A Psycho-Lexical Approach to the Assessment of Information Quality on Wikipedia

The great popularity of Wikipedia makes it one of the dominant knowledge source around the World. However, since one of the core principles of Wikipedia is being open for anyone to maintain it, Wikipedia cannot fully ensure the reliability of its articles, and thus sometimes suffered criticism for containing low-quality information. It is therefore essential to assess the quality of Wikipedia articles automatically. In this paper we describe how we approach that problem by using a psycho-lexical resource, i.e., the Language Inquiry and Word Count (LIWC) dictionary. By training a classifier on different LIWC categories, we discuss the implications of each category for Wikipedia quality assessment.