About Wikability

For language learners, Wikipedia could be a great tool for learning. However, there is no existing Wikipedia article recommendation system that takes user’s reading level into consideration. Also, traditional text recommendation systems evaluate readers’ reading levels and compare them to that of the text to make level-appropriate recommendations. These systems focus solely on the textual features such as sentence length and frequency of words to determine readability, without regards for inter-document similarity.

Wikability aims to address these weaknesses of existing text recommendation system by providing an interactive graphical tool that combines the calculated textual feature-based readability metrics with content similarity metrics to recommend Wikipedia articles that are reading-level and topic-appropriate. Considering the wide user base of Wikipedia and based on research that shows the benefit of visualization of document clustering, Wikability will offer a new dimension by which language learners and related personnel can discover challenging yet comprehensible articles that are similar to the articles that they had found interesting.