Feature Analysis Of Academic Misconduct Events Based On Text Mining

With the improvement of scientific research level,the phenomenon of academic misconduct is becoming more and more serious,which has become an urgent problem to be solved all over the world.The popularity of the Internet makes it easier to expose and spread academic misconduct at home and abroad.Therefore,the Internet has become an important source of information and data in the process of studying academic misconduct.However,the rapid growth of the amount of information and the continuous reduction of the threshold of media platform lead to the complexity of the content on the network,which is difficult to analyze directly.Based on this background,this paper analyzes and studies the characteristics of academic misconduct by taking the relevant information released by the media platform as the research object.This paper aims to collect information about academic misconduct events,collecting news,announcements,forum posts and commentary related to academic misconduct,which are published on 3 mainstream media websites,10 domestic media websites and 9 official account numbers.The text data can be filtered through the text classification of "academic misconduct related" and "academic misconduct unrelated";Tsinghua Chinese text classification toolkit THUCTC is used to classify academic misconduct;This paper uses named entity recognition method to extract the information of people and countries involved in academic misconduct events,and mine the attributes of the identified people,such as professional title and affiliated units;Using the rule-based matching algorithm to extract the names of journals involved in academic misconduct events,and mining the journal attributes;According to the professional category of the journal,the text data of academic misconduct related events are divided into disciplines,and finally the feature extraction and analysis of complex information are realized.Based on the statistical analysis of the characteristics of academic misconduct,this paper analyzes the causes and effects of academic misconduct from the perspectives of time,behavior,people,country,journal and discipline,and makes a comparative analysis of academic misconduct at home and abroad.Research findings:(1)The behavior of academic misconduct tends to be complicated and no longer simply shows as plagiarism.In terms of time distribution,the number of references to academic misconduct of "forgery" increased rapidly,and in many years,it was in the category with the most academic misconduct;From the behavior proportion,the proportion of "tampering" and "forgery" forms of misconduct is larger.Researchers are gradually focusing on the research of "hidden academic misconduct" such as tampering and plagiarism of charts and formulas,cross language plagiarism,data tampering;(2)The articles collected in the top international journals are likely to be academic misconduct,while domestic high-quality journals are not easy to have academic misconduct.Due to the small number of papers issued by domestic high-quality journals,journal editors will carefully choose the papers to submit.Meanwhile,high-quality journals attach importance to peer expert review,which makes academic misconduct effectively contained;The existence of predatory journals and the collection of layout fees connive the occurrence of academic misconduct.(3)Academic misconduct in biomedical field is more serious at home and abroad.Its special disciplinary nature and current talent system make it difficult to curb academic misconduct.In domestic disciplines,"R medicine and health" was first mentioned academic misconduct in 2004,and fluctuated greatly in recent 18 years,with four years being the highest mentioned fields in the same period;In the field of international disciplines,since 2013,in addition to "multidisciplinary Sciences",medicine is the field with the most academic misconduct mentioned every year.In addition,among the six disciplines most mentioned in academic misconduct,"medicine","general &International","Biochemistry &Molecular Biology" and "Oncology" are all related to biomedicine.The purpose of this paper is to understand and grasp the characteristics of academic misconduct events published by domestic and foreign media,and analyze their development rules and internal reasons,so that the relevant departments can timely find the problems in the process of scientific research policy-making,and then make strategic adjustments,so as to avoid the recurrence of similar events and affect China’s academic reputation and scientific research level.