An Implementation of Citation Parser Based On Sequence Alignment

Abstract

Most of the network services failed to provide exact set of data about the publishers of academic publications due to drastic increase in their number around the world. Here the main problem is that various publishers use various citation styles [1]. They format citation string in one of the thousands of different styles. So, it is difficult for the researchers to extract accurate metadata from thousands of existing citation styles. To avoid these difficulties in extracting metadata from citations, a parser called “Bibpro” is implemented in this project. The parser converts the components of a citation as its structural properties. The main principle of Bibpro parser is to create template database containing a set of sequence templates of different formats and then use the suitable sequence template to parse the given citation string[2]. It provides accuracy and quality data representation for academic publications. Using machine learning techniques this parser also provides efficient extraction templates for effective implementation.