Page 146 - JOURNAL OF LIBRARY SCIENCE IN CHINA 2018 Vol. 42

Types and description rules of knowledge elements

about methods in academic papers ①a

HUA Bolin ①b *
Department of Information Management, Peking University, Beijing 100871, China

Abstract
In academic papers, there are many knowledge elements about the method. How to extract these
methods to form a structured knowledge base is an important research of knowledge organization.
Through content analysis of a large number of documents, method knowledge elements are classified
as the following five types: knowledge element of method definition, knowledge element of method
classification, knowledge element of method features, knowledge element of method process, and
knowledge element of method functions. Sentences of the method are extracted from academic papers,
then the sentence description structure is formed by filtering keywords in the sentences, and the
description rules of method are formed after revising and classification in order to extract knowledge
elements of the method.

Keywords
Academic papers, Knowledge element of method, Knowledge mining, Pattern recognition

0 Introduction

As an important medium of knowledge, the academic paper reflects the development of science
and technology innovation in academic papers, and descriptions of the method are important
types of scientific knowledge. As the number of academic paper grows rapidly, manual processing
alone is not enough for method detection and the construction of the method knowledge base.
Today we rely more and more on information technology. In order to quickly and effectively
construct a comprehensive method knowledge base, we need to make full use of technology to
extract and mine valuable information and knowledge from the massive amount of literatures.
Aiming at automatic or semi-automatic extraction of method knowledge elements from papers,
types and description rules of method knowledge elements should be analyzed and studied more
deeply.
Methods include not only general methods like questionnaire, expert interviews, case studies,
but also bibliometric analysis, cluster analysis, correlation analysis, co-occurrence analysis, multi-
objective decision analysis, etc. These methods appear commonly in academic papers. Therefore,

①a Special contribution for the Youth Academic Forum sponsored by this Journal.
This article is an outcome of the project “Mining and Discovery of Knowledge Cell from Academic Papers” (No. 2014M560857)
supported by China Postdoctoral Science Foundation.
* Correspondence should be addressed to HUA Bolin, Email: huabolin@pku.edu.cn，ORCID: 0000-0001-9248-6455