Classi cation of Scienti c Papers Using Machine Learning Minh

The project aims to develop a domain-independent and adaptive approach for scienti c document classi cation using both information from document contents and citation links. We evaluate several content-based classi cation methods including K-nearest neighbours, nearest centroid, naive Bayes and decision trees and nd that the naive Bayes outperform other when training set is sufciently large. Using phrases in addition to words and a good feature selection strategy such as information gain is found to improve system accuracy in comparison with using words only. To combine citation links for classi cation, the project proposes two methods, linear labelling update and probabilistic labelling update. The two methods iteratively update the labellings of classi ed documents using categories information from neighbouring documents. Our experiments on the two methods show that, combining contents and citations signi cantly improves the system performance.