Intelligent auxiliary method, system and device for scientific research innovation topic selection

Abstract

The invention relates to an intelligent auxiliary method, system and device for scientific research and innovation topic selection, and the method comprises the steps: obtaining scientific and technological literatures, processing the scientific and technological literatures, carrying out the keyword extraction of a preprocessed data set through a TextRank algorithm, and carrying out the classification of the scientific and technological literatures, such as papers / patent standards, according to the keywords. And then calculating the number of science and technology literatures such as papers / patent standards under each category, performing scoring according to the number to obtain a hot word and new word list, obtaining a recent hot theme list according to the hot word and new word list, displaying the list to a user, and intelligently assisting the user to perform scientific research innovation topic selection. According to the method, the auxiliary decision-making effect of artificial intelligence and a big data technology in scientific research topic selection is exerted, the topic with the most potential and prospect is rapidly screened out from massive scientific research literatures, data and trends, scientific research personnel and innovation teams can select the research direction in a more targeted mode, blind exploration is avoided, and the research efficiency is improved. And scientific and technological innovation efficiency and achievement quality are improved.

Description

Technical Field

[0001] The present invention belongs to the technical field of natural language processing and data mining, and specifically relates to an intelligent auxiliary method, system and device for selecting research and innovation topics. Background Art

[0002] With the development of the times and the progress of science and technology, the social division of labor has become more and more detailed, and everyone has to bear more and more heavy work for their own survival and development needs. Therefore, science and technology have become an indispensable and important tool for contemporary human production and life.

[0003] The selection of research topics is the primary issue of scientific and technological innovation. In scientific research practice, being able to accurately find and overcome research topics that meet the needs of the national development strategy and the future development trend of scientific and technological innovation is one of the important signs of achieving high-level scientific and technological self-reliance. In order to take the initiative and fight a proactive battle for scientific and technological innovation, it is necessary to establish a scientific, independent and open research topic selection mechanism, improve the ability to select research topics, and enhance the level of layout of major scientific research projects.

[0004] At present, my country's economic and social development, energy system optimization and upgrading, and new power system construction are facing many practical problems that need to be solved. It is required to speed up the advancement of technologies that can quickly break through and solve problems in a timely manner; and to deploy in advance technologies that are strategic and require long-term efforts. Grid companies need to increase investment in basic research and applied research, strengthen industry-university-research cooperation, establish a research topic selection mechanism that transforms major practical problems into scientific and technological problems, condense scientific and technological problems from industrial innovation needs, and break through a number of underlying principles and key core technologies.

[0005] The emergence and development of science is inseparable from the division of disciplines. With the development of social economy and the vigorous promotion of modern information technology and computer networks, effective communication platforms and technical support have been provided for scientific research. At the same time, the speed of knowledge update in today's society is also accelerating, resulting in a gradual decline in people's ability to understand knowledge in depth, while their mastery of knowledge breadth is getting higher and higher. As a result, professionals have reached a certain peak in their research on their profession, and if they want to achieve further success, they must enter new technical fields. However, when faced with a complex knowledge system and a large number of research results, researchers often don't know where to start to find a suitable solution. In the actual application process of the applicant of the present invention, it is found that the current solutions to the above problems mainly rely on human experience summary or automatic recommendation of search software to determine a relatively suitable research direction, but these methods still have certain limitations, which are mainly reflected in the following aspects: (1) These solutions are all automatically given results by manually designed programs, without considering personal interests and concerns, and cannot meet personalized needs; (2) Most of the current technical solutions only provide analysis and screening based on existing results, which cannot directly reflect the real intention of the user, and can only provide partial reference opinions, and cannot play the role of intelligent guidance; (3) The information provided by the existing technical solutions is mostly obtained through simple text reading, which easily makes the operator ignore some key information points or even mistakenly regard relevant information as the key points, thereby causing misjudgment, ultimately affecting the progress of subsequent research or causing unnecessary waste of resources. In addition, although the search technology used in the prior art can well realize the rapid retrieval of data and display relevant content, it is difficult to accurately identify truly valuable information and make effective use of it. Due to the large differences between different fields, the same problem may have different solutions in different occasions; and there may be multiple potential solutions to the same problem. From the above, we can see that there is currently a lack of a comprehensive data query service system that can be automatically executed and has intelligent characteristics. Summary of the invention

[0006] The purpose of the present invention is to overcome the defects of the above-mentioned prior art and to provide an intelligent auxiliary method, system and device for selecting research and innovation topics.

[0007] The purpose of the present invention can be achieved by the following technical solutions:

[0008] The present invention provides an intelligent auxiliary method for selecting a research innovation topic, comprising the following steps:

[0009] Obtain scientific and technological literature and its publication time as the original data set;

[0010] The original data set is classified by literature form to obtain paper data set, patent data set and technical standard data set;

[0011] According to the China Library Classification System, the paper datasets are classified into different fields to obtain paper datasets in different fields.

[0012] Classify the patent dataset by field according to patent type to obtain patent datasets in various fields;

[0013] Classify the technical standard data sets by field according to the standard type to obtain the technical standard data sets in various fields;

[0014] The paper datasets, patent datasets and technical standard datasets in various fields are merged by field categories to obtain datasets in various fields;

[0015] Extract titles and abstracts from paper datasets and patent datasets in various fields, extract titles and scopes from technical standard datasets, and use the extracted data as the dataset to be processed in this field;

[0016] Preprocess the data set to be processed;

[0017] Use the TextRank algorithm to extract keywords from the preprocessed data set;

[0018] The extracted keywords are used as the research topics of their corresponding scientific and technological literature;

[0019] Statistical analysis is performed on all topics in each field by publication time and research topic to obtain a hot word list and a new word list in each field, wherein the hot word list includes the topics of the literature published in the first preset time period and their publication frequency, and the new word list includes the topics of the literature published only in the second preset time period and their publication frequency;

[0020] Sort the hot word list and new word list according to the frequency of publication;

[0021] Get the user input of the field of interest;

[0022] Intercept the document subject and its publication frequency at the first preset value position before the new word list in the field of interest, find the publication frequency of its document subject in the hot word list in the field of interest, add the frequency to the intercepted list, and obtain a preliminary list of recent hot topics;

[0023] Sorting the primary list of recent hot topics by the frequency of publication in the first preset time period to obtain a list of recent hot topics;

[0024] Through visualization technology, a list of recent hot topics is displayed to users, intelligently assisting users in selecting topics for scientific research and innovation.

[0025] Furthermore, the scientific and technological documents include: academic journals, academic papers, conference papers, scientific and technological achievements, patent documents and technical standard documents.

[0026] Furthermore, the paper data set includes academic journals, academic papers, conference papers and scientific and technological achievements; the patent data set includes patent documents; and the technical standard data set includes technical standard documents.

[0027] Furthermore, the data to be processed includes unstructured data such as word and pdf.

[0028] Furthermore, the preprocessing of the data set to be processed includes the following steps:

[0029] Read the data to be processed;

[0030] Convert unstructured data into structured data text through Python pre-processing scripts;

[0031] Delete the blank text fields generated in the structured data text during the extraction process, and for individual abnormal text fields, select the basic stop word library and supplement some stop words.

[0032] Furthermore, the use of the TextRank algorithm to extract keywords from the preprocessed data set includes the following steps:

[0033] Text preprocessing: including word segmentation, stop word removal, part-of-speech tagging, and converting text into a format suitable for algorithm processing;

[0034] Constructing a graph model: taking the words or sentences in the processed text as nodes, and constructing a graph model based on the relationships between the words or sentences;

[0035] Calculate node weight: Use the iterative idea of ​​the PageRank algorithm to calculate the weight of each node in the graph;

[0036] Sorting and extraction: Sort by node weights, and use the node with the second preset value before sorting as the keyword of the text.

[0037] Furthermore, the relationship between the nodes includes co-occurrence relationship and semantic similarity, and the weight of the node is determined by the weights of other nodes and the strength of the relationship between them.

[0038] Furthermore, the first preset time period is greater than the second preset time period, the first preset time period is nearly five years, and the second preset time period is nearly one year.

[0039] On the other hand, the present invention provides an intelligent auxiliary system for selecting topics for scientific research and innovation, including: a data acquisition module, a data classification module, a data extraction module, a data preprocessing module, a keyword extraction module, a hot word and new word processing module, an input module and a visualization module. The keyword extraction module is used to extract keywords from the preprocessed data set through the TextRank algorithm. The hot word and new word processing module is used to perform statistical analysis on all topics in each field by publication time and research topic, and obtain a hot word list and a new word list in each field. The input module is used to obtain the field of interest input by the user, and send the field of interest to the visualization module. The visualization module is used to receive the field of interest sent by the input module, and display the recent hot topic list of the corresponding field to the user through visualization technology according to the field of interest.

[0040] The present invention also provides a non-transitory computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the intelligent auxiliary method for selecting a topic for scientific research innovation as described in any one of claims 1 to 8.

[0041] Compared with the prior art, the present invention has the following advantages:

[0042] The present invention extracts keywords from scientific and technological literature through the TextRank algorithm, and obtains a list of recent hot topics by statistically sorting the keywords. In this way, it assists scientific researchers in selecting topics for scientific research innovation, gives full play to the auxiliary decision-making role of artificial intelligence and big data technology in scientific research topic selection, provides effective tools for scientific and technological innovation, and quickly screens out the most potential and promising topics from massive scientific research literature, data and trends, so that scientific researchers and innovation teams can choose research directions more specifically, avoid blind exploration, and improve the efficiency of scientific and technological innovation and the quality of results. BRIEF DESCRIPTION OF THE DRAWINGS

[0043] Figure 1 Flow chart of the method of the present invention DETAILED DESCRIPTION

[0044] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work should fall within the scope of protection of the present invention.

[0045] Embodiment 1:

[0046] This embodiment provides an intelligent auxiliary method for selecting research and innovation topics. Figure 1 As shown, the following steps are included:

[0047] The scientific and technological documents and their publication time are obtained as the original data set, and the scientific and technological documents include: academic journals, academic papers, conference papers, scientific and technological achievements, patent documents and technical standard documents.

[0048] The original data set is classified by literature form to obtain paper data set, patent data set and technical standard data set;

[0049] According to the China Library Classification System, the paper datasets are classified into different fields to obtain paper datasets in different fields.

[0050] Classify the patent dataset by field according to patent type to obtain patent datasets in various fields;

[0051] Classify the technical standard data sets by field according to the standard type to obtain the technical standard data sets in various fields;

[0052] The paper dataset includes academic journals, academic papers, conference papers and scientific and technological achievements; the patent dataset includes patent documents; and the technical standard dataset includes technical standard documents.

[0053] The paper datasets, patent datasets and technical standard datasets in various fields are merged by field categories to obtain datasets in various fields;

[0054] Extract titles and abstracts from paper datasets and patent datasets in various fields, extract titles and scopes from technical standard datasets, and use the extracted data as the dataset to be processed in this field;

[0055] Preprocessing the data set to be processed, wherein the data to be processed includes unstructured data such as word and pdf;

[0056] The preprocessing of the data set to be processed includes the following steps:

[0057] Write relevant python pre-processing scripts;

[0058] Read the data to be processed;

[0059] Convert unstructured data into structured data text through Python pre-processing scripts;

[0060] Delete the blank text fields generated in the structured data text during the extraction process, and for individual abnormal text fields, select the basic stop word library and supplement some stop words.

[0061] The method of extracting keywords from the preprocessed data set using the TextRank algorithm comprises the following steps:

[0062] Text preprocessing: including word segmentation, stop word removal, part-of-speech tagging, and converting text into a format suitable for algorithm processing;

[0063] Constructing a graph model: taking the words or sentences in the processed text as nodes, and constructing a graph model based on the relationships between the words or sentences;

[0064] Calculate node weight: Use the iterative idea of ​​the PageRank algorithm to calculate the weight of each node in the graph. The relationship between the nodes includes co-occurrence relationship and semantic similarity. The weight of the node is determined by the weights of other nodes and the strength of the relationship between them.

[0065] Sorting and extraction: Sort by node weights, and use the node with the second preset value before sorting as the keyword of the text.

[0066] The extracted keywords are used as the research topics of their corresponding scientific and technological literature;

[0067] Statistical analysis is performed on all topics in each field by publication time and research topic to obtain a hot word list and a new word list in each field, wherein the hot word list includes the literature topics and their publication frequency published within a first preset time period, and the new word list includes the literature topics and their publication frequency only published within a second preset time period, the first preset time period is greater than the second preset time period, the first preset time period is nearly five years, and the second preset time period is nearly one year.

[0068] Sort the hot word list and new word list according to the frequency of publication;

[0069] Get the user input of the field of interest;

[0070] Intercept the document subject and its publication frequency at the first preset value position before the new word list in the field of interest, find the publication frequency of its document subject in the hot word list in the field of interest, add the frequency to the intercepted list, and obtain a preliminary list of recent hot topics;

[0071] Sorting the primary list of recent hot topics by the frequency of publication in the first preset time period to obtain a list of recent hot topics;

[0072] Through visualization technology, a list of recent hot topics is displayed to users, intelligently assisting users in selecting topics for scientific research and innovation.

[0073] This embodiment mainly describes the specific implementation steps of an intelligent auxiliary method for selecting topics for scientific research innovation based on text mining and intelligent analysis. Step 1: Data collection and classification. First, the paper data sets, patent data sets and technical standard data sets in various fields are obtained through field classification. These data sets can be obtained from public literature databases, patent databases and technical standard databases. Then, these data sets are further classified according to patent type and standard type to obtain more detailed data sets. Finally, these data sets are merged by field categories to obtain data sets in various fields. Step 2: Data preprocessing. The data set to be processed is preprocessed, including steps such as converting it into structured data text, deleting blank text fields and abnormal text fields, and supplementing stop words. This step mainly uses Python programming language and some natural language processing (NLP) libraries, such as NLTK, TextBlob, etc. Specifically, the steps of converting unstructured data into structured data text include word segmentation, removing stop words, part-of-speech tagging, etc. The steps of deleting blank text fields and abnormal text fields are mainly implemented through regular expressions and string operations. The step of supplementing stop words is mainly to supplement some specific stop words from the existing stop word library according to specific fields and needs. Step 3: Keyword extraction. Use the TextRank algorithm to extract keywords from the preprocessed data set. The TextRank algorithm is an algorithm based on a graph model that can effectively capture keywords in the text. This step mainly uses the Python programming language and some graph algorithm libraries, such as NetworkX. Specifically, first, the words or sentences in the processed text are used as nodes, and a graph model is constructed based on the relationship between words or sentences. Then, the weight of each node in the graph is calculated using the iterative idea of ​​the PageRank algorithm. Finally, the nodes are sorted according to their weights, and the nodes with the second preset value position after sorting are used as keywords for the text. Step 4: Topic statistics and analysis. All topics in each field are statistically analyzed by publication time and research topic, and a list of hot words and a list of new words in each field are obtained. This step mainly uses the Python programming language and some data analysis libraries, such as Pandas, Matplotlib, etc. Specifically, first, the data sets in each field are sorted according to the publication time, and then the topic of each document is classified and counted according to the research topic. Finally, a list of hot words and a list of new words in each field are obtained. Step 5: Visualization display. The recent hot topic list is displayed to users through visualization technology. This step mainly uses Python programming language and some visualization libraries, such as Matplotlib, Seaborn, etc. Specifically, first sort the recent hot topic list by publication frequency, and then use pie charts, word clouds and other visualization methods to display it to users. Step 6: User interaction.Get the field of interest input by the user, and intercept the literature topic and its publication frequency of the first preset value position before the new word list in the field, find its publication frequency in the hot word list, and get the primary list of recent hot topics. Then, sort the primary list to get the recent hot topic list. This step is mainly implemented through the user interface (UI), so that users can easily enter the field of interest and view the corresponding results. Step 7: Intelligent assistance. Through intelligent analysis technology, assist in the selection of scientific research and innovation topics. This step mainly uses the Python programming language and some machine learning libraries, such as Scikit-learn, TensorFlow, etc. Specifically, first take the field of interest and new word list input by the user as input, and then use the machine learning model for prediction and analysis to obtain recommended research directions and topics. The above is the specific implementation steps of this embodiment. It should be noted that the specific parameters, tools, and libraries mentioned in this embodiment are exemplary and can be adjusted and replaced according to specific needs.

[0074] Embodiment 2:

[0075] The present embodiment also provides an intelligent auxiliary system for selecting topics for scientific research and innovation, including: a data acquisition module, a data classification module, a data extraction module, a data preprocessing module, a keyword extraction module, a hot word and new word processing module, an input module and a visualization module. The keyword extraction module is used to extract keywords from the preprocessed data set through the TextRank algorithm. The hot word and new word processing module is used to perform statistical analysis on all topics in each field by publication time and research topic, and obtain a hot word list and a new word list in each field. The input module is used to obtain the field of interest input by the user and send the field of interest to the visualization module. The visualization module is used to receive the field of interest sent by the input module, and display the recent hot topic list of the corresponding field to the user through visualization technology according to the field of interest.

[0076] This embodiment also provides an electronic device, including: a processor and a memory, wherein the processor is used to execute a processing program for a table in a version document stored in the memory to implement the above-mentioned method for deduplicating and extracting text elements.

[0077] This embodiment further provides a computer-readable storage medium on which a computer program is stored, and when the computer program is executed by a processor, the above-mentioned method is implemented.

[0078] If the above functions are implemented in the form of software functional units and sold or used as independent products, they can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention, or the part that contributes to the prior art, or the part of the technical solution, can be embodied in the form of a software product. The computer software product is stored in a storage medium, including several instructions for a computer device (which can be a personal computer, a server, or a network device, etc.) to perform all or part of the steps of the method described in each embodiment of the present invention. The aforementioned storage medium includes: U disk, mobile hard disk, read-only memory (ROM, Read-Only Memory), random access memory (RAM, Random Access Memory), disk or optical disk, etc., which can store program code.

[0079] The above is only a specific embodiment of the present invention, but the protection scope of the present invention is not limited thereto. Any technician familiar with the technical field can easily think of various equivalent modifications or replacements within the technical scope disclosed by the present invention, and these modifications or replacements should be included in the protection scope of the present invention. Therefore, the protection scope of the present invention shall be based on the protection scope of the claims.

Claims

1. An intelligent auxiliary method for selecting research and innovation topics, characterized in that: The following steps are involved: Obtain scientific and technological literature and its publication time as the original data set; The original data set is classified by literature form to obtain paper data set, patent data set and technical standard data set; According to the China Library Classification System, the paper datasets are classified into different fields to obtain paper datasets in different fields. Classify the patent dataset by field according to patent type to obtain patent datasets in various fields; Classify the technical standard data sets by field according to the standard type to obtain the technical standard data sets in various fields; The paper datasets, patent datasets and technical standard datasets in various fields are merged by field categories to obtain datasets in various fields; Extract titles and abstracts from paper datasets and patent datasets in various fields, extract titles and scopes from technical standard datasets, and use the extracted data as the dataset to be processed in this field; Preprocess the data set to be processed; Use the TextRank algorithm to extract keywords from the preprocessed data set; The extracted keywords are used as the research topics of their corresponding scientific and technological literature; Statistical analysis is performed on all topics in each field by publication time and research topic to obtain a hot word list and a new word list in each field, wherein the hot word list includes the topics of the literature published in the first preset time period and their publication frequency, and the new word list includes the topics of the literature published only in the second preset time period and their publication frequency; Sort the hot word list and new word list according to the frequency of publication; Get the user input of the field of interest; Intercept the document subject and its publication frequency at the first preset value position before the new word list in the field of interest, find the publication frequency of its document subject in the hot word list in the field of interest, add the frequency to the intercepted list, and obtain a preliminary list of recent hot topics; Sorting the primary list of recent hot topics according to the publication frequency in the first preset time period to obtain a list of recent hot topics; Through visualization technology, a list of recent hot topics is displayed to users, intelligently assisting users in selecting topics for scientific research and innovation.

2. The intelligent auxiliary method for selecting research and innovation topics according to claim 1, characterized in that: The scientific and technological documents include: academic journals, academic papers, conference papers, scientific and technological achievements, patent documents and technical standard documents.

3. The intelligent auxiliary method for selecting research and innovation topics according to claim 1, characterized in that: The paper dataset includes academic journals, academic papers, conference papers and scientific and technological achievements; the patent dataset includes patent documents; and the technical standard dataset includes technical standard documents.

4. The intelligent auxiliary method for selecting research and innovation topics according to claim 1, characterized in that: The data to be processed includes unstructured data such as word and pdf.

5. An intelligent auxiliary method for selecting research and innovation topics according to claim 1 or 4, characterized in that: The preprocessing of the data set to be processed includes the following steps: Read the data to be processed; Convert unstructured data into structured data text through Python pre-processing scripts; Delete the blank text fields generated in the structured data text during the extraction process, and for individual abnormal text fields, select the basic stop word library and supplement some stop words.

6. The intelligent auxiliary method for selecting research and innovation topics according to claim 1, characterized in that: The method of using the TextRank algorithm to extract keywords from the preprocessed data set includes the following steps: Text preprocessing: including word segmentation, stop word removal, part-of-speech tagging, and converting text into a format suitable for algorithm processing; Constructing a graph model: taking the words or sentences in the processed text as nodes, and constructing a graph model based on the relationship between the words or sentences; Calculate node weight: Use the iterative idea of ​​the PageRank algorithm to calculate the weight of each node in the graph; Sorting and extraction: Sort by node weights, and use the node with the second preset value before sorting as the keyword of the text.

7. The intelligent auxiliary method for selecting research and innovation topics according to claim 6, characterized in that: The relationship between the nodes includes co-occurrence relationship and semantic similarity, and the weight of the node is determined by the weights of other nodes and the strength of the relationship between them.

8. The intelligent auxiliary method for selecting research and innovation topics according to claim 1, characterized in that: The first preset time period is greater than the second preset time period, the first preset time period is nearly five years, and the second preset time period is nearly one year.

9. An intelligent auxiliary system for selecting research and innovation topics, characterized in that: include: Data acquisition module, data classification module, data extraction module, data preprocessing module, keyword extraction module, hot word and new word processing module, input module and visualization module. The keyword extraction module is used to extract keywords from the preprocessed data set through the TextRank algorithm. The hot word and new word processing module is used to perform statistical analysis on all topics in each field by publication time and research topic, and obtain hot word lists and new word lists in each field. The input module is used to obtain the field of interest input by the user and send the field of interest to the visualization module. The visualization module is used to receive the field of interest sent by the input module, and display the recent hot topic list of the corresponding field to the user through visualization technology according to the field of interest.

10. A non-transitory computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, it implements the intelligent auxiliary method for selecting a topic for scientific research and innovation as described in any one of claims 1 to 8.

Images