Corresponding Author

Hongzhou Shen, Nanjing University of Posts and Telecommunications, Nanjing, China,

Document Type

Work in Progress


The rigorousness, professionalism and seriousness of answer contents in social Q&A communities have declined sharply. It is of great significance to tell users how to provide excellent answer contents. The purpose of this research is to explore what features of answer content are more likely to make a high-quality answer. For this purpose, the research, taking “Zhihu” as an example, collects the data of answer contents with a crawler, combines initial analysis of the answer contents with related research literature, and selects 9 features that may have impacts on the quality of answer contents. Then the supervised machine learning method will be used to explore the features that can really affect the quality of answer contents, so as to provide reference for guiding users to provide high-quality answer contents.