Entity influence evaluation system and method based on network data

Abstract

The present invention discloses an entity influence evaluation system and method based on network data, wherein the target entity input module in the system is used to obtain the target entity and determine the set characteristic parameter structure; the characteristic parameter acquisition module is used to obtain the characteristic parameter group of the target entity from the network database based on the set characteristic parameter structure, and use it as a dimension parameter; the influence index determination module is used to obtain the influence index of the target entity through the web page ranking algorithm based on the preset reference relationship in the network database, and use it as a dimension parameter; the diversity index determination module is used to obtain the diversity index of the target entity through the preset knowledge complexity acquisition method, and use it as a dimension parameter; the central processing unit performs weighted operations on all dimensional parameters of the target entity to obtain the influence parameter of the target entity. A high-quality entity influence acquisition system is provided, which can be fully automated, updated in real time, has adjustable multi-dimensional parameters, and has many participating parameters and high parameter quality.

Description

Technical Field

[0001] The present invention relates to the technical field of entity influence assessment, and in particular to an entity influence assessment system and method based on network data. Background Art

[0002] In many fields, operators often need to evaluate the influence of one or more entities in the corresponding fields. For example, when applying for a university or evaluating university qualifications, it is necessary to know the influence of the university in the overall academic field or a certain academic field; when evaluating hospital specialties, it is necessary to know the influence of the hospital in the corresponding specialty. In addition, personal academic ability, journals, works, etc. also need to be evaluated or promoted, and the corresponding influence needs to be evaluated. By evaluating the influence of these entities, we can find entities with outstanding overall or special capabilities and greater influence, and then formulate corresponding operation strategies for users to achieve the reasonable push of corresponding data.

[0003] At present, when evaluating the influence of a specific object in a field, operators often use market research to obtain data, and perform specific parameter configuration operations based on the obtained data. This method requires a lot of manpower and material resources to collect data, so it is impossible to achieve real-time data updates and automatic acquisition of entity influence; at the same time, the parameters configured in the above method cannot be flexibly adjusted, and the data obtained are all parameters of the order of magnitude, without quality-level parameters, so the quality of the influence parameters finally obtained is not high, resulting in a large discrepancy between the final evaluation results and the actual situation. Therefore, it is of vital significance to study a system that can truly reflect the influence evaluation of a specific entity in a specific field. Summary of the invention

[0004] In view of the shortcomings of the prior art mentioned above, the purpose of the present application is to provide an entity influence evaluation system and method based on network data, which is used to solve the problem that the current entity influence evaluation methods mostly rely on large amounts of data investigation by manpower, and cannot achieve real-time data updating and automatic acquisition of entity influence. At the same time, the existing entity influence evaluation methods cannot achieve flexible adjustment of configuration parameters, and the data obtained are all parameters of the order of magnitude, without quality-level parameters, which affects the quality of the influence parameters finally obtained.

[0005] In a first aspect, the present application provides an entity influence evaluation system based on network data, comprising a central processing unit and a target entity input module, a characteristic parameter acquisition module, an influence index determination module and a diversity index determination module connected to the central processing unit, and also comprising a communication module, wherein the communication module is used to realize the communication between the characteristic parameter acquisition module, the influence index determination module and the diversity index determination module and a network database respectively;

[0006] The target entity input module is used to obtain the target entity and determine and set the characteristic parameter structure based on the target entity;

[0007] The characteristic parameter acquisition module is used to obtain the characteristic parameter group of the target entity from the network database through the communication module based on the set characteristic parameter structure, and use all characteristic parameters in the characteristic parameter group as dimension parameters of the target entity;

[0008] The influence index determination module is used to obtain the influence index of the target entity through a web page ranking algorithm based on the preset reference relationship in the network database, and use the influence index of the target entity as a dimension parameter of the target entity;

[0009] The diversity index determination module is used to obtain the diversity index of the target entity through a preset knowledge complexity acquisition method, and use the diversity index of the target entity as a dimension parameter of the target entity;

[0010] The central processor is used to send the target entity to the influence index determination module and the diversity index determination module, send the set characteristic parameter structure to the characteristic parameter acquisition module, and perform weighted operation on all the dimensional parameters of the target entity to obtain the influence parameter of the target entity;

[0011] Wherein, the network database includes a preset knowledge base and / or a search database in the network.

[0012] In one embodiment of the present application, the set feature parameter structure includes a specific feature parameter type group and an additional feature parameter type group; the specific feature parameter type group includes multiple specific feature parameter types and a specific weight corresponding to each of the specific feature parameter types; the additional feature parameter type group includes multiple additional feature parameter types and an adjustable weight corresponding to each of the additional feature parameter types.

[0013] In one embodiment of the present application, the characteristic parameter acquisition module acquires the characteristic parameter group of the target entity from the network database through the communication module based on the set characteristic parameter structure, including:

[0014] The feature parameter acquisition module captures the structured data of the target entity from the network database through the communication module, obtains the corresponding specific feature parameters from the structured data of the target entity based on each specific feature parameter type in the specific feature parameter type group, and obtains the corresponding additional feature parameters from the structured data of the target entity based on each additional feature parameter type in the additional feature parameter type group, and combines all the specific feature parameters and all the additional feature parameters into a feature parameter group of the target entity.

[0015] In one embodiment of the present application, when the preset reference relationship is an entity reference relationship, it is determined whether the target entity can be a category in the network database. If so, the influence index of the target entity is the sum of the influence indexes of all entities under the target entity when the target entity is a category in the network database. Otherwise, the influence index of the target entity is the influence index of the corresponding entity.

[0016] When the preset citation relationship is a paper citation relationship, the influence index of the target entity is the sum of the influence indexes of all papers under the target entity.

[0017] In one embodiment of the present application, the diversity index determination module obtains the diversity index of the target entity by a preset knowledge complexity acquisition method, including:

[0018] The target entity and the preset entities under the category to which the target entity belongs are all regarded as target knowledge production units, and based on the category to which the target entity belongs, all target knowledge products are obtained and the export method when the target knowledge production unit exports the target knowledge product is identified;

[0019] On the basis of the export mode, based on the export advantages of each of the target knowledge products in each of the target knowledge production units, obtaining the diversity initialization index of each of the target knowledge production units and the scarcity initialization index of each of the target knowledge products;

[0020] Based on the diversity initialization index of each target knowledge production unit and the scarcity initialization index of each target knowledge product, bidirectional iteration of the diversity index of each target knowledge production unit and the scarcity index of each target knowledge product is realized through a nested matrix to obtain the final diversity index of each target knowledge production unit, and the final diversity index of the target knowledge production unit corresponding to the target entity is used as the diversity index of the target entity.

[0021] In one embodiment of the present application, the entity influence assessment system based on network data further includes a sub-domain parameter determination module connected to the central processor, and the sub-domain parameter determination module communicates with the target network knowledge base through the communication module;

[0022] The sub-domain parameter determination module is used to set the target sub-domain category, obtain the semantic distance between the target entity and the target sub-domain category through a preset semantic distance acquisition method, and use the semantic distance between the target entity and the target sub-domain category as the dimensional parameter of the target entity.

[0023] In one embodiment of the present application, the sub-domain parameter determination module acquires the semantic distance between the target entity and the target sub-domain category by a preset semantic distance acquisition method, including:

[0024] Acquire a link set of a target entity from a target network knowledge base through the communication module as an entity link set of the target entity, acquire a link set of the target sub-domain category as an entity from the target network knowledge base through the communication module as an entity link set of the target sub-domain category, and use a semantic distance between the entity link set of the target entity and the entity link set of the target sub-domain category as a first semantic distance between the target entity and the target sub-domain category;

[0025] When the target entity is used as a category in the target network knowledge base, all entity sets under it are used as a temporary set to be ranked; a link set of each entity in the temporary set to be ranked is obtained from the target network knowledge base through the communication module and merged into a category link set of the target entity; all entity sets under the target sub-domain category are used as a sub-domain temporary set; a link set of each entity in the sub-domain temporary set is obtained from the target network knowledge base through the communication module and merged into a category link set of the target sub-domain category; and a semantic distance between the category link set of the target entity and the category link set of the target sub-domain category is used as a second semantic distance between the target entity and the target sub-domain category;

[0026] The first semantic distance between the target entity and the target sub-domain category, the second semantic distance between the target entity and the target sub-domain category, or the weighted average of the first semantic distance and the second semantic distance between the target entity and the target sub-domain category is used as the semantic distance between the target entity and the target sub-domain category.

[0027] In one embodiment of the present application, the central processing unit performs a linear weighted operation or a nonlinear weighted operation on all the dimensional parameters of the target entity to obtain the influence parameter of the target entity.

[0028] In one embodiment of the present application, the entity influence assessment system based on network data also includes a data storage module connected to the central processing unit, and the data storage module is used to store the characteristic parameter group of the target entity, the influence index of the target entity, the diversity index of the target entity and the influence parameter of the target entity.

[0029] In a second aspect, the present application provides an entity influence assessment method based on network data, comprising:

[0030] Acquire a target entity, and determine and set a characteristic parameter structure based on the target entity;

[0031] Acquire a characteristic parameter group of the target entity from the network database based on the set characteristic parameter structure, and use all characteristic parameters in the characteristic parameter group as dimension parameters of the target entity;

[0032] Based on the preset reference relationship in the network database, the influence index of the target entity is obtained through a web page ranking algorithm, and the influence index of the target entity is used as a dimension parameter of the target entity;

[0033] Acquire the diversity index of the target entity by a preset knowledge complexity acquisition method, and use the diversity index of the target entity as a dimension parameter of the target entity;

[0034] Performing a weighted operation on all the dimension parameters of the target entity to obtain an influence parameter of the target entity;

[0035] Wherein, the network database includes a preset knowledge base and / or a search database in the network.

[0036] Compared with the prior art, one or more embodiments of the above scheme may have the following advantages or beneficial effects:

[0037] The entity influence evaluation system based on network data provided by the embodiment of the present invention is applied. The characteristic parameter acquisition module captures structured data from the network database, and the characteristic parameter group of each target entity is screened out from the obtained structured data based on the corresponding set characteristic parameter structure to achieve the acquisition of each entity characteristic parameter; the influence index determination module provides quality-level parameters for the influence of the target entity based on the citation relationship of the entity or paper in the network database, and the influence index obtained by the web page ranking algorithm; at the same time, the diversity index obtained by the diversity index determination module through the preset knowledge complexity acquisition method also provides quality-level parameters for the influence of the target entity; the central processing unit realizes flexible configuration of multi-dimensional parameters by setting the adjustable weights of each dimension parameter, so that the user can obtain the influence parameters of the whole or a single field based on actual needs. The sub-field parameter determination module further provides quality-level parameters for the influence of the target entity by obtaining the semantic distance between the target entity and the target sub-field category, and the participation of the above semantic distance realizes the acquisition of the influence of the target entity in the target sub-field category. The system of the present invention provides a high-quality entity influence acquisition system that can be fully automated and updated in real time, with adjustable multi-dimensional parameters, many participating parameters and high parameter quality.

[0038] Other features and advantages of the present invention will be described in the following description, and partly become apparent from the description, or understood by practicing the present invention. The purpose and other advantages of the present invention can be realized and obtained by the structures particularly pointed out in the description, claims and drawings. BRIEF DESCRIPTION OF THE DRAWINGS

[0039] The accompanying drawings are used to provide a further understanding of the present invention and constitute a part of the specification. Together with the embodiments of the present invention, they are used to explain the present invention and do not constitute a limitation of the present invention. In the accompanying drawings:

[0040] Figure 1 Shown is a structural diagram of the entity influence evaluation system based on network data described in an embodiment of the present application.

[0041] Figure 2 Shown is a flow chart of the entity influence assessment method based on network data described in an embodiment of the present application. DETAILED DESCRIPTION

[0042] The following will describe the implementation methods of the present invention in detail with reference to the accompanying drawings and embodiments, so that the implementation process of how the present invention applies technical means to solve technical problems and achieve technical effects can be fully understood and implemented accordingly. It should be noted that as long as there is no conflict, the various embodiments of the present invention and the various features in the embodiments can be combined with each other, and the technical solutions formed are all within the protection scope of the present invention.

[0043] It should be noted that the illustrations provided in the following embodiments are only schematic illustrations of the basic concept of the present application, and thus the drawings only show components related to the present application rather than being drawn according to the number, shape and size of components in actual implementation. In actual implementation, the type, quantity and proportion of each component may be changed at will, and the component layout may also be more complicated.

[0044] An entity refers to independent things in the real world that can be distinguished from each other; an entity can be a concrete thing, such as a person, a place or an object, or an abstract concept or event; in the construction and application of a knowledge base, an entity is the most basic component unit used to represent and describe various things and concepts in the real world. A knowledge base usually contains relevant information and data in a certain field or subject, with corresponding rules, methods, concepts and other forms of knowledge representation; it can be classified, organized and retrieved according to a certain structure and semantics, thereby improving the comprehensibility, reusability and interoperability of knowledge. A category is a basic element in a knowledge structure or data structure, which is used to define a group of objects with the same characteristics or attributes; for example, humans, animals, and plants all belong to a category. The core idea of ​​the PageRank algorithm is to measure the importance of a web page by the number and quality of other web pages linked to it.

[0045] The following embodiments of the present application provide an entity influence assessment system and method based on network data, which solves the problem that the current entity influence assessment methods mostly rely on a large amount of data investigation by manpower, cannot achieve real-time data updating and automatic acquisition of entity influence, and at the same time, the existing entity influence assessment methods cannot achieve flexible adjustment of configuration parameters, and the data obtained are all parameters of order of magnitude, without quality-level parameters, which affects the quality of the influence parameters finally obtained.

[0046] The following will describe in detail the principles and implementation methods of a system and method for evaluating entity influence based on network data of this embodiment in conjunction with the accompanying drawings, so that those skilled in the art can understand the system and method for evaluating entity influence based on network data of this embodiment without creative work.

[0047] The entity influence evaluation system based on network data in this application can be applied to various entity influence evaluation scenarios, such as universities, middle schools, personal academics, works, hospitals, movies, or certain types of consumer goods (such as beer), etc. This entity influence evaluation system based on network data integrates the order of magnitude dimension parameters and the quality dimension parameters in the form of adjustable parameter weights to improve the quality of the acquired entity influence.

[0048] like Figure 1As shown, this embodiment provides an entity influence evaluation system based on network data, including a central processing unit and a target entity input module, a feature parameter acquisition module, an influence index determination module and a diversity index determination module connected to the central processing unit, and also includes a communication module that enables the feature parameter acquisition module, the influence index determination module and the diversity index determination module to communicate with the network database respectively.

[0049] The target entity input module is used to obtain the target entity and determine the set characteristic parameter structure based on the target entity. The specific user can input the target entity whose influence is to be evaluated through this module, and the target entity must be an entity in the network database itself to avoid failing to obtain data such as the characteristic parameters of the target entity in the network database.

[0050] The characteristic parameter acquisition module is used to obtain the characteristic parameter group of the target entity from the network database through the communication module based on the set characteristic parameter structure, and use all characteristic parameters in the characteristic parameter group as the dimension parameters of the target entity.

[0051] The network database may include a preset knowledge base pre-set in the network, and may also include a search database corresponding to a preset search engine pre-set in the network. When the target entity fails to capture the specific feature parameters contained in the specific feature parameter type group of the set feature parameter structure in the preset knowledge base, in order to ensure that the target entity can obtain all feature parameters in the feature parameter group, the preset search engine may be used to search for the corresponding specific feature parameters in the search database to complete the search. Furthermore, the preset knowledge base may also be a knowledge base as a whole composed of multiple knowledge bases, and the search database may also be a search database as a whole composed of multiple search databases, and the staff may set it as needed.

[0052] The influence index determination module is used to obtain the influence index of the target entity through a web page ranking algorithm based on the preset reference relationship in the network database, and use the influence index of the target entity as the dimension parameter of the target entity.

[0053] The diversity index determination module is used to obtain the diversity index of the target entity through a preset knowledge complexity acquisition method, and use the diversity index of the target entity as a dimensional parameter of the target entity.

[0054] The central processing unit is used to send the target entity to the influence index determination module and the diversity index determination module, send the set feature parameter structure to the feature parameter acquisition module, and perform weighted operations on all dimensional parameters of the target entity to obtain the influence parameter of the target entity.

[0055] The network database includes a preset knowledge base and / or a search database in the network.

[0056] The entity influence evaluation system based on network data provided by the embodiment of the present invention captures structured data from a network database through a feature parameter acquisition module, and screens out a feature parameter group of each target entity from the acquired structured data based on the corresponding set feature parameter structure, so as to obtain the feature parameters of each entity; the influence index determination module provides quality-level parameters for the influence of the target entity based on the influence index obtained by the web page ranking algorithm based on the citation relationship of the entity or paper in the network database; at the same time, the diversity index obtained by the diversity index determination module through the preset knowledge complexity acquisition method also provides quality-level parameters for the influence of the target entity; the central processing unit realizes flexible configuration of multi-dimensional parameters by adjustable setting of the weights of each dimensional parameter, so that users can obtain the influence parameters of the whole or a single field based on actual needs.

[0057] The target entity input module, characteristic parameter acquisition module, influence index determination module, diversity index determination module and central processing unit in the entity influence evaluation system based on network data of this embodiment are described in detail below.

[0058] The target entity input module is used to obtain the target entity and determine and set the feature parameter structure based on the target entity.

[0059] The target entity input module can be configured with a keyboard or a touch screen to enable the user to input the target entity. At the same time, the target entity input module is also configured to store a characteristic parameter structure table, which includes characteristic parameter structures corresponding to multiple types of entities. After the target entity is input into the target entity input module, the target entity input module can determine the characteristic parameter structure corresponding to the target entity based on the category to which the target entity belongs (i.e., entities with a common category have the same characteristic parameter structure), and set the characteristic parameter structure to a set characteristic parameter structure.

[0060] It should be noted that the feature parameter structures corresponding to various entities in the feature parameter structure table are set in advance by the staff based on actual conditions. The specific feature parameter structure includes a specific feature parameter type group and an additional feature parameter type group for the target entity.

[0061] The specific feature parameter type group includes all the specific feature parameter types that the target entity is to collect by default and the specific weights corresponding to each specific feature parameter type. For example, if the target entity is a university, the total number of university students, the faculty, the number of paper exports, the total number of graduate students and doctoral students, the number of various awards won, etc. are all specific feature parameter types that the target entity must consider by default when acquiring influence. At this time, the above-mentioned various feature parameter types can be set in the specific feature parameter type group. At the same time, in order to improve the flexibility of the specific feature parameter type group, the user can also make specific modifications to the specific weights corresponding to each specific feature parameter type in the specific feature parameter type group. The additional feature parameter type group includes multiple additional feature parameter types that may be collected by the target entity and the adjustable weight corresponding to each additional feature parameter type, wherein the additional feature parameter type is not the default feature parameter type for determining the target entity, but it will also have an impact on the influence parameter of the target entity. The staff can set the adjustable weight corresponding to the additional feature parameter type to adjust the additional feature parameter types included in the additional feature parameter type group and the proportion of each additional feature parameter type when the target entity obtains the influence parameter. This adjustment allows the staff to obtain the influence parameters of the target entity with multiple focuses. For example, if the target entity is a university, whether the university's founding year is involved in the target entity's influence parameter evaluation can be determined by setting the adjustable weight of the university's founding year as an additional feature parameter type, and the proportion of the university's founding year in the target entity's influence parameter evaluation can be determined by setting the adjustable weight value of the university's founding year.

[0062] Accordingly, for different categories to which the target entity belongs, there are pre-set and corresponding specific feature parameter type groups, and the staff can make corresponding settings before conducting the corresponding target entity influence parameter evaluation.

[0063] When determining the category to which the target entity belongs, the category to which the target entity belongs may be determined in a network database, or in a set database or knowledge base, and no further limitation is made here.

[0064] After the target entity input module obtains the target entity and sets the characteristic parameter structure, it can send the target entity and the set characteristic parameter structure to the central processing unit. The central processing unit sends the target entity to the influence index determination module and the diversity index determination module respectively, and sends the set characteristic parameter structure to the characteristic parameter acquisition module.

[0065] The characteristic parameter acquisition module is used to obtain the characteristic parameter group of the target entity from the network database through the communication module based on the set characteristic parameter structure, and use all characteristic parameters in the characteristic parameter group as the dimension parameters of the target entity.

[0066] The network database may include a preset knowledge base pre-set in the network, and may also include a search database corresponding to a preset search engine pre-set in the network. When the target entity fails to capture the specific feature parameter type contained in the specific feature parameter type group of the set feature parameter structure or the additional feature parameter type corresponding to the adjustment weight assigned with the corresponding value in the preset knowledge base, in order to ensure that the target entity can obtain all feature parameters in the feature parameter group, the preset search engine can be used to find the corresponding specific feature parameters in the search database for completion. Further, the preset knowledge base can be a knowledge base composed of multiple knowledge bases as a whole (for example, when the target entity is a university, the preset knowledge base can be set to include a paper knowledge base, a patent knowledge base, etc., and when the target entity is a certain type of consumer product, the preset knowledge base can be set to include an industrial and commercial knowledge database, etc.), and the search database can also be a database as a whole composed of multiple search databases, which can be set by the staff as needed. It should be noted that all kinds of knowledge bases and search databases in the network database are databases that can update data in real time or periodically.

[0067] After receiving the set characteristic parameter structure, the characteristic parameter acquisition module can obtain the characteristic parameter group of the target entity from the network database through the communication module based on the set characteristic parameter structure. The characteristic parameter acquisition module obtains the characteristic parameter group of the target entity from the network database based on the set characteristic parameter structure, specifically including: firstly, all structured data of the target entity can be captured from the network database, for example, all structured data of the target entity can be captured in the infobox of the target entity page corresponding to the network database (i.e., field attributes containing digital types), or all structured data of the target entity can be captured from the text information of the target entity page (i.e., field attributes containing digital types).

[0068] For example, the following information can be obtained from the corresponding "University A" page in the online database:

[0069] Table 1

[0070]

[0071] Usually, the feature parameter types in the captured structured data are often more than the feature parameter types in the specific feature parameter type group. At this time, the corresponding specific feature parameters can be obtained from the structured data of the target entity based on each specific feature parameter type in the specific feature parameter type group, and the corresponding additional feature parameters can be obtained from the structured data of the target entity based on each additional feature parameter type in the additional feature parameter type group. Then, all the specific feature parameters and all the additional feature parameters are combined to obtain the feature parameter group of the target entity. Finally, all the feature parameters in the feature parameter group of the target entity are used as the dimension parameters of the target entity.

[0072] In one embodiment of the present invention, if one wants to further implement the influence evaluation of the target entity in certain sub-fields, one can also control the feature parameter types in the specific feature parameter type group and the additional feature parameter type group to enable the target entity to only obtain feature parameters related to a certain sub-field.

[0073] The influence index determination module is used to obtain the influence index of the target entity through a web page ranking algorithm based on the preset reference relationship in the network database, and use the influence index of the target entity as the dimension parameter of the target entity.

[0074] The preset reference relationship may be a reference relationship of an entity, and specifically, a hyperlink of the entity in a corresponding page of a network database may be used to represent the reference relationship of the entity. At the same time, the preset reference relationship may also be a reference relationship of a paper, that is, the paper link cited in the paper is used as the reference relationship of the paper.

[0075] On the basis of the preset reference relationship as the reference relationship of the entity, the specific process of obtaining the influence index of the target entity through the web page ranking algorithm based on the preset reference relationship in the network database is: obtain the reference relationship of all entities in the network database, and then obtain the influence index of each entity through the web page ranking algorithm based on the reference relationship of all entities; then judge whether the target entity can be used as a category in the network database. If so, when the target entity is used as a category in the network database, the sum of the influence indexes of all entities under it is used as the influence index of the target entity, otherwise the influence index of the entity corresponding to the target entity is directly used as the influence index of the target entity.

[0076] On the basis of presetting the citation relationship as the citation relationship of the paper, the specific process of obtaining the influence index of the target entity through the web page ranking algorithm based on the preset citation relationship in the network database is as follows: obtain the citation relationship of all papers in the network database, and then obtain the influence index of each paper through the web page ranking algorithm based on the citation relationship of all papers; then take the sum of the influence indexes of all papers under the target entity as the influence index of the target entity.

[0077] Taking the citation relationship of the paper as an example, the influence index of each paper obtained by the web page ranking algorithm is described in detail. Specifically, it is assumed that there are n papers P1-Pn in the network database, and the initial value of each paper is set to 1 / n, and the difference range between the previous and next iterations is d. On this basis, it is assumed that the papers with paper links pointing to P1 are P2 and P3, of which P2 has x paper links and P3 has y paper links. Then the temporary influence index of P1 in the first iteration = 1 / n*1 / x+1 / n*1 / y. And so on, the same iteration is performed for all n papers. The initial value of the paper in the second iteration is the temporary influence index obtained in the first iteration. For example, the temporary influence index of P2 in the first iteration = x1, and the temporary influence index of P3 = y1, then the value of P1 in the second iteration = 1 / x1*1 / x+1 / y1*1 / y. Repeat the above iterative process until the difference between the temporary influence index of each paper in two iterations is within the set range d, and then it is considered to have converged. At this time, the temporary influence index of each paper is the influence index of the corresponding paper.

[0078] When the preset citation relationship is the citation relationship of the entity, the process of obtaining the influence index of each paper through the web page ranking algorithm is the same as the above process, which will not be described in detail here. Finally, the influence index of the target entity is also used as the dimension parameter of the target entity.

[0079] The influence index determination module obtains various links and other data from the network database through the communication module.

[0080] The diversity index determination module is used to obtain the diversity index of the target entity through a preset knowledge complexity acquisition method, and use the diversity index of the target entity as a dimensional parameter of the target entity.

[0081] First, the target entity and the preset entities under the category to which the target entity belongs are all taken as target knowledge production units, wherein the preset entities under the category to which the target entity belongs can be all entities under the category to which the target entity belongs, or can be random entities of a random number under the category to which the target entity belongs, or can be entities of other set forms under the category to which the target entity belongs, and this embodiment does not make fixed settings for this. At the same time, it is also necessary to obtain all target knowledge products based on the category to which the target entity belongs and to identify the export method when the target knowledge production unit exports the target knowledge products. Among them, for different categories to which the target entity belongs, there are multiple target knowledge products that are pre-set and corresponding to them and the export method when the target knowledge production unit exports the target knowledge products. For example, when the category to which the target entity belongs is a university, its corresponding target knowledge products can be set to include "mathematics", "English", "chemistry", "computer", "economics", "biology", etc., and the export method when the target knowledge production unit exports the target knowledge product can be set to be a paper. When the category to which the target entity belongs is other categories, there are also multiple target knowledge products corresponding to it and the export method when the target knowledge production unit exports the target knowledge product.

[0082] For different categories to which the target entity belongs, it is also possible to pre-set other different target knowledge products and the export method when the target knowledge production unit exports the target knowledge product. This embodiment does not impose a fixed limit on it. The knowledge product is an intellectual product such as a paper, patent, award, etc. The export method is usually the basis for the target knowledge unit to export the target knowledge product.

[0083] Secondly, on the basis of the export mode, based on the export advantages of each target knowledge product in each target knowledge production unit, the diversity initialization index of each target knowledge production unit and the scarcity initialization index of each target knowledge product are obtained.

[0084] Specifically, the comparison knowledge production unit is set as any one of all target knowledge production units, and the comparison knowledge product is set as any one of all target knowledge products. At the same time, based on the export method, the export volume of the comparison knowledge product exported by the comparison knowledge production unit is taken as the individual export volume, the export volume of all target knowledge products exported by the comparison knowledge production unit is taken as the individual total export volume, the export volume of the comparison knowledge product exported by all target knowledge production units is taken as the overall export volume, and the export volume of all target knowledge products exported by all target knowledge production units is taken as the overall total export volume. At this time, it can be judged whether the ratio of the individual export volume to the individual total export volume is greater than the ratio of the overall export volume to the overall total export volume. If so, it is determined that the comparison knowledge production unit has an export advantage when exporting the comparison knowledge product. Otherwise, it is determined that the comparison knowledge production unit does not have an export advantage when exporting the comparison knowledge product.

[0085] For example, assume that all target knowledge production units include target knowledge production unit A1, target knowledge production unit A2, target knowledge production unit A3, target knowledge production unit A4 and target knowledge production unit A5, and all target knowledge products include target knowledge product B1, target knowledge product B2, target knowledge product B3, target knowledge product B4 and target knowledge product B5. When the target knowledge production unit A1 is set as the comparison production unit and the target knowledge product B1 is set as the comparison knowledge product, the individual export volume of the comparison knowledge product B1 exported by the comparison knowledge production unit A1 is D1, the individual total export volume of all target knowledge products exported by the comparison knowledge production unit A1 is D2 (i.e. the total export volume of the comparison knowledge product B1, target knowledge product B2, target knowledge product B3, target knowledge product B4 and target knowledge product B5 exported by the comparison knowledge production unit A1 respectively), and the overall export volume of the comparison knowledge product B1 exported by all target knowledge production units is Z1 (i.e. the sum of the export volume of the comparison knowledge product B1 exported by the comparison knowledge production unit A1, the export volume of the comparison knowledge product B1 exported by the target knowledge production unit A2, the export volume of the comparison knowledge product B1 exported by the target knowledge production unit A3, the export volume of the comparison knowledge product B1 exported by the target knowledge production unit A4 and the export volume of the comparison knowledge product B1 exported by the target knowledge production unit A5). The overall total export volume of all target knowledge products exported by the units is Z2 (i.e. the export volume of comparative knowledge product B1, target knowledge product B2, target knowledge product B3, target knowledge product B4 and target knowledge product B5 exported by comparative knowledge production unit A1; the export volume of comparative knowledge product B1, target knowledge product B2, target knowledge product B3, target knowledge product B4 and target knowledge product B5 exported by target knowledge production unit A2; the export volume of comparative knowledge product B1, target knowledge product B2, target knowledge product B3, target knowledge product B4 and target knowledge product B5 exported by target knowledge production unit A3; the export volume of comparative knowledge product B1, target knowledge product B2, target knowledge product B3, target knowledge product B4 and target knowledge product B5 exported by target knowledge production unit A4; the export volume of comparative knowledge product B1, target knowledge product B2, target knowledge product B3, target knowledge product B4 and target knowledge product B5 exported by target knowledge production unit A5).

[0086] Furthermore, determine whether the ratio of the individual export volume D1 to the individual total export volume D2 is greater than the ratio of the overall export volume Z1 to the overall total export volume Z2, that is, whether D1 / D2 is greater than Z1 / Z2. If so, it is determined that the comparative knowledge production unit A1 has an export advantage when exporting the comparative knowledge product B1; otherwise, it is determined that the comparative knowledge production unit A1 does not have an export advantage when exporting the comparative knowledge product B1.

[0087] Based on the above method, it can be determined whether each target knowledge product has an export advantage in each target knowledge production unit, and based on the export advantage of each target knowledge product in each target knowledge production unit, the diversity initialization index of each target knowledge production unit and the scarcity initialization index of each target knowledge product can be obtained.

[0088] On the basis of setting the comparison knowledge production unit as any one of all the target knowledge production units, obtaining the diversity initialization index of the comparison knowledge production unit based on the export advantage of each target knowledge product in the comparison knowledge production unit specifically includes: first, sorting all the target knowledge products to obtain a knowledge product sequence; then setting a value N with an initial value of 1, judging whether the current N is less than or equal to the preset product threshold, if the current N is less than or equal to the preset product threshold, it means that all the target knowledge products have not been completely screened, and at this time it is necessary to judge whether the comparison knowledge production unit has an export advantage when exporting the Nth target knowledge product in the knowledge product sequence, if so, adding 1 to the diversity temporary index of the current comparison knowledge production unit, otherwise there is no need to modify the diversity temporary index of the current comparison knowledge production unit; then adding 1 to the current N to obtain a new N, and returning to the above process of judging whether the current N is less than or equal to the preset product threshold, that is, re-judging the current N value and the subsequent corresponding steps. If, when judging whether the current N is less than or equal to the preset product threshold, it is determined that the current N is greater than the preset product threshold, it means that the screening of all target knowledge products has been completed. At this time, the temporary diversity index of the current comparison knowledge production unit can be directly used as the diversity initialization index of the comparison knowledge production unit.

[0089] It should be noted that the above-mentioned preset product threshold is the number of target knowledge products in the knowledge product sequence, while the initial value of the temporary index of diversity of the comparison knowledge production unit is 0.

[0090] Through the above method, the diversity initialization index of each target knowledge production unit can be obtained based on the export advantage of each target knowledge product in each target knowledge production unit.

[0091] On the basis of setting the comparison knowledge product to be any one of all the target knowledge products, based on the export advantages of the comparison knowledge product in each target knowledge production unit, obtaining the scarcity initialization index of the comparison knowledge product specifically includes: first, sorting all the target knowledge production units to obtain a knowledge production unit sequence; then setting a value M with an initial value of 1, judging whether the current M is less than or equal to the preset production unit threshold, if the current M is less than or equal to the preset production unit threshold, it means that all the target knowledge production units have not been completely screened, and at this time it is necessary to judge whether the Mth target knowledge production unit in the knowledge production unit sequence has an export advantage when exporting the comparison knowledge product, if so, adding 1 to the scarcity temporary index of the current comparison knowledge product, otherwise there is no need to modify the scarcity temporary index of the current comparison knowledge product; then adding 1 to the current M to obtain a new M, and returning to the above process of judging whether the current M is less than or equal to the preset production unit threshold, that is, re-judging the current M value and the subsequent corresponding steps. If, when judging whether the current M is less than or equal to the preset production unit threshold, it is determined that the current M is greater than the preset production unit threshold, it means that the screening of all target knowledge production units has been completed. At this time, the temporary scarcity indicator of the current comparison knowledge product can be directly used as the scarcity initialization indicator of the comparison knowledge product.

[0092] It should be noted that the above-mentioned preset production unit threshold is the number of target knowledge production units in the knowledge production unit sequence, and the initial value of the temporary indicator of scarcity of the comparison knowledge product is 0.

[0093] Through the above method, the scarcity initialization index of each target knowledge product can be obtained based on the export advantage of each target knowledge product in each target knowledge production unit.

[0094] Finally, based on the diversity initialization index of each target knowledge production unit and the scarcity initialization index of each target knowledge product, the bidirectional iteration of the diversity index of each target knowledge production unit and the scarcity index of each target knowledge product is realized through the nested matrix to obtain the final diversity index of each target knowledge production unit.

[0095] After obtaining the diversity initialization index of each target knowledge production unit and the scarcity initialization index of each target knowledge product, the two-way iteration of the diversity index of each target knowledge production unit and the scarcity index of each target knowledge product can be achieved through nested matrices on this basis.

[0096] Furthermore, in the above-mentioned bidirectional iteration process, each target knowledge production unit diversity index and each target knowledge product scarcity index are bidirectionally iterated until a preset iteration stop condition is met.

[0097] In the above process, on the basis of setting the comparison knowledge production unit as any one of all the target knowledge production units, the single iterative calculation process of the diversity index of the comparison knowledge production unit realized by the nested matrix specifically includes: multiplying the scarcity index of each target knowledge product obtained in the previous iteration by its production index relative to the comparison knowledge production unit as the unit production value, and then summing all the unit production values ​​and dividing them by the diversity initialization index of the comparison knowledge production unit to obtain the diversity index of the comparison knowledge production unit in the current iteration. When the comparison knowledge production unit has an export advantage in exporting the target knowledge product, the production index of the target knowledge product relative to the comparison knowledge production unit is set to 1, otherwise the production index of the target knowledge product relative to the comparison knowledge production unit is set to 0. Furthermore, because the nature of the nested matrix iteration must converge, the above single iterative calculation process of the diversity index of the comparison knowledge production unit realized by the nested matrix is ​​actually the mean of the scarcity of all target knowledge products with export advantages in the previous iteration of the comparison knowledge production unit.

[0098] Repeat the above process to obtain the diversity indicators of all target knowledge production units in the current iteration.

[0099] Similarly, in the above process, on the basis of setting the comparative knowledge product as any one of all the target knowledge products, the single iterative calculation process of the scarcity index of the comparative knowledge product realized by the nested matrix specifically includes: multiplying the diversity index of each target knowledge production unit obtained in the previous iteration by its production index relative to the comparative knowledge product as the product production value, and then summing all product production values ​​and dividing them by the scarcity initialization index of the comparative knowledge product to obtain the scarcity index of the comparative knowledge product in the current iteration. When the target knowledge production unit has an export advantage in exporting the comparative knowledge product, the production index of the target knowledge production unit relative to the comparative knowledge product is set to 1, otherwise the production index of the target knowledge production unit relative to the comparative knowledge product is set to 0. Furthermore, because the nature of the nested matrix iteration must converge, the above single iterative calculation process of the scarcity index of the comparative knowledge product realized by the nested matrix is ​​actually the mean of the diversity of all knowledge production units whose comparative knowledge products can be exported with advantage in the previous iteration.

[0100] Repeat the above process to obtain the scarcity index of all target knowledge products in the current iteration.

[0101] Then, through the above process, multiple bidirectional iterations of the diversity index of each target knowledge production unit and the scarcity index of each target knowledge product are achieved (i.e., multiple bidirectional iterations are achieved), until the preset iteration stop condition is met and the iteration is stopped, and finally, the final diversity index of each target knowledge production unit is obtained based on the final even-numbered iteration results of the diversity index, and the final scarcity index of each target knowledge product is obtained based on the final odd-numbered iteration results of the scarcity index. The final diversity index of the target knowledge production unit corresponding to the target entity is used as the diversity index of the target entity, and the diversity index of the target entity is also used as the dimension parameter of the target entity.

[0102] In one embodiment of the present invention, the preset iteration stop condition can be set to the number of iterations reaching the preset number of iterations, or can be set to the diversity indicators of all target knowledge production units obtained converge, and the scarcity indicators of all target knowledge products obtained converge. The preset iteration stop condition can also be set to other reasonable conditions, which are not fixed here.

[0103] The diversity index determination module obtains data such as the preset entities under the category to which the target entity belongs and the export volume of the target knowledge products exported by each target knowledge production unit from the network database through the communication module.

[0104] The central processing unit is used to perform weighted operations on all dimensional parameters of the target entity to obtain the influence parameter of the target entity.

[0105] All dimension parameters obtained by the above target entity are weighted to obtain the influence parameter of the target entity. The influence parameter of the target entity can be obtained by linear weighted operation. For example, the influence parameter of the target entity can be expressed as:

[0106] score=aX1+bX2+…

[0107] Among them, score represents the influence parameter of the target entity, X1, X2... represent the parameters of each dimension of the target entity, and a, b... represent the weights of the parameters of each dimension of the target entity. Specifically, the sum of the weights of the parameters of each dimension of all target entities can be set to 1, or the sum of the weights of the parameters of each dimension of all target entities can be set not to 1.

[0108] The influence parameter of the target entity can also be obtained through nonlinear weighted operation. For example, the influence parameter can be expressed as:

[0109] score=aln(X1)+b(X2)^n+…

[0110] Among them, score represents the influence parameter of the target entity, X1, X2... represent the parameters of each dimension of the target entity, and the parameters of each dimension of the target entity are not linear in the above formula, and the corresponding complex operations such as log, multiple powers or square roots can be taken as needed. a, b... represent the weights of the parameters of each dimension of the target entity. Specifically, the sum of the weights of all the parameters of each dimension of all the target entities can be set to 1, or the sum of the weights of all the parameters of each dimension of all the target entities can be set not to 1.

[0111] The weights of the various dimensional parameters of the above-mentioned target entity, in addition to the specific weights of the specific feature parameters corresponding to each specific feature parameter type being obtained from the set feature parameter structure, and the adjustable weights of the additional feature parameters corresponding to each additional feature parameter type being obtained from the set feature parameter structure, the weights of the influence index of the target entity and the weights corresponding to the diversity index of the target entity can also be adjusted and set based on actual conditions. For example, since the influence index of the target entity and the diversity index of the target entity are both dimensional parameters representing the quality level, the corresponding weights of the two can be appropriately set to larger values.

[0112] At the same time, all dimension parameters can be standardized first to limit each dimension parameter to a specific range, and then the standardized parameter value is brought into the influence parameter formula for calculation. For example, the standardized calculation formula for the dimension parameter Ai of the target entity Wi can be set as:

[0113] Ai=RankAi*Up / RankA-max,

[0114] Where Up is the upper limit of the specific interval set for the dimension parameter type A, RankAi is the original dimension value of the dimension parameter Ai of the target entity Wi, and RankA-max is the maximum original dimension value of the dimension parameter type A for all target entities. If the distribution of all dimension parameters of a certain type after standardization is still uneven, the dimension parameters can be subjected to complex operations such as log, multiple powers or square roots before standardization.

[0115] The influence parameters of the target entity obtained above can also be used to promote the influence assessment based on geographical areas. The specific geographical coordinates can be extracted from map data sources. The influence parameters of the target entity can also be integrated with various map tools, which will not be described in detail here.

[0116] In one embodiment of the present invention, if one wants to further evaluate the influence parameters of certain sub-fields based on the existing target entity influence parameters, this embodiment can also add new dimensional parameters in the influence parameter evaluation process through the sub-field parameter determination module.

[0117] The entity influence evaluation system based on network data is specifically configured to also include a sub-domain parameter determination module connected to the central processor, and the sub-domain parameter determination module communicates with the target network knowledge base through the communication module.

[0118] The subdomain parameter determination module is used to set the target subdomain category, obtain the semantic distance between the target entity and the target subdomain category through a preset semantic distance acquisition method, and use the semantic distance between the target entity and the target subdomain category as the dimensional parameter of the target entity.

[0119] The target sub-domain category is set, and the target sub-domain category is the sub-domain in which the target entity is expected to achieve influence assessment in the field. The target network knowledge base can be pre-set based on actual needs, for example, it can be a certain type of data knowledge base, or a certain type of paper knowledge base, etc. The semantic distance between the target entity and the target sub-domain category is obtained by the preset semantic distance acquisition method, which specifically includes: obtaining the link set of the target entity from the target network knowledge base as the entity link set of the target entity, obtaining the link set of the target sub-domain category as an entity from the target network knowledge base as the entity link set of the target sub-domain category, and finally obtaining the semantic distance between the entity link set of the target entity and the entity link set of the target sub-domain category through the Jaccard formula or other semantic distance calculation formulas between sets, and using it as the first semantic distance between the target entity and the target sub-domain category. For example, assuming that the target sub-domain category is Computer science and University A is the target entity, it is necessary to query the entity link sets of Computer science and University A respectively in the target network knowledge base, and then obtain the semantic distance between the real link set of University A and the real link set of Computer science through the Jaccard formula or other semantic distance calculation formulas between sets as the first semantic distance between University A and Computer science.

[0120] At the same time, when the target entity is used as a category in the target network knowledge base, all entities under it are grouped as a temporary set to be ranked, and then the link set of each entity in the temporary set to be ranked is obtained from the target network knowledge base, and the link sets of all entities in the temporary set to be ranked are merged into the category link set of the target entity; and all entities under the target sub-domain category are grouped as a sub-domain temporary set, and the link set of each entity in the sub-domain temporary set is obtained from the target network knowledge base, and the link sets of all entities in the sub-domain temporary set are merged into the category link set of the target sub-domain category; finally, the semantic distance between the category link set of the target entity and the category link set of the target sub-domain category is obtained through the Jaccard formula or other semantic distance calculation formulas between sets, and it is used as the second semantic distance between the target entity and the target sub-domain category. For example, assuming that the target sub-domain category is Computer science and University A is the target entity, it is necessary to query the target network knowledge base for all entity sets under the category of University A as a temporary set to be ranked, and then obtain the link set of each entity in the temporary set to be ranked from the target network knowledge base, and merge the link sets of all entities in the temporary set to be ranked into the category link set of University A; at the same time, it is necessary to query the target network knowledge base for all entity sets under the category of Computer science as a temporary sub-domain set, and then obtain the link set of each entity in the temporary sub-domain set from the target network knowledge base, and merge the link sets of all entities in the temporary sub-domain set into the category link set of Computer science; finally, the semantic distance between the category link set of University A and the category link set of Computer science is obtained through the Jaccard formula or other semantic distance calculation formulas between sets, and it is used as the second semantic distance between University A and Computer science.

[0121] The link set of the target entity includes multiple links under the target entity, which can be a hyperlink in the corresponding page of the entity in the target network knowledge base, or a link to the paper corresponding to the entity in the target network knowledge base.

[0122] On the basis of the above, based on actual needs, only the first semantic distance between the target entity and the target sub-domain category can be used as the semantic distance between the target entity and the target sub-domain category. In this case, only the first semantic distance between the target entity and the target sub-domain category is obtained. Based on actual needs, only the second semantic distance between the target entity and the target sub-domain category can be used as the semantic distance between the target entity and the target sub-domain category. In this case, only the second semantic distance between the target entity and the target sub-domain category is obtained. Furthermore, based on actual needs, the weighted average of the first semantic distance and the second semantic distance between the target entity and the target sub-domain category can be used as the semantic distance between the target entity and the target sub-domain category. For example, when "University A" is closer to the semantic distance of "Computer science" than "University B", then it is determined that "University A" has a higher score in the semantic distance parameter.

[0123] The semantic distance between the target entity and the target sub-domain category can be obtained in the above manner, and the semantic distance between the target entity and the target sub-domain category is also used as a dimension parameter of the target entity.

[0124] At the same time, the central processing unit sets a corresponding weight for the semantic distance between the target entity and the target sub-domain category, so that it can participate in the influence evaluation calculation of the target entity as a dimensional parameter.

[0125] The entity influence assessment system based on network data also includes a data storage module connected to the central processing unit, and the data storage module is used to store the characteristic parameter group of the target entity, the influence index of the target entity, the diversity index of the target entity and the influence parameter of the target entity. When the semantic distance between the target entity and the target sub-domain category is used, the semantic distance between the target entity and the target sub-domain category also needs to be stored in the data storage module. The above-mentioned stored data can facilitate subsequent data query, and can also be used in the subsequent influence assessment process of other sub-domains or focuses by adjusting the weights of the parameters of each dimension.

[0126] The entity influence evaluation system based on network data provided by the embodiment of the present invention captures structured data from the network database through the feature parameter acquisition module, and selects the feature parameter group of each target entity from the acquired structured data based on the corresponding set feature parameter structure, so as to obtain the feature parameters of each entity; the influence index determination module provides quality-level parameters for the influence of the target entity based on the citation relationship of the entity or paper in the network database, and the influence index obtained by the web page ranking algorithm; at the same time, the diversity index obtained by the diversity index determination module through the preset knowledge complexity acquisition method also provides quality-level parameters for the influence of the target entity; the central processing unit realizes flexible configuration of multi-dimensional parameters by setting the adjustable weights of each dimension parameter, so that the user can obtain the influence parameters of the whole or a single field based on actual needs. The sub-field parameter determination module further provides quality-level parameters for the influence of the target entity by obtaining the semantic distance between the target entity and the target sub-field category, and the participation of the above semantic distance realizes the acquisition of the influence of the target entity in the target sub-field category. The system of the present invention provides a high-quality entity influence acquisition system that can be fully automated and updated in real time, with adjustable multi-dimensional parameters, many participating parameters and high parameter quality.

[0127] like Figure 2 As shown, this embodiment provides an entity influence evaluation method based on network data. The evaluation method is implemented by the above-mentioned entity influence evaluation system based on network data, and specifically includes the following steps.

[0128] Step S201, obtaining a target entity, and determining and setting a feature parameter structure based on the target entity.

[0129] Specifically, the target entity is obtained through a target entity input module in the entity influence evaluation system based on network data, and a characteristic parameter structure is set based on the target entity.

[0130] Step S202: acquiring a characteristic parameter group of a target entity from a network database based on a set characteristic parameter structure, and using all characteristic parameters in the characteristic parameter group as dimension parameters of the target entity.

[0131] Specifically, through the feature parameter acquisition module in the entity influence evaluation system based on network data, the feature parameter group of the target entity is obtained from the network database based on the set feature parameter structure, and all feature parameters in the feature parameter group are used as dimensional parameters of the target entity.

[0132] Step S203, based on the preset reference relationship in the network database, the influence index of the target entity is obtained through a web page ranking algorithm, and the influence index of the target entity is used as a dimension parameter of the target entity.

[0133] Specifically, the influence index determination module in the entity influence evaluation system based on network data is used to obtain the influence index of the target entity based on the preset reference relationship in the network database through the web page ranking algorithm, and the influence index of the target entity is used as the dimensional parameter of the target entity.

[0134] Step S204: obtaining a diversity index of the target entity by using a preset knowledge complexity acquisition method, and using the diversity index of the target entity as a dimension parameter of the target entity.

[0135] Specifically, through the diversity index determination module in the entity influence evaluation system based on network data, the diversity index of the target entity is obtained through a preset knowledge complexity acquisition method, and the diversity index of the target entity is used as the dimensional parameter of the target entity.

[0136] Step S205: performing a weighted operation on all dimensional parameters of the target entity to obtain an influence parameter of the target entity.

[0137] Specifically, a central processor in an entity influence evaluation system based on network data performs weighted operations on all dimensional parameters of a target entity to obtain the influence parameters of the target entity.

[0138] The network database includes a preset knowledge base and / or a search database in the network.

[0139] The entity influence evaluation method based on network data provided by the embodiment of the present invention captures structured data from the network database through the feature parameter acquisition module, and selects the feature parameter group of each target entity from the acquired structured data based on the corresponding set feature parameter structure to achieve the acquisition of each entity feature parameter; the influence index determination module provides quality-level parameters for the influence of the target entity based on the citation relationship of the entity or paper in the network database, and the influence index obtained by the web page ranking algorithm; at the same time, the diversity index obtained by the diversity index determination module through the preset knowledge complexity acquisition method also provides quality-level parameters for the influence of the target entity; the central processing unit realizes the flexible configuration of multi-dimensional parameters by setting the adjustable weights of each dimension parameter, so that the user can obtain the influence parameters of the whole or a single field based on actual needs. The sub-field parameter determination module further provides quality-level parameters for the influence of the target entity by obtaining the semantic distance between the target entity and the target sub-field category, and the participation of the above semantic distance realizes the acquisition of the influence of the target entity in the target sub-field category. The method of the present invention provides a high-quality entity influence acquisition method that can be fully automated and updated in real time, with adjustable multi-dimensional parameters, many participating parameters and high parameter quality.

[0140] Although the embodiments disclosed in the present invention are as above, the contents described are only embodiments adopted for facilitating the understanding of the present invention and are not intended to limit the present invention. Any technician in the technical field to which the present invention belongs can make any modifications and changes in the form and details of the implementation without departing from the spirit and scope disclosed in the present invention, but the protection scope of the present invention shall still be subject to the scope defined in the attached claims.

Claims

1. An entity influence evaluation system based on network data, characterized in that: It includes a central processing unit and a target entity input module, a characteristic parameter acquisition module, an influence index determination module and a diversity index determination module connected to the central processing unit, and also includes a communication module, wherein the communication module is used to realize the communication between the characteristic parameter acquisition module, the influence index determination module and the diversity index determination module and a network database respectively; The target entity input module is used to obtain the target entity and determine and set the characteristic parameter structure based on the target entity; The characteristic parameter acquisition module is used to obtain the characteristic parameter group of the target entity from the network database through the communication module based on the set characteristic parameter structure, and use all characteristic parameters in the characteristic parameter group as dimension parameters of the target entity; The influence index determination module is used to obtain the influence index of the target entity through a web page ranking algorithm based on the preset reference relationship in the network database, and use the influence index of the target entity as a dimension parameter of the target entity; The diversity index determination module is used to obtain the diversity index of the target entity through a preset knowledge complexity acquisition method, and use the diversity index of the target entity as a dimension parameter of the target entity; The central processor is used to send the target entity to the influence index determination module and the diversity index determination module, send the set characteristic parameter structure to the characteristic parameter acquisition module, and perform weighted operation on all the dimensional parameters of the target entity to obtain the influence parameter of the target entity; Wherein, the network database includes a preset knowledge base and / or a search database in the network; The diversity index determination module obtains the diversity index of the target entity by a preset knowledge complexity acquisition method, including: The target entity and the preset entities under the category to which the target entity belongs are all regarded as target knowledge production units, and based on the category to which the target entity belongs, all target knowledge products are obtained and the export method when the target knowledge production unit exports the target knowledge product is identified; On the basis of the export mode, based on the export advantages of each of the target knowledge products in each of the target knowledge production units, obtaining the diversity initialization index of each of the target knowledge production units and the scarcity initialization index of each of the target knowledge products; Based on the diversity initialization index of each target knowledge production unit and the scarcity initialization index of each target knowledge product, a bidirectional iteration of the diversity index of each target knowledge production unit and the scarcity index of each target knowledge product is realized through a nested matrix to obtain the final diversity index of each target knowledge production unit, and the final diversity index of the target knowledge production unit corresponding to the target entity is used as the diversity index of the target entity; The comparison knowledge production unit is set to be any one of all the target knowledge production units, and the comparison knowledge product is set to be any one of all the target knowledge products. At the same time, based on the export method, the export volume of the comparison knowledge product exported by the comparison knowledge production unit is taken as the individual export volume, the export volume of all the target knowledge products exported by the comparison knowledge production unit is taken as the individual total export volume, the export volume of all the target knowledge production units exporting the comparison knowledge product is taken as the overall export volume, and the export volume of all the target knowledge production units exporting all the target knowledge products is taken as the overall total export volume. At this time, it is determined whether the ratio of the individual export volume to the individual total export volume is greater than the ratio of the overall export volume to the overall total export volume. If so, it is determined that the comparison knowledge production unit has an export advantage when exporting the comparison knowledge product. Otherwise, it is determined that the comparison knowledge production unit does not have an export advantage when exporting the comparison knowledge product. On the basis of setting the comparison knowledge production unit as any one of all the target knowledge production units, obtaining the diversity initialization index of the comparison knowledge production unit based on the export advantage of each target knowledge product in the comparison knowledge production unit specifically includes: first, sorting all the target knowledge products to obtain a knowledge product sequence; then setting a value N with an initial value of 1, judging whether the current N is less than or equal to the preset product threshold, if the current N is less than or equal to the preset product threshold, it means that all the target knowledge products have not been completely screened, and at this time, it is necessary to judge whether the comparison knowledge production unit has an export advantage when exporting the Nth target knowledge product in the knowledge product sequence, if so, the diversity temporary index of the current comparison knowledge production unit is increased by 1, otherwise, there is no need to modify the diversity temporary index of the current comparison knowledge production unit; then, the current N is increased by 1 to obtain a new N, and at the same time, the above process of judging whether the current N is less than or equal to the preset product threshold is returned to realize the re-judgment of the current N value and the subsequent corresponding steps, and if when judging whether the current N is less than or equal to the preset product threshold, it is judged that the current N is greater than the preset product threshold, it means that all the target knowledge products have been screened, and at this time, the diversity temporary index of the current comparison knowledge production unit is directly used as the diversity initialization index of the comparison knowledge production unit; On the basis of setting the comparison knowledge product as any one of all the target knowledge products, based on the export advantages of the comparison knowledge product in each target knowledge production unit, obtaining the scarcity initialization index of the comparison knowledge product specifically includes: firstly sorting all the target knowledge production units to obtain the knowledge production unit sequence; then setting the value M with an initial value of 1, judging whether the current M is less than or equal to the preset production unit threshold, if the current M is less than or equal to the preset production unit threshold, it means that all the target knowledge production units have not been completely screened, and at this time it is necessary to judge whether the Mth target knowledge production unit in the knowledge production unit sequence has an export when exporting the comparison knowledge product. Advantage, if yes, then add 1 to the temporary index of scarcity of the current comparison knowledge product, otherwise there is no need to modify the temporary index of scarcity of the current comparison knowledge product; then add 1 to the current M to obtain a new M, and return to the above process of judging whether the current M is less than or equal to the preset production unit threshold, that is, to re-judge the current M value and the subsequent corresponding steps, and if when judging whether the current M is less than or equal to the preset production unit threshold, it is judged that the current M is greater than the preset production unit threshold, it means that the screening of all target knowledge production units has been completed, and at this time, the temporary index of scarcity of the current comparison knowledge product is directly used as the scarcity initialization index of the comparison knowledge product; Among them, the preset product threshold is the number of target knowledge products in the knowledge product sequence, and the initial value of the temporary indicator of diversity of the comparison knowledge production unit is 0; the preset production unit threshold is the number of target knowledge production units in the knowledge production unit sequence, and the initial value of the temporary indicator of scarcity of the comparison knowledge product is 0.

2. The influence assessment system according to claim 1, characterized in that: The set feature parameter structure includes a specific feature parameter type group and an additional feature parameter type group; the specific feature parameter type group includes multiple specific feature parameter types and a specific weight corresponding to each of the specific feature parameter types; the additional feature parameter type group includes multiple additional feature parameter types and an adjustable weight corresponding to each of the additional feature parameter types.

3. The influence assessment system according to claim 2, characterized in that: The characteristic parameter acquisition module acquires the characteristic parameter group of the target entity from the network database through the communication module based on the set characteristic parameter structure, including: The feature parameter acquisition module captures the structured data of the target entity from the network database through the communication module, obtains the corresponding specific feature parameters from the structured data of the target entity based on each specific feature parameter type in the specific feature parameter type group, and obtains the corresponding additional feature parameters from the structured data of the target entity based on each additional feature parameter type in the additional feature parameter type group, and combines all the specific feature parameters and all the additional feature parameters into a feature parameter group of the target entity.

4. The influence assessment system according to claim 1, characterized in that: When the preset reference relationship is an entity reference relationship, determine whether the target entity can be a category in the network database. If so, the influence index of the target entity is the sum of the influence indexes of all entities under the target entity when the target entity is a category in the network database. Otherwise, the influence index of the target entity is the influence index of the corresponding entity. When the preset citation relationship is a paper citation relationship, the influence index of the target entity is the sum of the influence indexes of all papers under the target entity.

5. The influence assessment system according to claim 1, characterized in that: It also includes a sub-domain parameter determination module connected to the central processor, and the sub-domain parameter determination module communicates with the target network knowledge base through the communication module; The sub-domain parameter determination module is used to set the target sub-domain category, obtain the semantic distance between the target entity and the target sub-domain category through a preset semantic distance acquisition method, and use the semantic distance between the target entity and the target sub-domain category as the dimensional parameter of the target entity.

6. The influence assessment system according to claim 5, characterized in that: The sub-domain parameter determination module obtains the semantic distance between the target entity and the target sub-domain category by a preset semantic distance acquisition method, including: Acquire a link set of a target entity from a target network knowledge base through the communication module as an entity link set of the target entity, acquire a link set of the target sub-domain category as an entity from the target network knowledge base through the communication module as an entity link set of the target sub-domain category, and use a semantic distance between the entity link set of the target entity and the entity link set of the target sub-domain category as a first semantic distance between the target entity and the target sub-domain category; When the target entity is used as a category in the target network knowledge base, all entity sets under it are used as a temporary set to be ranked; a link set of each entity in the temporary set to be ranked is obtained from the target network knowledge base through the communication module and merged into a category link set of the target entity; all entity sets under the target sub-domain category are used as a sub-domain temporary set; a link set of each entity in the sub-domain temporary set is obtained from the target network knowledge base through the communication module and merged into a category link set of the target sub-domain category; and a semantic distance between the category link set of the target entity and the category link set of the target sub-domain category is used as a second semantic distance between the target entity and the target sub-domain category; The first semantic distance between the target entity and the target sub-domain category, the second semantic distance between the target entity and the target sub-domain category, or the weighted average of the first semantic distance and the second semantic distance between the target entity and the target sub-domain category is used as the semantic distance between the target entity and the target sub-domain category.

7. The influence assessment system according to claim 1, characterized in that: The central processor performs a linear weighted operation or a nonlinear weighted operation on all the dimensional parameters of the target entity to obtain the influence parameter of the target entity.

8. The influence assessment system according to claim 1, characterized in that: It also includes a data storage module connected to the central processing unit, and the data storage module is used to store the characteristic parameter group of the target entity, the influence index of the target entity, the diversity index of the target entity and the influence parameter of the target entity.

9. A method for evaluating entity influence based on network data based on the entity influence evaluation system based on network data according to any one of claims 1 to 8, comprising: Acquire a target entity, and determine and set a characteristic parameter structure based on the target entity; Acquire a characteristic parameter group of the target entity from a network database based on the set characteristic parameter structure, and use all characteristic parameters in the characteristic parameter group as dimension parameters of the target entity; Based on the preset reference relationship in the network database, the influence index of the target entity is obtained through a web page ranking algorithm, and the influence index of the target entity is used as a dimension parameter of the target entity; Acquire the diversity index of the target entity by a preset knowledge complexity acquisition method, and use the diversity index of the target entity as a dimension parameter of the target entity; Performing a weighted operation on all the dimension parameters of the target entity to obtain an influence parameter of the target entity; Wherein, the network database includes a preset knowledge base and / or a search database in the network.

Images