Component Classification of References

Abstract

References for books and articles require specific components. Identification of various components of the references is an important task in publishing industry. It also can facilitate the automatic categorization of the articles. This work is an attempt to represent the components of references using a list of parameters using various visual and syntactic features. The components are then classified using multi layer perceptron (MLP). The effectiveness of the proposed classifier is demonstrated on journal articles and is found to produce satisfactory result. It can be applied for other references as well. Moreover, other classification tools such as Bayes’ classifier or support vector machine (SVM) can be used in this regard.