Electric power scientific research information retrieval result optimization method and device

Abstract

The invention provides an electric power scientific research information retrieval result optimization method and device, and relates to the technical field of electric power scientific research information.The method comprises the steps that an electric power field scientific research database is mined, entity recognition is conducted, and an electric power field scientific research knowledge base is constructed; performing classified coding and knowledge integration on the scientific research knowledge to obtain a power field scientific research retrieval library; inputting a target retrieval word, and performing associated retrieval in a scientific research retrieval library in the power field to obtain a matched scientific research information set; and constructing a dynamic weight sorting model, carrying out weight analysis sorting on the matched scientific research information set, determining an electric power scientific research sequence retrieval result, and carrying out visual optimization display. By means of the method and device, the technical problem that in the prior art, due to the fact that electric power scientific research information is scattered and difficult to efficiently retrieve and lacks effective information sorting, the quality of the retrieval result is not high is solved, through dynamic weight adjustment and real-time feedback, the scientificity and practicability of the sorting result are ensured, and the efficiency and accuracy of electric power scientific research information retrieval are improved.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of electric power scientific research information, and particularly relates to an electric power scientific research information retrieval result optimization method and device. BACKGROUND

[0002] With the rapid development of the electric power industry, a large amount of scientific research achievements and technical literature have been produced in the field of electric power scientific research, which are distributed in various channels such as databases, periodical papers, conference papers, patent literature and project reports, and form a huge and complex information resource system. The electric power scientific research information resources are scattered, and lack effective integration mechanisms, which leads to the fact that researchers need to access multiple independent data sources in the retrieval process, and consume a lot of time and effort. The existing retrieval methods are mostly based on simple keyword matching, lack in-depth identification of electric power field professional terms and scientific research entities, are difficult to accurately understand the user retrieval intention, contain a large amount of irrelevant information in the retrieval results, and reduce the accuracy and efficiency of retrieval. In addition, the sorting method of the retrieval results is relatively single, and is usually only sorted according to the keyword matching degree or the literature publication time, lacks comprehensive consideration of the literature quality, academic influence and multi-dimensional weight, and makes it difficult for users to quickly obtain high-quality scientific research information. The display form of the retrieval results is also relatively single, lacks intuitive and interactive visualization support, and affects the understanding and utilization of the scientific research information by the users.

[0003] In summary, in the prior art, there is a technical problem that the electric power scientific research information is scattered, difficult to efficiently retrieve, and lacks effective information sorting, which leads to low quality of the retrieval results. SUMMARY

[0004] The purpose of the present application is to provide an electric power scientific research information retrieval result optimization method and device, to solve the technical problem in the prior art that the electric power scientific research information is scattered, difficult to efficiently retrieve, and lacks effective information sorting, which leads to low quality of the retrieval results.

[0005] In view of the above problems, the present application provides an electric power scientific research information retrieval result optimization method and device.

[0006] In a first aspect, the application provides a power scientific research information retrieval result optimization method, which is implemented by a power scientific research information retrieval result optimization device. The power scientific research information retrieval result optimization method comprises the following steps: mining a power field scientific research database, performing entity recognition on the power field scientific research database, and constructing a power field scientific research knowledge base; classifying and integrating knowledge in the power field scientific research knowledge base to obtain a power field scientific research retrieval library; inputting a target search term, performing associated retrieval in the power field scientific research retrieval library based on the target search term to obtain a matching scientific research information set; constructing a dynamic weight ordering model, performing weight analysis and ordering on the matching scientific research information set by using the dynamic weight ordering model, determining a power scientific research sequence retrieval result, and performing visual optimization display based on the power scientific research sequence retrieval result.

[0007] Optionally, the power field scientific research database is subjected to repeated data removal and error data correction to obtain an available power field scientific research database; a power entity recognizer is constructed according to power field professional knowledge and language rules; and the available power field scientific research database is subjected to entity recognition and knowledge base construction based on the power entity recognizer to obtain the power field scientific research knowledge base.

[0008] Optionally, a power knowledge classification standard is constructed, a power scientific research knowledge coding system is designed according to the power knowledge classification standard, each scientific research knowledge in the power field scientific research knowledge base is classified and coded according to the power scientific research knowledge coding system to obtain a scientific research knowledge coding information set, and the power field scientific research knowledge base is integrated based on the scientific research knowledge coding information set to obtain the power field scientific research retrieval library.

[0009] Optionally, the power knowledge classification standard is divided into multiple coding levels to obtain a multi-level coding level, coding rules are designed based on the multi-level coding level to determine a multi-level power coding rule, the multi-level power coding rule comprises a coding identifier and a coding length, and the power scientific research knowledge coding system is constructed according to the multi-level coding level and the multi-level power coding rule.

[0010] Optionally, the target search term is subjected to word segmentation processing and power term association to obtain an extended power keyword combination, a power retrieval formula is constructed according to the extended power keyword combination, and associated retrieval is performed in the power field scientific research retrieval library by using the power retrieval formula to obtain the matching scientific research information set.

[0011] Optionally, according to the characteristics of the power scientific research information and the user demand, a set of information sorting indexes is determined; weight distribution is performed on each index information in the set of information sorting indexes to obtain a set of sorting index weight information; and the set of information sorting indexes is weighted and fitted based on the set of sorting index weight information to construct the dynamic weight sorting model.

[0012] Optionally, each scientific research information in the set of matched scientific research information is weighted and calculated by using the dynamic weight sorting model to obtain a set of scientific research information comprehensive scores; a set of scientific research information weight factors is determined according to the set of scientific research information comprehensive scores; and the set of matched scientific research information is arranged in descending order according to the set of scientific research information weight factors to determine the power scientific research sequence retrieval result.

[0013] In a second aspect, the application further provides a power scientific research information retrieval result optimization device for executing the power scientific research information retrieval result optimization method as described in the first aspect, wherein the power scientific research information retrieval result optimization device comprises: an entity recognition module for mining a power field scientific research database, performing entity recognition on the power field scientific research database, and constructing a power field scientific research knowledge base; a knowledge integration module for classifying and coding each scientific research knowledge in the power field scientific research knowledge base and integrating the knowledge to obtain a power field scientific research retrieval library; an associated retrieval module for inputting a target retrieval word, performing associated retrieval in the power field scientific research retrieval library based on the target retrieval word, and obtaining a set of matched scientific research information; and a weight sorting module for constructing a dynamic weight sorting model, performing weight analysis and sorting on the set of matched scientific research information by using the dynamic weight sorting model, determining a power scientific research sequence retrieval result, and performing visualized optimization display based on the power scientific research sequence retrieval result.

[0014] The one or more technical solutions provided in the application have at least the following beneficial effects:

[0015] The power field scientific research database is mined, entity recognition is performed on the power field scientific research database, and a power field scientific research knowledge base is constructed; each scientific research knowledge in the power field scientific research knowledge base is classified and coded and knowledge integration is performed, to obtain a power field scientific research search library; a target search term is input, and based on the target search term, associated search is performed in the power field scientific research search library to obtain a matching scientific research information set; a dynamic weight ordering model is constructed, the dynamic weight ordering model is used to perform weight analysis ordering on the matching scientific research information set, a power scientific research sequence search result is determined, and visualized optimal display is performed based on the power scientific research sequence search result. That is, by constructing a power field scientific research knowledge base, each scientific research knowledge in the power field scientific research knowledge base is classified and coded and knowledge integration is performed, to obtain a power field scientific research search library, associated search is performed in the power field scientific research search library according to a target search term, a matching scientific research information set is output, weight analysis ordering is performed on the matching scientific research information set according to a dynamic weight ordering model, a power scientific research sequence search result is determined, visualized optimal search is performed, the accuracy and efficiency of information search are improved, and meanwhile, the understandability and analyzability of information are also enhanced.

[0016] The above description is only a summary of the technical solutions of the present application. In order to enable the technical means of the present application to be more clearly understood, and to be implemented according to the content of the description, and in order to enable the above and other purposes, characteristics and advantages of the present application to be more apparent and easy to understand, the following specific embodiments of the present application are described. It should be understood that the content described in this section is not intended to identify key or important features of the embodiments of the present application, nor is it intended to limit the scope of the present application. Other features of the present application will become apparent from the following description. BRIEF DESCRIPTION OF DRAWINGS

[0017] In order to more clearly illustrate the technical solutions in the present application or the prior art, the following will briefly introduce the drawings needed to be used in the embodiments or the prior art description. Obviously, the drawings in the following description are only exemplary, and other drawings can also be obtained by the provided drawings without creative labor for those skilled in the art.

[0018] Figure 1 The flowchart of the power scientific research information search result optimization method of the present application.

[0019] Figure 2 The structural diagram of the power scientific research information search result optimization device of the present application.

[0020] Explanation of reference signs: entity recognition module 11, knowledge integration module 12, associated search module 13, weight ordering module 14. DETAILED DESCRIPTION

[0021] The application provides an electric power scientific research information retrieval result optimization method and device, solves the technical problem of low quality of retrieval results due to dispersion of electric power scientific research information, difficulty in efficient retrieval, and lack of effective information sorting in the prior art. By constructing an electric power field scientific research knowledge base, classifying and coding each scientific research knowledge in the electric power field scientific research knowledge base, and integrating knowledge, an electric power field scientific research retrieval library is obtained. According to the target search term, associated retrieval is performed in the electric power field scientific research retrieval library, a matching scientific research information set is output, the matching scientific research information set is sorted according to a dynamic weight sorting model, the electric power scientific research sequence retrieval result is determined, visual optimization retrieval is performed, the accuracy and efficiency of information retrieval are improved, and the intelligibility and analyzability of information are also enhanced.

[0022] The technical solutions in the application will be described clearly and completely below with reference to the drawings. Obviously, the described embodiments are only some of the embodiments of the application, rather than all the embodiments of the application. It should be understood that the application is not limited to the example embodiments described herein. Based on the embodiments of the application, all other embodiments obtained by a person of ordinary skill in the art without creative work fall within the scope of protection of the application. In addition, it should be noted that, for the convenience of description, only parts related to the application are shown in the drawings rather than all parts.

[0023] Embodiment one, please refer to the accompanying Figure 1 The application provides an electric power scientific research information retrieval result optimization method, wherein the electric power scientific research information retrieval result optimization method is executed by an electric power scientific research information retrieval result optimization device, and the electric power scientific research information retrieval result optimization method specifically comprises the following steps:

[0024] S100: Mining an electric power field scientific research database, performing entity recognition on the electric power field scientific research database, and constructing an electric power field scientific research knowledge base.

[0025] Further, the S100 of the application comprises:

[0026] The electric power field scientific research database is subjected to repeated data removal and error data correction to obtain an available electric power field scientific research database; an electric power entity recognizer is constructed according to electric power field professional knowledge and language rules; the available electric power field scientific research database is subjected to entity recognition and knowledge base construction based on the electric power entity recognizer to obtain the electric power field scientific research knowledge base.

[0027] Specifically, the power field research database is mined, that is, power-related research data resources are collected and analyzed, including collection from multiple databases such as journal papers, patents, project reports, conference literature, and technical standards. To ensure the effectiveness of the data, data preprocessing is performed, including duplicate data removal (using hash comparison, similarity algorithms such as cosine similarity to remove duplicate documents) and error data correction (using a combination of rule checking and manual review, automatically detecting format errors, missing fields, and correcting them). Duplicate data removal refers to deleting duplicate records in the database to avoid data redundancy. The power field research database is subjected to duplicate data removal using data cleaning tools, which usually involves comparing the similarity of records and deleting identical records. Error data correction refers to correcting erroneous information in the database to improve data quality. The remaining data is subjected to error data correction, which can be achieved through manual review or the use of error detection algorithms. For example, assume that there are 10,000 records in the original database, of which 1,000 are duplicates and 500 contain erroneous information. Through duplicate data removal and error data correction, 9,500 usable records are obtained.

[0028] After data cleaning, a power entity recognizer is constructed based on the professional knowledge and language rules of the power field. The power entity recognizer is a model or tool based on natural language processing (NLP) technology that automatically identifies key entities with professional significance in research literature, such as power equipment names, technical parameters, professional terms, and research objects. The power field professional knowledge and language rules refer to the professional terms and grammatical structures in the power field. The power field professional knowledge and language rules are collected, which usually involves consulting professional books, papers, standards, etc. Natural language processing techniques such as named entity recognition and syntax analysis are used to construct the power entity recognizer.

[0029] The power entity recognizer is used to perform entity recognition on the text in the available power field research database, extracting entities and their attributes, relationships, and other information from the text. The entities are organized into a structured knowledge base, such as using a knowledge graph to represent the relationships between entities, to establish a power field research knowledge base. For example, 1,000 power research papers were subjected to entity recognition using the power entity recognizer, and a total of 5,000 entities were identified, such as transformers and power grids. The entities and their relationships are represented using a knowledge graph to construct a knowledge base containing these entities. Through data cleaning and professional entity recognition technology, the redundancy and errors of power research data are successfully eliminated, ensuring a high-quality foundation for the data, achieving data quality improvement and information structuring, and significantly improving the usability and retrieval efficiency of power research information.

[0030] S200: Classify and code each research knowledge in the power field research knowledge base and integrate the knowledge to obtain a power field research search library.

[0031] Further, the application S200 comprises:

[0032] The power knowledge classification standard is constructed, and a power scientific research knowledge coding system is designed according to the power knowledge classification standard. Each scientific research knowledge in the power field scientific research knowledge base is classified and coded according to the power scientific research knowledge coding system, and a scientific research knowledge coding information set is obtained. The power field scientific research knowledge base is integrated based on the scientific research knowledge coding information set, and a power field scientific research search library is obtained.

[0033] Further, the application further comprises the following steps:

[0034] The power knowledge classification standard is coded and divided into levels to obtain a multi-level coding level. Coding rules are designed based on the multi-level coding level to determine a multi-level power coding rule, and the multi-level power coding rule includes coding identification and coding length. The power scientific research knowledge coding system is constructed according to the multi-level coding level and the multi-level power coding rule.

[0035] Specifically, according to the professional requirements and existing scientific research contents of the power field, the power knowledge classification standard is constructed, and the scientific research knowledge is clearly divided into several categories (such as power generation technology, power transmission and distribution technology, smart grid, etc.), and is refined to sub-categories and more detailed categories, forming a hierarchical classification system. According to the power knowledge classification standard, the coding level corresponding to each classification level is determined, and a multi-level coding structure is established. For example, the first level coding uses 2-digit numbers to represent the major category, the second level coding uses 3-digit numbers to represent the fine classification, and the third level coding uses 4-digit numbers to refine to the specific sub-field. The multi-level coding level is a coding system that contains multiple levels, and each level corresponds to different classification fineness, such as first level classification coding representing power major category, second level classification coding representing sub-field, and third or more level coding further refining.

[0036] According to the multi-level coding level, the coding rule is designed, and the coding rule of each coding level is designed, including coding identification and coding length. The rule includes the length limit of each level coding, the allowed character range (numbers, letters), the connection symbol between levels, etc. The coding rules are combined to form a multi-level power coding rule. The multi-level power coding rule refers to the coding rule designed according to the multi-level coding level, including coding identification and coding length, etc. For example, the first level coding length is 2 digits, and the coding identification is a number; the second level coding length is 3 digits; the third level coding length is 4 digits, etc., to ensure that the coding is unique and the hierarchical structure is clear.

[0037] According to the multi-level coding hierarchy and the multi-level power coding rules, a power scientific research knowledge coding system is built to assign a corresponding code to each piece of scientific research knowledge in the knowledge base, supporting fast classification query and management. The coding rules corresponding to each classification level are applied to the actual power scientific research knowledge to generate the corresponding code. The power scientific research knowledge coding system refers to a coding system built according to the multi-level coding hierarchy and the multi-level power coding rules, which is used to code and identify the power scientific research knowledge.

[0038] According to the power scientific research knowledge coding system, each piece of scientific research knowledge in the power field research knowledge base is classified and coded, i.e. each piece of scientific research knowledge in the power field research knowledge base is coded one by one. An automatic coding tool or program is used to automatically generate the corresponding code according to the classification level to which the scientific research knowledge belongs, forming a scientific research knowledge coding information set, i.e. each piece of scientific research knowledge corresponds to a unique code.

[0039] According to the scientific research knowledge coding information set, the power field research knowledge base is integrated. Scientific research knowledge with similar or related codes is aggregated, de-duplicated and associated to eliminate redundant information and establish a logical association network between knowledge. Through integration, a unified, structured and easily managed and retrieved power field research search library is formed. The power field research search library is a collection of power scientific research knowledge for searching after classification and coding and knowledge integration, supporting efficient and accurate information query. For example, the power field research search library containing 1000 papers is built by integrating the power field research knowledge base using the scientific research knowledge coding information set. Users can search using the code or keywords of the paper, such as inputting the code "0123" to retrieve all papers belonging to this classification.

[0040] By building a scientific coding system and classifying and integrating the power scientific research knowledge, the standardized management and structured organization of scientific research knowledge are realized, the accuracy and completeness of information are improved, and the subsequent search efficiency and search quality are significantly optimized.

[0041] S300: input a target search term, perform associated search in the power field research search library based on the target search term, and obtain a matching scientific research information set.

[0042] Further, the S300 of the present application comprises:

[0043] The target search term is subjected to word segmentation processing and power term association to obtain an expanded power keyword combination. The power search formula is constructed according to the expanded power keyword combination. The power search formula is used to perform associated search in the power field research search library to obtain the matching scientific research information set.

[0044] Specifically, the input target retrieval word, i.e., the keyword used by the user for query, represents the power research topic or content that the user wants to search. The user inputs relevant keywords or phrases in the input interface of the retrieval system according to his own needs, which are usually related to the research topics, technical difficulties or problems in the power field.

[0045] The target retrieval word is processed by using a word segmentation algorithm (such as Jieba segmentation, HanLP, etc.), and the input text string is segmented into independent lexical units. The segmented words are matched with the term library in the power field to find related power terms and form an extended power keyword combination. Power term association refers to matching and associating the segmented words with the professional terms in the power field. The extended power keyword combination is a set of keyword combinations formed after word segmentation and term expansion, which is used to enhance the coverage and accuracy of the retrieval.

[0046] In other words, the basic words are expanded by term association using the built-in power field term library or ontology library to obtain possible synonyms, hyponyms or technical association words, forming an extended power keyword combination. According to the extended power keyword combination, a power retrieval formula is constructed, i.e., a structured retrieval sentence generated by using the keyword combination, which is used to perform matching retrieval in the database.

[0047] The power retrieval formula is used to search in the power field research retrieval library, and the associated retrieval is performed in the power field research retrieval library through keyword matching to obtain a matching research information set. The matching research information set refers to the set of related research information obtained by searching in the power field research retrieval library according to the power retrieval formula, which is a set of research information such as papers, reports, patents, etc. that matches the retrieval formula.

[0048] Through word segmentation processing, power term association, construction of power retrieval formula and associated retrieval, relevant research information is found in the power field research retrieval library according to the user input target retrieval word, which improves the accuracy and relevance of information retrieval. At the same time, the use of extended power keyword combination and power retrieval formula also improves the flexibility and coverage of the retrieval.

[0049] S400: Construct a dynamic weight ordering model, use the dynamic weight ordering model to analyze and sort the weights of the matching research information set, determine the power research sequence retrieval result, and perform visual optimization display based on the power research sequence retrieval result.

[0050] Further, the S400 of the present application includes:

[0051] According to the characteristics of the power research information and the user demand, a set of information sorting indexes is determined; weight distribution is performed on each index information in the set of information sorting indexes to obtain a set of sorting index weight information; and the set of information sorting indexes is weighted and fitted based on the set of sorting index weight information to construct the dynamic weight sorting model.

[0052] Further, the application further includes the following steps:

[0053] The dynamic weight sorting model is used to perform weighted calculation on each research information in the set of matched research information to obtain a set of research information comprehensive scores; a set of research information weight factors is determined according to the set of research information comprehensive scores; and the set of matched research information is arranged in descending order according to the set of research information weight factors to determine the power research sequence retrieval result.

[0054] Specifically, a set of information sorting indexes is determined according to the characteristics of the power research information and the user demand. The characteristics of the power research information are inherent attributes of the power field research data, such as technical field classification (such as new energy, power grid control, etc.), information source (journal, conference), update timeliness, cited times, research method, achievement type (patent, paper, etc.), etc. The user demand is the focus of the research information of the retrieval user in actual use, such as the priority of the latest research, the priority of the high-cited value literature or the priority of the authoritative achievement of a specific technical field. The characteristics of the power research information are analyzed, such as the use of professional terms, the diversity of research directions, etc. The demand and preference of the user are understood through user research, questionnaire survey, etc. According to the characteristics of the power research data and the user research results, a plurality of indexes affecting the sorting quality of the retrieval result are selected.

[0055] According to the importance of each index information in the set of information sorting indexes, a corresponding weight is given, and the determination of the weight usually needs to comprehensively consider the characteristics of the power research information and the user demand to obtain a set of sorting index weight information, which is a set containing each index and the corresponding weight.

[0056] According to the set of sorting index weight information, the set of information sorting indexes is weighted and fitted, and a dynamic weight sorting model is constructed through a weighted summation formula. Weighted fitting refers to the process of mathematical modeling of the index set according to the weight of each index, and the dynamic weight sorting model refers to a mathematical model for sorting information according to weight information. By determining the set of information sorting indexes, performing weight distribution and constructing the dynamic weight sorting model, the information is sorted according to the characteristics of the power research information and the user demand, which improves the accuracy and relevance of information retrieval.

[0057] The dynamic weight ranking model is used to perform weighted calculation on each scientific research information in the matched scientific research information set, to obtain a comprehensive score of each information. Through the weighted calculation, the total score set of each scientific research information is the basis for ranking. The score of each information = the sum of each index x weight. The comprehensive score of each scientific research information is the weight factor of the item in the final ranking, which constitutes the scientific research information weight factor set. The scientific research information weight factor set represents the reference value of each information in the ranking, which is usually the score calculated finally, i.e., the numerical quantification of the matching degree of each scientific research information with the user's search intention.

[0058] According to the scientific research information weight factor set, the matched scientific research information set is arranged in descending order, i.e., ranked from high to low according to the scientific research information weight factor set, to obtain the final power scientific research sequence search result. The power scientific research sequence search result refers to the result list obtained by ranking the matched scientific research information set according to the scientific research information weight factor set. The power scientific research sequence search result is visualized using data visualization tools such as bar charts and scatter plots. An interactive interface is provided to enable users to filter and optimize the search results according to their own needs. The comprehensive score is calculated by the dynamic weight weighting model, which accurately reflects the matching degree of the scientific research information with the user's intention. Through descending arrangement and visualization display, the user's perception quality and use efficiency of the search results are significantly improved.

[0059] In summary, the power scientific research information search result optimization method provided by the present application has the following beneficial effects:

[0060] By mining the power field scientific research database, entity recognition is performed on the power field scientific research database to construct a power field scientific research knowledge base. Each scientific research knowledge in the power field scientific research knowledge base is classified and coded and knowledge is integrated to obtain a power field scientific research search library. The target search term is input, and based on the target search term, associated search is performed in the power field scientific research search library to obtain a matched scientific research information set. A dynamic weight ranking model is constructed, and the dynamic weight ranking model is used to perform weight analysis and ranking on the matched scientific research information set to determine a power scientific research sequence search result, and the power scientific research sequence search result is visualized and optimized for display. That is, by constructing a power field scientific research knowledge base, each scientific research knowledge in the power field scientific research knowledge base is classified and coded and knowledge is integrated to obtain a power field scientific research search library. According to the target search term, associated search is performed in the power field scientific research search library to output a matched scientific research information set. According to the dynamic weight ranking model, the matched scientific research information set is subjected to weight analysis and ranking to determine a power scientific research sequence search result, which is visualized and optimized for search, improving the accuracy and efficiency of information search, and also enhancing the understandability and analyzability of information.

[0061] Embodiment Two, based on the same inventive concept as the power research information retrieval result optimization method in the aforementioned Embodiment One, the present application also provides a power research information retrieval result optimization device, please refer to the attached Figure 2 , the power research information retrieval result optimization device comprises:

[0062] An entity recognition module 11 is configured to mine a power field research database, perform entity recognition on the power field research database, and construct a power field research knowledge base; a knowledge integration module 12 is configured to classify and code each research knowledge in the power field research knowledge base and integrate knowledge to obtain a power field research retrieval library; an associated retrieval module 13 is configured to input a target search term, perform associated retrieval in the power field research retrieval library based on the target search term, and obtain a matching research information set; a weight ordering module 14 is configured to construct a dynamic weight ordering model, use the dynamic weight ordering model to perform weight analysis and ordering on the matching research information set, determine a power research sequence retrieval result, and perform visual optimization display based on the power research sequence retrieval result.

[0063] Further, the entity recognition module 11 in the power research information retrieval result optimization device is also configured to:

[0064] perform duplicate data removal and error data correction on the power field research database to obtain a usable power field research database; construct a power entity recognizer according to power field professional knowledge and language rules; perform entity recognition and knowledge base construction on the usable power field research database based on the power entity recognizer to obtain the power field research knowledge base.

[0065] Further, the knowledge integration module 12 in the power research information retrieval result optimization device is also configured to:

[0066] construct a power knowledge classification standard, design a power research knowledge coding system according to the power knowledge classification standard; classify and code each research knowledge in the power field research knowledge base according to the power research knowledge coding system to obtain a research knowledge coding information set; integrate knowledge of the power field research knowledge base based on the research knowledge coding information set to obtain the power field research retrieval library.

[0067] Further, the knowledge integration module 12 in the power research information retrieval result optimization device is also configured to:

[0068] The power knowledge classification standard is coded and hierarchical, and a multi-level coding hierarchy is obtained; coding rules are designed based on the multi-level coding hierarchy, and multi-level power coding rules are determined, including coding identification and coding length; and the power scientific research knowledge coding system is constructed according to the multi-level coding hierarchy and the multi-level power coding rules.

[0069] Further, the association retrieval module 13 in the power scientific research information retrieval result optimization device is further used for:

[0070] The target retrieval word is subjected to word segmentation processing and power term association, and an extended power keyword combination is obtained; a power retrieval formula is constructed according to the extended power keyword combination; and the power retrieval formula is used for association retrieval in the power field scientific research retrieval library, and the matching scientific research information set is obtained.

[0071] Further, the weight ordering module 14 in the power scientific research information retrieval result optimization device is further used for:

[0072] According to the characteristics of power scientific research information and user demand, an information ordering index set is determined; weight distribution is performed on each index information in the information ordering index set, and an ordering index weight information set is obtained; the information ordering index set is weighted and fitted based on the ordering index weight information set, and the dynamic weight ordering model is constructed.

[0073] Further, the weight ordering module 14 in the power scientific research information retrieval result optimization device is further used for:

[0074] Each scientific research information in the matching scientific research information set is subjected to weighted calculation by using the dynamic weight ordering model, and a scientific research information comprehensive score set is obtained; a scientific research information weight factor set is determined according to the scientific research information comprehensive score set; the matching scientific research information set is arranged in descending order according to the scientific research information weight factor set, and the power scientific research sequence retrieval result is determined.

[0075] Each embodiment in the specification is described in a progressive manner, and each embodiment focuses on the difference from other embodiments. Figure 1 The power scientific research information retrieval result optimization method and specific examples in Embodiment One are also applicable to the power scientific research information retrieval result optimization device of the present embodiment. Through the foregoing detailed description of the power scientific research information retrieval result optimization method, those skilled in the art can clearly understand the power scientific research information retrieval result optimization device in the present embodiment, so for the sake of brevity of the specification, it will not be described in detail here.

[0076] The above description of disclosed embodiments enables one of ordinary skill in the art to make or use the application. Various modifications to these embodiments will be readily apparent to those skilled in the art, and the generic principles defined herein can be applied to other embodiments without departing from the spirit or scope of the application. Therefore, the present application is not intended to be limited to the embodiments shown herein but is to be accorded the widest scope consistent with the principles and novel features disclosed herein.

[0077] Obviously, many modifications and changes can be made to the application as set forth above without departing from the spirit and scope of the application. It is to be understood that the above description is intended to be illustrative, and not restrictive. Many other embodiments will be apparent to those of skill in the art upon reviewing the above description. The scope of the application should, therefore, be determined not with reference to the above description, but instead with reference to the appended claims, along with their full scope of equivalents.

Claims

1. A method for optimizing electric power research information retrieval results, characterized in that: include: Mining the scientific research database in the electric power field, performing entity recognition on the scientific research database in the electric power field, and building a scientific research knowledge base in the electric power field; Classify and code the scientific research knowledge in the power field scientific research knowledge base and integrate the knowledge to obtain a power field scientific research retrieval base; Input a target search term, and perform an associated search in the power field scientific research search database based on the target search term to obtain a matching scientific research information set; A dynamic weight ranking model is constructed, and the matching scientific research information set is weighted analyzed and ranked using the dynamic weight ranking model to determine the electric power scientific research sequence retrieval results, and a visual optimization display is performed based on the electric power scientific research sequence retrieval results.

2. The method for optimizing electric power scientific research information retrieval results according to claim 1, characterized in that: The construction of the scientific research knowledge base in the field of electric power includes: Deduplicating data and correcting erroneous data from the power field scientific research database to obtain a usable power field scientific research database; Build an electric entity recognizer based on electric power domain expertise and language rules; Based on the electric power entity identifier, entity identification and knowledge base construction are performed on the available electric power field scientific research database to obtain the electric power field scientific research knowledge base.

3. The method for optimizing electric power research information retrieval results according to claim 1, wherein: The said obtaining of the scientific research retrieval database in the field of electric power includes: Establish a classification standard for electric power knowledge and design an electric power scientific research knowledge coding system based on the classification standard; Classify and encode each scientific research knowledge in the scientific research knowledge base in the electric power field according to the electric power scientific research knowledge coding system to obtain a scientific research knowledge coding information set; Based on the scientific research knowledge coding information set, the scientific research knowledge base in the electric power field is integrated to obtain the scientific research retrieval base in the electric power field.

4. The method for optimizing electric power scientific research information retrieval results according to claim 3, characterized in that: The design of the power research knowledge coding system includes: Dividing the power knowledge classification standard into coding levels to obtain a multi-level coding hierarchy; Designing coding rules based on the multi-level coding hierarchy to determine a multi-level power coding rule, wherein the multi-level power coding rule includes a coding identifier and a coding length; The power scientific research knowledge coding system is constructed based on the multi-level coding hierarchy and the multi-level power coding rules.

5. The method for optimizing electric power scientific research information retrieval results according to claim 1, characterized in that: The obtained matching scientific research information set includes: Performing word segmentation processing and electric power term association on the target search term to obtain an expanded electric power keyword combination; Constructing an electric power search formula according to the expanded electric power keyword combination; The electric power search formula is used to perform an associated search in the electric power field scientific research search database to obtain the matching scientific research information set.

6. The method for optimizing electric power research information retrieval results according to claim 1, wherein: The construction of the dynamic weight ranking model includes: Determine the information ranking index set based on the characteristics of power research information and user needs; Performing weight assignment on each indicator information in the information sorting indicator set to obtain a sorting indicator weight information set; The information sorting index set is weightedly fitted based on the sorting index weight information set to construct the dynamic weight sorting model.

7. The method for optimizing electric power research information retrieval results according to claim 1, wherein: The determination of the search results of the electric power scientific research sequence includes: Using the dynamic weight ranking model to perform weighted calculation on each scientific research information in the matching scientific research information set to obtain a comprehensive score set of scientific research information; Determining a scientific research information weight factor set based on the scientific research information comprehensive score set; The matching scientific research information set is arranged in descending order according to the scientific research information weight factor set to determine the power scientific research sequence retrieval result.

8. The device for optimizing the results of electric power research information retrieval is characterized by: The steps for implementing the method for optimizing the electric power scientific research information retrieval results according to any one of claims 1 to 7, wherein the electric power scientific research information retrieval result optimization device comprises: An entity recognition module is used to mine the scientific research database in the electric power field, perform entity recognition on the scientific research database in the electric power field, and build a scientific research knowledge base in the electric power field; A knowledge integration module is used to classify and code various scientific research knowledge in the power field scientific research knowledge base and integrate the knowledge to obtain a power field scientific research retrieval base; An associated search module, used for inputting a target search term, and performing an associated search in the power field scientific research search database based on the target search term to obtain a matching scientific research information set; The weight sorting module is used to construct a dynamic weight sorting model, use the dynamic weight sorting model to perform weight analysis and sorting on the matching scientific research information set, determine the power scientific research sequence retrieval results, and perform visual optimization display based on the power scientific research sequence retrieval results.

Images