Data di Pubblicazione:
2014
Abstract:
The Internet is naturally a simple and immediate mean to retrieve information. However, not everything one can find is equally ac- curate and reliable. In this paper, we continue our line of research towards effective techniques for assessing the quality of online content. Focusing on the Wikipedia Medicinal Portal, in a previous work we implemented an automatic technique to assess the quality of each article and we com- pared our results to the classification of the articles given by the portal itself, obtaining quite different outcomes. Here, we present a lightweight instantiation of our methodology that reduces both redundant features and those not mentioned by the WikiProject guidelines. What we obtain is a fine-grained assessment and a better discrimination of the articles' quality, w.r.t. previous work. Our proposal could help to automatically evaluate the maturity of Wikipedia medical articles in an efficient way.
Tipologia CRIS:
04.01 Contributo in Atti di convegno
Keywords:
Quality Assurance; Machine Learning
Elenco autori:
Marzini, Emanuel; Spognardi, Angelo; Conti, Riccardo; Petrocchi, Marinella; Mori, Paolo; Matteucci, Ilaria
Link alla scheda completa: