Paper search method and device

Abstract

The invention discloses a method for searching for a paper. The method comprises the steps that search terms are received; a plurality of search results are generated according to the search terms and are provided, each search result comprises a plurality of attribute parameters, and at least part of the attribute parameters has corresponding entity identifiers; when the attribute parameters in the search results are triggered, new search results are generated according to the entity identifiers corresponding to the attribute parameters and are provided. According to the method, the problem of partial matching of entity name duplication and long search terms is thoroughly solved, the accuracy of the search results is improved, and user experience is improved. The invention further discloses a device for searching for the paper.

Description

technical field [0001] The invention relates to the technical field of search engines, in particular to a method and device for searching papers. Background technique [0002] At present, in the field of academic research, the research results of scholars / students / research workers on a certain knowledge will eventually be published in the form of papers, and the published papers themselves include many attributes, such as the name of the author of the paper, the place where the paper was published (including journals, conferences, dissertations, etc.). However, the same attributes of multiple papers form their own attribute sets, and the association between the paper collection and attribute sets forms the knowledge graph of the paper. Users can view other related attributes through any attribute in the graph. attribute information, but because of the huge number of existing papers, the number of attribute sets formed is also large, and there are a large number of attribute...

AI Technical Summary

Problems solved by technology

[0004] However, there are problems in related technologies: (1) disambiguation of authors with the same name cannot be realized. For example, when a user clicks on the author "Li Ming" as shown in Figure 1(a) to initiate a search, the user hopes to find the "Study on the Allelopathic Effects of Pumpkin Bacon Root Exudates" related to the results of the publisher "Li Ming", not "Li Ming" from other institutions

However, through the search method in the related technology, the "Li Ming" in the search results obtained comes from various fields, and is completely different from the target Li Ming, which cannot meet the user's precise name query needs; (2) cannot solve the long query ( search terms), for example, as shown in Figure 1(b), use "journal: (Journal of Ecology)" to search for papers published in "Journal of Ecology", and you can see that "Journal of Applied Ecology" will also be detected , the traditional search based on keyword matching cannot solve the partial matching problem, resulting in inaccurate search results and poor user experience

Images