Authorship verification is the task of determining whether written documents have been authored by a specific person. This is a very complex task that, in order to be solved properly, requires of a careful analysis of the writing style of authors. Automatic methods for authorship verification use a sample of documents and, by means of machine learning techniques, generate a coarse model of authors’ writing style. While, promising results have been obtained with these techniques in general scenarios, online education poses additional complications that require of tailored authorship verification techniques.
Perhaps the biggest challenge for automatic authorship verification techniques in the context of TeSLA is the scarcity of samples. In general, the performance of automatic methods is directly related with the number of samples used to build the authorship model. In online learning environments, however, collecting samples through enrollment activities is not easy. For instance, it is unrealistic to ask students to provide a few dozens of sample documents to build an accurate authorship verification instrument. Instead, instruments have to be adapted to work with a few sample documents at the beginning and to implement incremental learning and adaptive mechanisms that allow them to improve the model with new samples that authors can provide on the fly.
Another major challenge has to do with the lack of agreement between sample documents used to build the model and the documents that the instrument has to analyze when it is in operation. Enrollment activities usually require students to write about anything to generate samples. Whereas writing style is not dependent on the thematic content, students use a different language structure when writing informal vs formal documents. Therefore, author verification methods should be robust to this mismatch in the writing style of documents. Other challenges have to do with the type of documents (e.g., essay like vs math documents), their length, and learning-activity specific characteristics.
Although there are several challenges inherent to the online education scheme, authorship verification is critical to guarantee a trustable assessment of learning activities. In this context, TeSLA’s authorship verification instrument is able to deal with the associated difficulties, thus guaranteeing its effectiveness for the online education scenario.
Hugo Jair Escalante, Manuel Montes, Pastor López (INAOE team)
FUNDED BY THE EUROPEAN UNION
TeSLA is not responsible for any contents linked or referred to from these pages. It does not associate or identify itself with the content of third parties to which it refers via a link. Furthermore TESLA is not liable for any postings or messages published by users of discussion boards, guest books or mailing lists provided on its page. We have no control over the nature, content and availability of any links that may appear on our site. The inclusion of any links does not necessarily imply a recommendation or endorse the views expressed within them.
TeSLA is coordinated by Universitat Oberta de Catalunya (UOC) and funded by the European Commission’s Horizon 2020 ICT Programme. This website reflects the views only of the authors, and the Commission cannot be held responsible for any use which may be made of the information contained therein.