Retrieval query method and system oriented to science and technology resource discipline and research theme information

Abstract

The invention provides a retrieval query method and system oriented to science and technology resource discipline and research theme information, and the method comprises the steps: obtaining keywords of science and technology resource information data, and constructing a keyword library; generating a word vector of the keyword and a word vector of the subject name of each subject through a BERT pre-training model; performing clustering analysis on the keywords through a clustering algorithm based on the word vectors of the subject names and the word vectors of the keywords; obtaining achievement publishing historical data under the research themes to which the keywords belong in a historical predetermined time period, and predicting achievement publishing development trends of the research themes to which the keywords belong in a future predetermined time period according to the achievement publishing historical data; determining an influence index of each research theme according to the number of achievements under the research themes to which various keywords belong in the future predetermined time period; and according to the achievement publication development trends of the research themes and the influence indexes of the research themes, determining an arrangement sequence of the research themes, and according to the arrangement sequence, determining a retrieval query result.

Description

technical field [0001] The invention relates to the technical field of natural language processing, in particular to a retrieval method and system for science and technology resource disciplines and research subject information. Background technique [0002] In recent years, with the continuous development of various information retrieval platforms, more and more users tend to find the scientific and technological resource information they need on the retrieval platform. develop. Under such circumstances, when users want to browse and refer to scientific and technological resource information in a certain field, such as natural foundations, academic papers, patent information, etc., they will search different resource platforms and systems. However, these platforms have the problem of high content repetition in terms of the attributes of scientific and technological resources. Users need to retrieve, manually integrate and filter these data one by one to obtain the desired ...

AI Technical Summary

Problems solved by technology

However, due to the uniqueness of scientific and technological resources, traditional theories and models are difficult to directly apply to the analysis and processing of information and data of scientific and technological resources with multi-source heterogeneous characteristics

For example, in a scientific research field, there are both published papers and applied for natural fund projects, as well as relevant news and news. The attributes of these information are different, but they are all closely related to the scientific research field. In addition, due to the interdisciplinary nature of disciplines, the information may belong to different disciplines; and the current traditional theories and models cannot accurately retrieve the disciplines and research topics hidden in the scientific and technological resource information for the keywords entered by users, so It is impossible to realize efficient retrieval and query of science and technology resource disciplines and research subject information

Images