Furthermore, in the solution of this embodiment of the invention, after each collection of literature data, the feature data of candidate users with the same name are distinguished. By distinguishing candidate users with the same name, the accuracy of the correspondence between literature data and candidate users can be improved, making the evaluation values ​​of candidate users calculated subsequently more realistic and accurate, thereby improving the accuracy of information push. Attached Figure Description

[0020] Figure 1 This is a flowchart illustrating an information push method according to an embodiment of the present invention.

[0021] Figure 2 yes Figure 1 A partial flowchart of the first specific step in step S102.

[0022] Figure 3 yes Figure 1 A partial flowchart of the second specific step in step S102.

[0023] Figure 4 This is a schematic diagram of the structure of an information push device according to an embodiment of the present invention. Detailed Implementation

[0024] As described in the background section, there is an urgent need for an information delivery method that can accurately deliver information based on literature data.

[0025] To address the aforementioned technical problems, this invention provides an information push method. In this embodiment, literature data is collected according to a first preset time step. For each collected literature data, characteristic data of each candidate user is calculated and determined. An evaluation value at the time of collection is calculated based on the characteristic data. A growth index is calculated based on the evaluation value at the time of collection, the evaluation value at the time of the previous collection, and the first preset time step. Since the growth index can represent the rate of change of the evaluation value over time, the obtained growth index can accurately represent the changes in the academic information, such as literature data, of the candidate users within a preset time range. Determining the information to be pushed based on this change can avoid the impact of outdated literature data on information push errors, thereby improving the accuracy of information push. Candidate users with current development potential can be selected as the information to be pushed.

[0026] To make the above-mentioned objectives, features and beneficial effects of the present invention more apparent and understandable, specific embodiments of the present invention will be described in detail below with reference to the accompanying drawings.

[0027] Reference Figure 1 , Figure 1This is a flowchart illustrating an information push method according to an embodiment of the present invention. The information push method can be executed by a terminal, which can be various suitable terminals, such as mobile phones, computers, IoT devices, servers, etc., but is not limited thereto. The pushed information can be academic information of multiple candidate users. The academic information can be user identifiers of multiple candidate users, or at least a portion of the literature data of multiple candidate users, but is not limited thereto.

[0028] Figure 1 The information push method shown may include the following steps:

[0029] Step S101: Collect the target dataset according to the first preset time step, wherein the target dataset includes the literature data of the candidate users;

[0030] Step S102: Calculate the feature data of the candidate users based on the literature data, and input the feature data of the candidate users into the trained network model to obtain the evaluation value of the candidate users. The trained network model is obtained by training a preset network model with the feature data and evaluation values ​​of multiple user samples as training data.

[0031] Step S103: Based on the evaluation values ​​of the candidate users during this collection, the evaluation values ​​of the candidate users during the previous collection, and the first preset time step, calculate and determine the rate of change of the evaluation values ​​of the candidate users during this collection over time, so as to obtain the growth index of the candidate users during this collection.

[0032] Step S104: Push information based on the growth index of the candidate users mentioned in this data collection.

[0033] In the specific implementation of step S101, the terminal can collect literature data from multiple candidate users according to a first preset time step, that is, the terminal can collect literature data periodically. The first preset time step can be pre-set, for example, the first preset time step is 1 year. Each time the terminal collects literature data, it can use web crawling technology to collect literature data from multiple databases. Regarding the specific implementation of web crawling technology, refer to the prior art, and the embodiments of the present invention do not impose any limitations on it.

[0034] It should be noted that the literature referred to in the embodiments of the present invention can be a paper, a book, a patent document, a conference report, a scientific research project report, or any other form of literature, and the embodiments of the present invention do not impose any restrictions on this.

[0035] Therefore, the literature data in this embodiment of the invention may include the title, abstract, keywords, co-authors, publications, and citation data of papers published by each candidate user; it may also include the name, publisher, publication date, and citation data of books published by the candidate user; it may also include bibliographic information of patents for which the candidate user is the inventor or applicant, such as patent name, applicant, priority information, co-inventors, etc.; it may also include the abstract of the patent specification; and it may also include the project name, project leader, project summary, and project participants of research projects.

[0036] It should also be noted that the candidate user to which the literature data belongs each time can be different. That is, the embodiments of the present invention do not impose any restrictions on the candidate users, and the user to which the literature data belongs can be the candidate user in the embodiments of the present invention.

[0037] Furthermore, after each collection of literature data, data cleaning and data integration processing can be performed on the literature data.

[0038] Specifically, each document data entry should include the name of the candidate user to whom the document data belongs, such as the author of the paper, the inventor or applicant of the patent, or the person in charge of the research project. During the data cleaning process, document data that does not include the name of the candidate user can be removed.

[0039] Furthermore, the terminal can also standardize fields such as the name of the candidate user, the institution to which the candidate user belongs, the publishing institution, and the time in the document data according to a preset format. For example, it can standardize the different abbreviations of the same candidate user's name in different documents, distinguish between candidate users with the same name, and standardize the description methods of different institutions.

[0040] In practical applications, different candidate users may have the same name. It is necessary to distinguish between candidate users with the same name in order to avoid errors in subsequent processing and inaccurate push notifications if they are not distinguished.

[0041] Specifically, the terminal can determine whether candidate users with the same name are the same person based at least on the topic similarity between pairs of document data. Topic similarity can be calculated based on data such as titles, keywords, and abstracts between documents. In the specific calculation, natural language processing algorithms can be used to determine semantic similarity, such as the Chinese Lexical Analysis System (ICTCLAS) of the Institute of Computing Technology, jieba word segmentation, and term frequency–inverse document frequency (TF-IDF) algorithms, etc. This embodiment of the invention does not limit this.

[0042] If the topic similarity between two sets of literature reaches a preset threshold, such as 60%, it can be determined that the two sets of literature belong to the same field, and the candidate users with the same name can be verified as the same person; if the topic similarity is found to be below the preset threshold, it is considered that the candidate users with the same name are not the same person.

[0043] In a non-limiting embodiment of the present invention, distinguishing between literature data of candidate users with the same name may include the following steps: calculating the topic similarity between literature data; if the topic similarity is less than a first preset threshold, determining the proportion of candidate users with the same name whose literature data appears in the same time period as institutions with the same name; if the proportion is greater than a second preset threshold, determining the proportion of candidate users with the same name whose literature data has the same proportion of collaborating users within the same time period and the same institution; if the proportion of collaborating users with the same name within the same time period and the same institution is greater than a third preset threshold, determining that candidate users with the same name are the same candidate user, otherwise determining that candidate users with the same name are different candidate users.

[0044] Specifically, due to the different ways of writing the names of candidate users, such as surname first and given name last, given name first and surname last, and abbreviation of names, the names of the same candidate user can be described in different ways. Therefore, the terminal can pre-identify the names of candidate users for each document data and then distinguish between candidate users with the same name.

[0045] Further, the preliminary identification of candidate user names can include the following steps: First, extract the name information of candidate users from all literature data and sort all names. Scan candidate users with identical names in the sorted list and perform a first merge to form a preliminary candidate user set. Then, arrange the remaining candidate users in alphabetical order according to their names and compare them pairwise. If two strings completely match the order of the shorter string, then the two names are considered to be likely to belong to the same candidate user. To further confirm, the last name and first name of the two original names need to be extracted and compared according to different arrangements of the last name and first name (the comparison method is the same as above, but the alphabetical order is not adjusted). If a shortest matching arrangement is found after traversing all arrangements (e.g., Thomas Huang and T. Huang), then the two names are further confirmed to be likely to belong to the same candidate user, and a second merge is performed, adding them to the preliminary candidate user set.

[0046] Furthermore, within the initial candidate user set, candidates with the same name are differentiated. Specifically, all document data for candidate users with the same name are traversed, and the similarity of each document is compared pairwise to confirm whether they belong to the same field. Similarity comparison can be performed using the title, keywords, and abstract of each document, for example, by using ICTCLAS, jieba word segmentation, and TF-IDF algorithms to calculate topic similarity. If the topic similarity exceeds a first preset threshold (e.g., 60%), the two documents can be determined to belong to the same field, preliminarily verifying that the candidate user with the same name is the same person. If the topic similarity is found to be below the first preset threshold, it is considered that the candidate user with the same name may not be the same person, and further differentiation processing is required.

[0047] Furthermore, if the topic similarity is below a first preset threshold, the institution to which the candidate user belongs is identified. Specifically, if the preliminary verification steps above initially determine that the candidate users with the same name are the same, then based on the time field in the literature data, the candidate user's time at that institution can be further identified. If the preliminary verification shows they are not the same, then the institution needs to be further identified. The proportion of candidate users with the same name appearing in the literature data of the same-named candidate users within the same time period (e.g., 5 years) is calculated. If the proportion exceeds a second preset threshold (e.g., 0.5), it indicates two situations: one is that two candidate users with the same name but different specialties appear in the same institution, and the other is that the candidate users with the same name are the same candidate user, but their research direction has changed. Therefore, if the proportion does not exceed the second preset threshold, then the candidate users with the same name are determined to be different candidate users; if the proportion exceeds the second preset threshold, then their collaborating users are further identified.

[0048] Furthermore, if the proportion of identical collaborating users within the same time period and institution exceeds a third preset threshold (e.g., 30%), then the candidates with the same name are considered to be the same candidate user, only their research direction has changed; if the proportion is below the third preset threshold, then the candidates with the same name are different candidates within the same institution. After the above processing, according to the aforementioned rules, it is possible to accurately distinguish candidates with the same name from different literature data, improving the accuracy of subsequent calculations of evaluation values ​​and information delivery. The aforementioned first, second, and third preset thresholds can be pre-set.

[0049] Furthermore, the terminal can also set a user identifier for each candidate user. The user identifier corresponds one-to-one with the candidate user and can uniquely identify the corresponding candidate user. This allows the acquisition of the target dataset, which includes the literature data of multiple candidate users, each candidate user having a unique user identifier.

[0050] In the specific implementation of step S102, after each collection of the target dataset, the terminal can calculate and determine the characteristic data of each candidate user based on the collected literature data. The characteristic data refers to data that can be used to indicate the number of publications, citation frequency, etc.

[0051] Furthermore, the feature data in the embodiments of the present invention may include a first evaluation index for each document and / or a second evaluation index for candidate users.

[0052] refer to Figure 2 , Figure 2 It shows Figure 1 A partial flowchart illustrating the first specific step in step S102. Figure 2 The steps shown can be used to calculate the first evaluation index for each document.

[0053] Figure 2 The method shown for calculating the first evaluation index for each document may include the following steps:

[0054] Step S201: Extract the citation relationships between each document from the document data, and determine the citation network based on the citation relationships between each document. The citation network includes multiple nodes and connecting lines between nodes. The nodes represent documents, and the connecting lines represent citation relationships between documents.

[0055] Step S202: Determine the central iteration node;

[0056] Step S203: In the reference network, identify the reference nodes that have a reference relationship with the central iteration node, and allocate the weight of the reference node to the weight of the central iteration node according to the number of nodes referenced by the reference node.

[0057] Step S204: Iteratively determine the reference nodes in the reference network that have a reference relationship with the reference nodes determined in the previous iteration, and allocate the weight of the reference node to the weight of the reference node it references according to the number of nodes it references in the current iteration. Allocate the weight change value of the reference node whose weight changed in the previous iteration to the weight of the reference node it references. When the number of iterations reaches a preset value, or when the weight change value of the central iteration node falls into a preset value range, the weight of the central iteration node is the first evaluation index of the literature corresponding to the central iteration node.

[0058] It should be noted that in this embodiment of the invention, each node represents a document, and the weight of the node is the first evaluation index of the document corresponding to that node. Since the citation network describes the citation relationship between nodes through connecting lines, the first evaluation index calculated based on the citation relationship can characterize the situation in which the document represented by the node is cited by multiple documents represented by various nodes in the citation network, thereby reflecting the importance of the document.

[0059] The first evaluation index in this embodiment can also be called the Cite Article (CA) index, or any other feasible name, and this embodiment of the invention does not limit it.

[0060] In the specific implementation of step S201, since the documents obtained by the terminal can have citation relationships, the terminal can construct a citation network for multiple documents based on the obtained document data. The citation network includes multiple nodes and connections between nodes. Each node in the citation network represents a document, and the connections between nodes indicate a citation relationship between two documents.

[0061] Furthermore, the connecting line has a direction, the node to which the connecting line points represents the referenced node, and the node connected to the other side of the connecting line represents the referencing node.

[0062] In a non-limiting embodiment of the present invention, before constructing the citation network, the field to which each candidate user belongs can be determined based on the literature data of each candidate user, and more literature data in that field can be obtained as literature data for calculating the first evaluation index. That is, the literature data used by the user to calculate the first evaluation index may include literature data that does not belong to the target dataset. Therefore, all literature with direct or indirect citation relationships to the candidate user's literature can be used to construct the citation network, thereby making the calculated first evaluation index more realistic and accurate in reflecting the citation status of the candidate user's literature, thus reflecting the importance of the candidate user's literature.

[0063] In a specific implementation of step S202, a central iteration node is determined from among multiple nodes in the reference network. The central iteration node can be randomly selected, or it can be a designated node, etc., and this embodiment of the invention does not impose any restrictions on this.

[0064] In the specific implementation of steps S203 and S204, the terminal performs multiple iterative calculations, and the weight of the central iteration node is updated once after each iteration.

[0065] In the specific implementation of step S203, the terminal performs the first iterative calculation.

[0066] Specifically, firstly, the referencing nodes in the referencing network that have a referencing relationship with the central iteration node are identified. Then, the weights of the referencing nodes are allocated to the weights of the central iteration node based on the number of nodes referenced by each referencing node. Identifying the referencing nodes that have a referencing relationship with the central iteration node means identifying other nodes in the referencing network that reference the central iteration node.

[0067] In the specific implementation of step S204, the terminal performs subsequent multiple iterative calculations until the number of iterations reaches a preset value, or the weight change value of the central iterative node falls into a preset numerical range.

[0068] Specifically, in each subsequent iteration, the reference nodes that have a reference relationship with the reference nodes determined in the previous iteration are first identified in the reference network. Then, the weight of the reference node is assigned to the weight of the reference node it references based on the number of nodes it references in the current iteration. Finally, the weight change value of the reference node whose weight has changed in the previous iteration is assigned to the weight of the reference node it references.

[0069] Furthermore, in each iteration, if the referencing node refers to only a single node, then the weight of the referencing node is entirely allocated to the weight of the node it refers to. That is, when a referencing node refers to only one node, all the weight of the referencing node is allocated to the node it refers to. For example, if the weight of a referencing node is 1, and it refers to only one node (such as node A), then its weight 1 is entirely allocated to node A.

[0070] In each iteration, if a node references multiple nodes, the ratio of the weight of the referencing node to the number of nodes is calculated, and this ratio is allocated to the weight of each node it references. That is, when a referencing node references multiple nodes, its weight can be averaged across the referenced nodes. For example, if a referencing node has a weight of 1 and references two nodes (such as node A and node B), then the referencing node allocates 0.5 of its weight to node A and 0.5 to node B.

[0071] Furthermore, if the difference between the weight of the central iteration node after the (N+1)th iteration and the weight after the Nth iteration falls within a preset numerical range, the iteration process is stopped; or, if the number of iterations reaches a preset value, the iteration process is forcibly stopped.

[0072] Furthermore, in each iteration, the weight of the referencing node can be determined by the following steps: calculating the difference between the current weight of the referencing node and the weight assigned to the nodes it refers to, as the weight of the node after allocation. Thus, after the referencing node allocates its weight to the nodes it refers to, the weight of the referencing node will also decrease accordingly, ensuring the balance of weight values ​​among nodes in the referencing network.

[0073] In a non-limiting embodiment of the present invention, the weight of a reference node can be determined in each iteration by the following steps: calculating the sum of the weights assigned to reference nodes whose weights have changed during the current iteration, as the weight change value; and continuing to assign the weight change value to the weights of the referenced nodes according to the number of nodes referenced by the reference node whose weight has changed.

[0074] If the node referenced by the node whose weight has changed is not the central iteration node, the weight change value of the referenced node continues to be assigned to the node it references until the referenced node is the central iteration node.

[0075] At this point, the weight of the central iteration node can be determined. The weight change value of the central iteration node is the sum of the weights assigned to the central iteration node after multiple iterations. The weight value of the central iteration node is the sum of its initial weight and the weights assigned to the central iteration node after multiple iterations. That is, the first evaluation index of the literature corresponding to the central iteration node can be determined.

[0076] Furthermore, for each node in the citation network, to calculate its first evaluation index, each node can be used as the central iteration node to execute steps S203 and S204 above. Thus, the first evaluation index of each document in the target dataset can be obtained. It should be noted that if the document data used to calculate the first evaluation index includes document data that does not belong to the candidate users, then it is not necessary to use the nodes corresponding to the documents that do not belong to the candidate users as the central iteration nodes.

[0077] refer to Figure 3 , Figure 3 It shows Figure 1 A partial flowchart illustrating the second specific step of step S102. (Through...) Figure 3 The steps shown can be used to calculate the second evaluation index for each candidate user.

[0078] Figure 3 The calculation shown to determine the second evaluation index for each candidate user may include the following steps:

[0079] Step S301: Extract the published literature data and cited literature data of the candidate users based on the literature data. The published literature data includes the number of published literatures, and the cited literature data includes the number of cited literatures.

[0080] Step S302: Calculate and determine the second evaluation index of the candidate users based at least on the published literature data and the cited literature data.

[0081] In the specific implementation of step S301, for each candidate user, the published literature data of the candidate user can be determined. The published literature data may include the number of publications published by the candidate user and the data of each publication published by the candidate user, such as the title, year, source, institution, keywords and abstract of each publication. The cited literature data of each candidate user can also be determined. The cited literature refers to the literature that cites the publications of the candidate user. The cited literature data may include the number of cited literature, the title, year, source, institution, keywords and abstract of the cited literature.

[0082] In the specific implementation of step S302, the second evaluation index can be calculated using at least the number of published documents and the number of cited documents. The second evaluation index is positively correlated with both the number of published documents and the number of cited documents; the more documents a candidate user publishes, the higher their second evaluation index; the more documents a candidate user cites, the higher their second evaluation index.

[0083] Therefore, the second evaluation index is positively correlated with both the number of publications and the number of citations of the candidate user, thus avoiding the distortion caused by using a single parameter. For candidate users with few publications but high citation counts, the above calculation method can obtain a higher second evaluation index, solving the problem in existing technologies that cannot simultaneously determine the publication data and citation counts of documents based solely on the number of publications or citations. In subsequent steps, the evaluation value and growth index calculated based on feature data including the second evaluation index are more realistic and accurate, improving the accuracy of information delivery.

[0084] It should be noted that the second evaluation index can also be called the H++ index, or any other feasible name, and the embodiments of the present invention do not limit this.

[0085] In a non-limiting embodiment of the present invention, step S302 may include the following steps: calculating the total number of cited references using the cited reference data; determining the number of zero-cited references in the published reference data, and calculating the weighted sum of the number of zero-cited references and the number of remaining references; and calculating the second evaluation index of the at least some users based on the total number of cited references and the weighted sum.

[0086] Specifically, some of the candidate users' published works contain zero-citation documents, meaning these documents have no cited references. The impact of zero-citation documents on a user's second evaluation index is relatively small. Therefore, the second evaluation index can be calculated by weighting the number of zero-citation documents with the number of remaining documents, thus reducing the impact of zero-citation documents on the second evaluation index. This weighted sum reflects the number of documents published by the candidate users.

[0087] In addition, the total number of cited references can be calculated. This involves summing the number of citations for all publications by each candidate user. Based on this total number of citations and the weighted sum mentioned above, a second evaluation index for that candidate user can be calculated.

[0088] It should be noted that any feasible mathematical operation can be used to calculate the total number of cited documents and the aforementioned weighted sum to obtain the second evaluation index for candidate users. For example, the total number of cited documents and the aforementioned weighted sum can be summed or multiplied directly; or, if the total number of cited documents is large, to avoid its excessive influence on the second evaluation index, operations such as taking its square root, cube root, or logarithm can be performed, and then summed or multiplied with the aforementioned weighted sum, etc. The embodiments of this invention do not impose any restrictions on this.

[0089] In another non-limiting embodiment of the present invention, step S302 may include the following steps: calculating the total number of cited documents using the cited document data; determining the number of zero-cited documents in the published document data, and calculating the weighted sum of the number of zero-cited documents and the remaining document count; calculating the H-index of each candidate user; and combining the H-index with the total number of cited documents and the weighted sum to obtain a second evaluation index for each candidate user. That is, in addition to calculating the total number of cited documents and the aforementioned weighted sum, the H-index of each candidate user is also calculated, and the second evaluation index of each candidate user is calculated based on the H-index, the total number of cited documents, and the aforementioned weighted sum. It should be noted that any feasible mathematical operation can be used to operate on the H-index, the total number of cited documents, and the aforementioned weighted sum to obtain the second evaluation index of the candidate user. The specific implementation method for calculating the H-index can refer to existing methods, and will not be repeated here in the embodiments of the present invention.

[0090] In a specific implementation of step S301, the terminal may also establish a publication database and a citation database based on the literature data. The publication database includes the literature data published by each candidate user, and the citation database includes the literature data cited by the literature data published by each candidate user. The number of publications published by each candidate user and the number of cited literatures are determined based on the publication database and the citation database.

[0091] Specifically, before calculating the second evaluation index, the terminal can split and supplement the acquired literature data to form a publication database and a citation database. The publication database is mainly used to store all literature data published by the candidate users (including title, publication year, source journal, affiliated institution, keywords, abstract, etc.), while the citation database is mainly used to store data on how the candidate users' published literature is cited by other literature, thus enabling the determination of citation data such as the number of cited literature for each candidate user.

[0092] Furthermore, the determination of the published literature data of the candidate users in step S301 can be based on the data in the publication database; the determination of the cited literature data of the candidate users can be based on the data in the citation database.

[0093] Continue to refer to Figure 1 In another non-limiting embodiment of the present invention, the feature data may further include a third evaluation index, which can be used to indicate the degree of matching between each candidate user's literature and a preset field. The preset field may be pre-set, and the specific content of the preset field is not limited in this embodiment of the present invention.

[0094] The literature data includes a time field, such as the publication year of a paper, the application date of a patent, and the publication date of a book, but is not limited to these. The calculation method for the third evaluation index includes: for each candidate user, filtering literature data belonging to a preset time range based on the time field in the literature data; filtering literature data containing the preset field from the academic data belonging to the preset time range; and calculating the proportion of academic data containing the preset field to the literature data belonging to the preset time range, to obtain the third evaluation index.

[0095] Specifically, for each candidate user's literature data, the terminal can filter out literature data belonging to a preset time range. This preset time range can be pre-set, for example, the most recent three years (January 1, 2018 to December 31, 2020). For the filtered literature data belonging to the preset time range, further filtering is performed to select literature data containing preset fields. These preset fields can be fields indicating the main content of the literature, such as "artificial intelligence" or "image processing." The proportion of academic data containing the preset fields to the literature data belonging to the preset time range is calculated to obtain the third evaluation index. The larger the proportion, the more the main content of the candidate user's published literature within the preset time range matches the preset fields.

[0096] In practical applications, the preset field can be determined according to the academic information to be pushed. For example, if the information to be pushed is the academic data of a preset number of candidate users in the field of image processing, the preset field can be set to "image processing". The more relevant the candidate user's published literature is to "image processing", the higher the third evaluation index of the candidate user will be.

[0097] It should be noted that the feature data in the embodiments of the present invention may also include other evaluation indices. For example, it may also include a fourth evaluation index, which can be used to indicate the number of collaborating users of the candidate user and the citation status of the documents published by the collaborating users; it may also include a fifth evaluation index, which can be used to indicate the number of valid patents and patent licenses of the candidate user as an inventor or applicant.

[0098] Furthermore, after calculating and determining the feature data of each candidate user, the feature data can be input into the trained network model to obtain the evaluation value of each candidate user. The trained network model is obtained by training a preset network model using the feature data and evaluation values ​​of multiple user samples as training data.

[0099] Specifically, before inputting the feature data of each candidate user into the trained network model, the terminal can use the feature data and evaluation values ​​of multiple user samples as training data to train the preset network model. Using the feature data and evaluation values ​​of multiple user samples to train the network model allows the trained network model to learn the relationship between the feature data and the evaluation values. Therefore, when the feature data of candidate users is subsequently input into the trained network model, the evaluation values ​​of the candidate users can be obtained. These evaluation values ​​can be numerical scores.

[0100] Before training the preset network model, the terminal can acquire literature data from multiple user samples and calculate and determine the feature data and evaluation value of each user sample based on the literature data from multiple user samples.

[0101] It should be noted that the specific details of calculating and determining the feature data of each user sample based on the literature data of multiple user samples can be found in the above description of determining the feature data of each candidate user based on the literature data of candidate users, and will not be repeated here.

[0102] Furthermore, the evaluation value for each user sample can be calculated using the analytic hierarchy process (AHP) on the literature data of each user sample. Specific methods can be found in existing resources and will not be elaborated upon here.

[0103] Furthermore, after the training process is completed, the feature data of multiple candidate users can be input into the trained network model to obtain the evaluation value of each candidate user when collecting literature data.

[0104] In a non-limiting embodiment of the present invention, the terminal may locally store a preset user dataset, which may include literature data of multiple candidate users, and each candidate user in the user dataset has a unique user identifier.

[0105] Before step S102, the following steps may also be included: sequentially searching for user identifiers in the target dataset within a preset user dataset; if the user identifier is found in the user dataset, comparing the literature data of the candidate user corresponding to the user identifier in the target dataset with the literature data in the user dataset to determine whether the literature data of the candidate user in the user dataset has been updated; for candidate users whose literature data has been updated, updating their literature data in the target dataset to the user dataset.

[0106] Specifically, the terminal can sequentially search for the user identifier of each candidate user in the target dataset collected this time in the user dataset; if the user identifier of each candidate user in the target dataset is not found in the user dataset, it means that the preset user dataset does not include the literature data of the candidate user, and the user identifier and literature data of the candidate user can be directly stored in the user dataset.

[0107] If the user identifier is found in the user dataset, the literature data of the candidate user corresponding to the user identifier in the target dataset is compared with the literature data in the user dataset to determine whether the literature data of the candidate user has been updated; for candidate users whose literature data has been updated, their literature data in the target dataset is updated in the user dataset.

[0108] Therefore, after collecting the target dataset, the collected literature data can be updated to the user dataset, which stores the latest literature data of multiple candidate users.

[0109] It should be noted that the terminal can also directly store the literature data of multiple candidate users obtained in this transaction into the user dataset, and overwrite the original literature data in the user dataset to update the data in the user dataset.

[0110] It should also be noted that the preset user dataset can also be stored on an external server, and the terminal can read or write data from the external server through interaction.

[0111] Furthermore, during this data collection, the terminal can create an update flag for candidate users whose literature data in the user dataset has been updated. It should be noted that "updated literature data" here can refer to changes in the literature data of candidate users in the user dataset, or it can refer to the addition of literature data from one or more candidate users in the target dataset to the user dataset.

[0112] Furthermore, the terminal can scan the user dataset according to a second preset time step. When an update marker is detected, it calculates the feature data of the candidate user indicated by the update marker based on the literature data in the user dataset. Then, it inputs the feature data of the candidate user indicated by the update marker into the trained network model to obtain the evaluation value of the candidate user indicated by the update marker. The calculated evaluation value is saved to the user dataset, and a mapping relationship is established with the candidate user. Thus, the user dataset can store the evaluation values ​​of candidate users after each collection of literature data. The second preset time step can be pre-set.

[0113] Furthermore, when scanning the user dataset according to the second preset time step, for candidate users who do not have an update marker, the evaluation value of the candidate user in the previous collection is used as the evaluation value in this collection and saved to the user dataset.

[0114] In the specific implementation of step S103, after the terminal calculates the evaluation value of each candidate user at the time of this collection, it can calculate and determine the rate of change of the evaluation value of the candidate user with time at the time of this collection based on the evaluation value of the candidate user at the time of this collection, the evaluation value of the candidate user at the time of the previous collection, and the first preset time step.

[0115] Specifically, the terminal can create an "evaluation value-time" relationship graph for each candidate user. The horizontal axis of the relationship graph is time, and the vertical axis is the user's evaluation value. After each evaluation value is calculated, the evaluation point at this time can be marked on the relationship graph based on the time of this literature data collection and the evaluation value calculated at this time. The slope is calculated based on the horizontal and vertical coordinates of the evaluation point marked on the relationship graph at the time of the last literature data collection. The slope is the growth index of the candidate user at the time of this collection.

[0116] In a non-limiting embodiment of the present invention, after calculating the evaluation value each time literature data is collected, the evaluation value obtained in this collection can be saved to the user database for subsequent calculation of the growth index.

[0117] Specifically, when scanning the user dataset according to the second preset step size, the calculated evaluation value can be saved to the user dataset. Since the evaluation value obtained each time is saved in the user dataset, the rate of change of the evaluation value over time can be calculated based on the evaluation value saved in the user dataset during the previous collection and the evaluation value during the current collection, as well as the first preset time step size.

[0118] In a non-limiting embodiment of the present invention, after calculating the growth index, the terminal can further calculate and determine the comprehensive growth index for the current collection based on the growth index of each collection of candidate users, and push information based on the comprehensive growth index. For example, the terminal calculates the comprehensive growth index based on the growth index of the current collection and the growth index of the previous first preset number of collections. The comprehensive growth index is calculated based on the growth index of the current collection and the previous first preset number of collections, and information can be pushed based on the comprehensive growth index. The comprehensive growth index can be calculated by performing mathematical operations such as summation, product, and averaging on multiple growth indices. The first preset number can be pre-set.

[0119] In the specific implementation of step S104, after calculating and determining the growth index of each candidate user, information can be pushed based on the growth index of multiple candidate users.

[0120] Furthermore, the pushed information can be the user identifiers of one or more candidate users. For example, multiple candidate users can be sorted according to their growth index, and then a second preset number of candidate user identifiers can be selected as the pushed information in descending order of growth index. Alternatively, the user identifiers of candidate users with growth indices higher than a preset value can be used as the pushed information. The second preset number can be pre-set.

[0121] Furthermore, while pushing the user identifier, the growth index of the candidate user being pushed can also be pushed at the same time, as can at least a portion of the candidate user's literature data, but it is not limited thereto. The embodiments of the present invention do not impose any restrictions on the specific content and form of the pushed information.

[0122] refer to Figure 4 This invention also provides an information push device, comprising: a collection module 41, used to collect a target dataset according to a first preset time step, the target dataset including literature data of candidate users; a first calculation module 42, used to calculate feature data of the candidate users based on the literature data, and input the feature data of the candidate users into a trained network model to obtain the evaluation value of the candidate users, the trained network model being obtained by training a preset network model using feature data and evaluation values ​​of multiple user samples as training data; a second calculation module 42, used to calculate and determine the rate of change of the evaluation value of the candidate users over time at the time of this collection based on the evaluation value of the candidate users at the time of this collection, the evaluation value of the candidate users at the time of the previous collection, and the first preset time step, to obtain the growth index of the candidate users at the time of this collection; and a push module 44, used to push information based on the growth index of the candidate users at the time of this collection.

[0123] For more information on the working principle, operation mode, and beneficial effects of the aforementioned information push device, please refer to the above section. Figures 1 to 3 The relevant descriptions will not be repeated here.

[0124] This invention also provides a storage medium, which is a computer-readable storage medium storing a computer program thereon, the computer program being executable during runtime. Figures 1 to 3 The steps of the method shown are illustrated. The storage medium may include ROM, RAM, disk, or optical disk, etc. The storage medium may also include non-volatile memory or non-transitory memory, etc.

[0125] This invention also provides a terminal, which may include a memory and a processor. The memory stores a computer program that can run on the processor. When the processor runs the computer program, it can execute... Figures 1 to 3 The steps of the method shown are as follows. The terminal includes, but is not limited to, mobile phones, computers, tablets, and other terminal devices.

[0126] In this application, the term "terminal" can refer to various forms of terminals, access terminals, user units, user stations, mobile stations, mobile stations (MS), remote stations, remote terminals, mobile devices, user terminals, terminal equipment, wireless communication equipment, user agents, or user devices. Terminal equipment can also be cellular phones, cordless phones, Session Initiation Protocol (SIP) phones, Wireless Local Loop (WLL) stations, Personal Digital Assistants (PDAs), handheld devices with wireless communication capabilities, computing devices, or other processing devices connected to a wireless modem, in-vehicle devices, wearable devices, terminal equipment in future 5G networks, or terminal equipment in future evolved Public Land Mobile Networks (PLMNs), etc. This application does not limit the scope of these terminal devices.

[0127] It should be understood that the term "and / or" in this article is merely a description of the relationship between related objects, indicating that three relationships can exist. For example, A and / or B can represent: A existing alone, A and B existing simultaneously, and B existing alone. Additionally, the character " / " in this article indicates that the preceding and following related objects have an "or" relationship.

[0128] In the embodiments of this application, "multiple" refers to two or more.

[0129] The descriptions of "first," "second," etc., appearing in the embodiments of this application are for illustrative purposes and to distinguish the objects being described. They have no order and do not indicate any special limitation on the number of devices in the embodiments of this application, nor do they constitute any limitation on the embodiments of this application.

[0130] In this application embodiment, "connection" refers to various connection methods such as direct connection or indirect connection to achieve communication between devices. This application embodiment does not limit this in any way.

[0131] It should be understood that in the embodiments of this application, the processor can be a central processing unit (CPU), or it can be other general-purpose processors, digital signal processors (DSPs), application-specific integrated circuits (ASICs), field-programmable gate arrays (FPGAs), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. The general-purpose processor can be a microprocessor or any conventional processor.

[0132] It should also be understood that the memory in the embodiments of this application can be volatile memory or non-volatile memory, or may include both volatile and non-volatile memory. The non-volatile memory can be read-only memory (ROM), programmable read-only memory (PROM), erasable programmable read-only memory (EPROM), electrically erasable programmable read-only memory (EEPROM), or flash memory. The volatile memory can be random access memory (RAM), which is used as an external cache. By way of example, but not limitation, many forms of random access memory (RAM) are available, such as static RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), double data rate SDRAM (DDR SDRAM), enhanced synchronous DRAM (ESDRAM), synchronous linked DRAM (SLDRAM), and direct rambus RAM (DR RAM).

[0133] The above embodiments can be implemented, in whole or in part, by software, hardware, firmware, or any other combination thereof. When implemented using software, the above embodiments can be implemented, in whole or in part, as a computer program product. The computer program product includes one or more computer instructions or computer programs. When the computer instructions or computer programs are loaded or executed on a computer, all or part of the processes or functions described in the embodiments of this application are generated. The computer can be a general-purpose computer, a special-purpose computer, a computer network, or other programmable device. The computer instructions can be stored in a computer-readable storage medium or transmitted from one computer-readable storage medium to another. For example, the computer instructions can be transmitted from one website, computer, server, or data center to another website, computer, server, or data center via wired or wireless means. The computer-readable storage medium can be any available medium that a computer can access or a data storage device such as a server or data center that includes one or more sets of available media. The available medium can be a magnetic medium (e.g., floppy disk, hard disk, magnetic tape), an optical medium (e.g., DVD), or a semiconductor medium. A semiconductor medium can be a solid-state drive.

[0134] In the several embodiments provided in this application, it should be understood that the disclosed methods, apparatuses, and systems can be implemented in other ways. For example, the apparatus embodiments described above are merely illustrative; for example, the division of units is merely a logical functional division, and other division methods may exist in actual implementation; for example, multiple units or components may be combined or integrated into another system, or some features may be ignored or not executed. Furthermore, the coupling or direct coupling or communication connection shown or discussed may be through some interfaces, and the indirect coupling or communication connection between devices or units may be electrical, mechanical, or other forms.

[0135] The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the units can be selected to achieve the purpose of this embodiment according to actual needs.

[0136] Furthermore, the functional units in the various embodiments of the present invention can be integrated into one processing unit, or each unit can be physically comprised separately, or two or more units can be integrated into one unit. The integrated unit described above can be implemented in hardware or in the form of hardware plus software functional units.

[0137] The integrated units implemented as software functional units described above can be stored in a computer-readable storage medium. These software functional units, stored in a storage medium, include several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute some steps of the methods described in the various embodiments of the present invention. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0138] While the present invention has been disclosed above, it is not limited thereto. Any person skilled in the art can make various modifications and alterations without departing from the spirit and scope of the invention; therefore, the scope of protection of the present invention should be determined by the scope defined in the claims.

Claims

1. An information push method, characterized in that, The method includes: The target dataset is collected according to a first preset time step, and the target dataset includes literature data of candidate users; The feature data of the candidate users are calculated based on the literature data. The feature data of the candidate users are then input into the trained network model to obtain the evaluation value of the candidate users. The trained network model is obtained by training a preset network model using the feature data and evaluation values ​​of multiple user samples as training data. Based on the evaluation values ​​of the candidate users during this data collection, the evaluation values ​​of the candidate users during the previous data collection, and the first preset time step, the rate of change of the evaluation values ​​of the candidate users during this data collection is calculated to obtain the growth index of the candidate users during this data collection. Information will be pushed based on the growth index of the candidate users mentioned during this data collection. The feature data includes a first evaluation index for each document, and the calculation method for the first evaluation index of a document includes: The citation relationships between the documents are extracted from the document data, and a citation network is determined based on the citation relationships between the documents. The citation network includes multiple nodes and connecting lines between the nodes. The nodes represent documents, and the connecting lines represent citation relationships between documents. Determine the central iteration node; In the reference network, identify the reference nodes that have a reference relationship with the central iteration node, and allocate the weight of the reference node to the weight of the central iteration node according to the number of nodes referenced by the reference node. Iteratively, reference nodes that have a reference relationship with the reference nodes determined in the previous iteration are identified in the reference network. The weight of the reference node is assigned to the weight of the reference node it references based on the number of nodes it references in the current iteration. The weight change value of the reference node whose weight changed in the previous iteration is assigned to the weight of the reference node it references. When the number of iterations reaches a preset value, or when the weight change value of the central iteration node falls into a preset value range, the weight of the central iteration node is the first evaluation index of the literature corresponding to the central iteration node.

2. The information push method according to claim 1, characterized in that, The feature data includes a second evaluation index, and the calculation method for the second evaluation index includes: Based on the literature data, the user's published literature data and cited literature data are extracted. The user is the candidate user or the user sample. The published literature data includes the number of published literatures, and the cited literature data includes the number of cited literatures. The user's second evaluation index is calculated based at least on the published literature data and the cited literature data.

3. The information push method according to claim 1, characterized in that, Before calculating the characteristic data of the candidate users based on the literature data, the method further includes: Calculate the topic similarity between the literature data of candidate users with the same name; If the topic similarity is less than a first preset threshold, then the proportion of the literature data of the candidate users with the same name appearing with the same institution in the same time period is determined. If the proportion of institutions with the same name appearing in the same time period is greater than the second preset threshold, then it is determined that the proportion of the literature data of the candidate users with the same name is the same as that of the collaborating users in the same time period and the same institution. If the proportion of cooperating users within the same time period and the same institution is greater than a third preset threshold, then the candidate users with the same name are determined to be the same candidate user; otherwise, the candidate users with the same name are determined to be different candidate users.

4. The information push method according to claim 1, characterized in that, Information pushes are made based on the growth index of the candidate users mentioned during this data collection, including: The comprehensive growth index of the candidate users at the time of this collection is calculated based on the growth index of the candidate users at each time the target dataset is collected. The information is pushed based on the comprehensive growth index of the candidate users mentioned in this data collection.

5. The information push method according to claim 1, characterized in that, The target dataset includes user identifiers of the candidate users, with each user identifier corresponding one-to-one with a candidate user. Before calculating the feature data based on the literature data, the method further includes: Search for the user identifier sequentially in the preset user dataset; If the user identifier is found in the user dataset, the literature data of the candidate user corresponding to the user identifier in the target dataset is compared with the literature data in the user dataset to determine whether the literature data of the candidate user has been updated. For candidate users whose literature data has been updated, their literature data in the target dataset is updated in the user dataset.

6. The information push method according to claim 5, characterized in that, The user dataset has update markers, which are used to indicate candidate users whose literature data has been updated during this collection. The feature data is calculated based on the literature data, and the feature data of the candidate users is input into the trained network model to obtain the evaluation values ​​of the candidate users, including: The user dataset is scanned according to a second preset time step. When the update marker is scanned, the feature data of the candidate user indicated by the update marker is calculated based on the literature data. The feature data of the candidate user indicated by the update marker is input into the trained network model to obtain the evaluation value of the candidate user indicated by the update marker.

7. An information push device, characterized in that, The device includes: The acquisition module is used to acquire a target dataset according to a first preset time step, wherein the target dataset includes literature data of candidate users; The first calculation module is used to calculate the feature data of the candidate users based on the literature data, and input the feature data of the candidate users into the trained network model to obtain the evaluation value of the candidate users. The trained network model is obtained by training a preset network model with the feature data and evaluation values ​​of multiple user samples as training data. The second calculation module is used to calculate and determine the rate of change of the evaluation value of the candidate user over time during the current collection based on the evaluation value of the candidate user during the previous collection, the evaluation value of the candidate user during the previous collection, and the first preset time step, so as to obtain the growth index of the candidate user during the current collection. The push module is used to push information based on the growth index of the candidate users mentioned in this data collection. The feature data includes a first evaluation index for each document, and the calculation method of the first evaluation index for the document by the first calculation module includes: The citation relationships between the documents are extracted from the document data, and a citation network is determined based on the citation relationships between the documents. The citation network includes multiple nodes and connecting lines between the nodes. The nodes represent documents, and the connecting lines represent citation relationships between documents. Determine the central iteration node; In the reference network, identify the reference nodes that have a reference relationship with the central iteration node, and allocate the weight of the reference node to the weight of the central iteration node according to the number of nodes referenced by the reference node. Iteratively, reference nodes that have a reference relationship with the reference nodes determined in the previous iteration are identified in the reference network. The weight of the reference node is assigned to the weight of the reference node it references based on the number of nodes it references in the current iteration. The weight change value of the reference node whose weight changed in the previous iteration is assigned to the weight of the reference node it references. When the number of iterations reaches a preset value, or when the weight change value of the central iteration node falls into a preset value range, the weight of the central iteration node is the first evaluation index of the literature corresponding to the central iteration node.

8. A storage medium, characterized in that, It stores a computer program, characterized in that, when the computer program is run by a processor, it executes the steps of the information push method according to any one of claims 1 to 6.

9. A terminal comprising a memory and a processor, wherein the memory stores a computer program executable on the processor, characterized in that, When the processor runs the computer program, it executes the information push method according to any one of claims 1 to 6.

Images