Font Size: a A A

Application Of Author-topic Model In Interdisciplinary Research

Posted on:2021-02-14Degree:MasterType:Thesis
Country:ChinaCandidate:M R WangFull Text:PDF
GTID:2439330611969761Subject:Applied statistics
Abstract/Summary:PDF Full Text Request
With the advancement of sciences and technologies and popularization of the Internet,people from different academic disciplines have more opportunities to work together and exchange ideas.Studies have also shown that interdisciplinary research,now also supported by national policies and funds,is often more likely to generate scientific and technological innovation.In addition to promoting technological innovation and development of knowledge in subjects,interdisciplinary studies have also become features of the time.Specifically,measurement and analysis of interdisciplinarity in a topic allow focusing on a specific area of research studies,encouraging researchers to collaborate closely with each other and bring in different perspectives to tackle real-world problems.More importantly,they could help scientists get inspiration from the crossing of multiple subjects,overcome challenges in researches,and resolve technology barriers.This article contains four parts.The first part introduces the history of interdisciplinary research from China and overseas.It also presents a summary of the research frameworks for interdisciplinary studies and an overview of measurements of interdisciplinarity.Based on the research objects sorted out,it then expands the dimensions of research—the topic dimension,and analyzes related technologies of topic recognition.The second part builds a dataset from the field of gene editing,selects indicators of the interdisciplinary nature of a topic,and establish an indicator system.The third part,considering the imbalance of articles mapped to corresponding disciplines in multi-disciplinary journals,introduces the subject-topic model into interdisciplinary research and conducts topic mining on data from the field of gene editing.Subject modeling uses the perplexity index as a standard to measure the generalization ability,from which the topic-subject matrix and the subject term matrix are obtained.The fourth part first calculates the similarity between subjects using the subject-topic model and compares it with the result obtained from the model based on the citation relationship.Next,the correlation analysis of measurement of interdisciplinarity leads to the establishment of the TD index of the ST model and the DIV* index based on the references from different subjects.Finally,it investigates the subjects with strong interdisciplinarity in the field of gene editing and analyzes the distribution of disciplines involved in each topic.The fourth part first calculates the similarity between subjects using the subject-topic model and compares it with the result obtained from the model based on the citation relationship.Next,calculate and analyze the interdisciplinarity of topics,and excavate the strong interdisciplinary topics in the field of gene editing.What’s more,the results of the research impact show that,the correlation analysis of measurement of interdisciplinarity leads to the establishment of the TD index of the ST model and the DIV* index based on the references from different subjects.Finally,combination of both,it investigates the subjects with strong interdisciplinarity in the field of gene editing and analyzes the distribution of disciplines involved in each topic.Through empirical research,this article can provide references for policymakers and research directions for scientific researchers.At the same time,this paper provides a new measurement method for calculating the similarity among subjects,enriching the theoretical research of interdisciplinarity.Finally,transferring the research angle to the topic level to conduct interdisciplinary knowledge diffusion measurement research,it expands the dimensions of interdisciplinary research objects.
Keywords/Search Tags:interdisciplinarity, gene editing, subject-topic model, topic mining
PDF Full Text Request
Related items