|
|
Frontiers of Information Technology & Electronic Engineering
ISSN 2095-9184 (print), ISSN 2095-9230 (online)
2021 Vol.22 No.2 P.170-184
Learning natural ordering of tags in domain-specific Q&A sites
Abstract: Tagging is a defining characteristic of Web 2.0. It allows users of social computing systems (e.g., question and answering (Q&A) sites) to use free terms to annotate content. However, is tagging really a free action? Existing work has shown that users can develop implicit consensus about what tags best describe the content in an online community. However, there has been no work studying the regularities in how users order tags during tagging. In this paper, we focus on the natural ordering of tags in domain-specific Q&A sites. We study tag sequences of millions of questions in four Q&A sites, i.e., CodeProject, SegmentFault, Biostars, and CareerCup. Our results show that users of these Q&A sites can develop implicit consensus about in which order they should assign tags to questions. We study the relationships between tags that can explain the emergence of natural ordering of tags. Our study opens the path to improve existing tag recommendation and Q&A site navigation by leveraging the natural ordering of tags.
Key words: Question and answering (Q&A) sites, Tagging, Natural order, Skip gram
贾俊芳1，李国强2
1山西大同大学计算机与网络工程学院，中国大同市，037009
2上海交通大学软件学院，中国上海市，200240
摘要：标注是Web 2.0的一个重要特征。它使得社会计算系统（如问答网站）的用户们可以自由地标记内容。然而，标注真的是自由不受限的吗？现有工作表明，用户们常常可以隐性地就哪种标签最能描述在线社区的内容达成共识。然而，目前还没有针对用户在标注过程中对标签排序的规律性开展研究。本文专注于研究特定领域问答网站中的标签自然排序，并对CodeProject，SegmentFault，Biostars以及CareerCup 4个问答网站上数以百万计的问题中的标签序列进行研究。结果表明，这些问答网站的用户可以就问题标签的排序达成隐性共识。研究了标签之间的关系，这些关系可以解释标签自然顺序的出现。该研究为利用标签的自然顺序提升现有标签推荐以及问答站点导航提供了可能。
关键词组：
References:
Open peer comments: Debate/Discuss/Question/Opinion
<1>
DOI:
10.1631/FITEE.1900645
CLC number:
TP182
Download Full Text:
Downloaded:
2796
Download summary:
<Click Here>Downloaded:
1198Clicked:
4673
Cited:
0
On-line Access:
2021-02-01
Received:
2019-11-24
Revision Accepted:
2020-02-12
Crosschecked:
2020-08-19