A Comparison of Manual and Automatic Quality Assessment for Wikipedia Articles

Thesis Type Bachelor
Thesis Status
Student Ramona Huber
Thesis Supervisor
Research Field

Wikipedia is the biggest online encyclopedia in the world-wide web and is due its collaborative structure one of the five most popular websites in the world. Due to its popularity and its size, it is necessary for Wikipedia to provide its users with correct information. The Wikipedia platform assesses the article quality via its users, who edited and grade the articles. However, the user’s ratings can be subjective and different from user to user. This bachelor thesis combines 39 quality metrics with machine learning algorithms to achieve an automatic quality assessment for English Wikipedia articles. Moreover, the usage of feature selection will be depicted and the resulting subset of metrics will be compared to other models. The classification performance of those machine learning algorithms will be measured with known multi-class classification performance metrics.