Font Size: a A A

Research On Answer Quality Judgement On Academic Social Q&A Sites

Posted on:2019-07-07Degree:DoctorType:Dissertation
Country:ChinaCandidate:L LiFull Text:PDF
GTID:1368330602961121Subject:Management Science and Engineering
Abstract/Summary:PDF Full Text Request
The traditional way of obtaining academic information is mainly through academic journals,academic essays or academic monographs.Their quality can be easily obtained through peer review,cited information,and the authority of the publisher.With the prosperity and development of social media,new ways of acquiring academic information have emerged,namely through general social media websites(such as forums,microblogs,blogs)or specialized academic social media websites(such as:ResearchGate.net,Academia.Edu)to share and access academic information.More and more scholars have expressed their willingness to share and obtain academic information on social media.Therefore,various types of social media are filled with a large amount of academic user-generated content.However,because of the lack of peer review,the quality of the academic information on it is uneven,which affects the scholars' satisfaction with the use of social media for academic information exchange and acquisition.Therefore,the quality of academic user generated content on social media is a problem that needs to be solved urgently.As one of the types of social media platforms,social question and answer(Q&A)can support scholars or researchers to exchange academic information by presenting academic questions and answers.Because of the low cost,quick communication,and easy establishment of social capital,social Q&A has become more and more popular among scholars.ResearchGate Q&A is a well-known social Q&A platform with the largest number of users dedicated to supporting scholars to conduct academic questions and answers.A wealth of research is available to explore how users judge the quality of various user-generated content on social media.In theresearch on general social Q&A sites,such as Yahoo!Answers,Baidu knows,etc,how users evaluate the quality of answers to identify high-quality answers is also an important research topic.With the emergence of the academic social Q&A platforms,the answer on it as a new type of user-generated content,and how scholars evaluate its quality also needs to be an important research topic.However,there is a lack of research on this issue in the existing research.This paper aims to fill this gap.Therefore,this dissertation will use the answers on the academic social Q&A platform-ResearchGate Q&A as the research object.Based on the research process of finding problems and solving problems,a series of two phases of researches will be used to study this issue.The first phase of the research was the exploration of problems and the status exploring,including two sub-studies.First,the objective characteristics of the answer on ResearchGate Q&A were analyzed,and the obtained objective characteristics of the answer on the ResearchGate Q&A were compared with the objective features of the answer on the general social Q&A.It was found that the objective characteristics of the answers on the general social Q&A were very different from the objective characteristics of the answers on the academic social Q&A,and the answers on the academic social Q&A also had their own unique objective characteristics.Then,in the first research phase,I tried to only use the objective characteristics of the answer to predict the high-quality answers recommended by the scholars on ResearchGate Q&A.However,it was found that only using the objective evaluation criteria based on the characteristics of the answer itself could not accurately predic the quality of the answer.There was a need to explore more effective model for answer quality evaluation.The second research stage was problem solving and model construction.It will directly obtain various subjective and objective evaluation criteria for evaluating the quality of answers from the user's point of view,and realized a more accurate answer quality evaluation model.This phase included three sub-studies.First of all,it was to design question items that scholars may use to evaluate the quality of answers on the academic social Q&A.Since the existing researches do not provide the criteria for the quality evaluation of the answers on the academic social Q&A,the question items were first constructed through the survey of scholars to obtain the evaluation criteria used in evaluating the answers in their subject areas.Therefore,combined with the evaluation criteria obtained by the user survey and the quality evaluation criteria proposed in the existing quality researches,the question items for the quality evaluation criteria of the answer on the academic social Q&A were constructed.Then,through the questionnaire survey of scholars who had actually recommended high-quality answers on ResearchGate Q&A,the scholars scored the importance of each evaluation criteria.This section mainly conducted basic statistical analysis and evaluation criteria model construction on the questionnaire data.The basic statistical analysis obtained the important evaluation criteria used by scholars to evaluate the quality of the answers on the academic social Q&A,and further used the difference analysis method to find out the demographic characteristics and the subjects of the scholars that would influence the importance of each evaluation criteria.In addition,this part finally uses the principal component analysis method to construct an evaluation model for evaluating the quality of answers on the academic social Q&A that integrates various subjective and objective evaluation criteria.The last part of the research was to verify the evaluation model obtained above.In the case of limiting external influence factors,this study employed scholars who had the same demographic characteristics as possible to verify the constructed model.The results proved that the model obtained in this study can effectively identify higher quality answers.This dissertation inherits the research route from finding problems to solving problems,and carries out a series of relatively complete researches on how scholars evaluate the quality of answers on academic social Q&A sites,making up for the lack of research on this issue.At the same time,this study also provides support for academic social Q&A platforms to recommend high-quality answers for scholars and improve the platform quality recommendation mechanism design,thereby saving scholars time for obtaining high-quality information.The ultimate goal is to increase academics' satisfaction with the use of academic social Q&A platforms and scholars' research work efficiency,and promote the exchange and sharing of academic information on social media.
Keywords/Search Tags:Answer quality judgement, Subjective and objective judgement criteria, Quality prediction of answers, Academic social Q&A, Academic social media
PDF Full Text Request
Related items