Title :
Reputation on Stack Exchange: Tag, You´re It!
Author_Institution :
Dept. of Comput. Sci., Univ. of Victoria, Victoria, BC, Canada
Abstract :
Question and Answer sites like the Stack Exchange network allow users to contribute knowledge in a variety of topics. Of particular interest are the habits of expert users, or users with high reputation scores. Expanding on previous work done by Bosu et al., we performed an exploratory analysis of the data to determine if there is a positive correlation between user reputation scores and the diversity of tags a user contributed to. For our analysis, we used data available from Stack Exchange to created a visual network of user tagging habits. We found that there is a correlation between these two measures, suggesting that expert users contribute to a wide variety of tags. Our research also shows poor community structures in the networks studied. This is consistent with the prevailing literature on Stack Exchange use and the power law. Our findings confirms one of the recommendations put forward by Bosu et al. Through the use of a much larger population and put forward ideas for new areas of exploration on question and answer sites.
Keywords :
data analysis; question answering (information retrieval); Stack Exchange network; exploratory data analysis; question-and-answer sites; tags diversity; user reputation scores; user tagging habits; Communities; Correlation; Data analysis; Data mining; Sociology; Tagging; Community Detection; Software Engineering; Stack Overflow; Tagging;
Conference_Titel :
Advanced Information Networking and Applications Workshops (WAINA), 2014 28th International Conference on
Conference_Location :
Victoria, BC
Print_ISBN :
978-1-4799-2652-7
DOI :
10.1109/WAINA.2014.108