Publication

Abstract : Citation recommendation is a technique that assists an academic publisher or scholar in identifying a selection of relevant works that can be cited whilst writing a paper. In comparison with prior years, it is clear that there is currently a deluge of articles being published every year. As a result, the traditional way to identify valid citations appears to be a very difficult process, as there appears to be massive flow of data. In this research, we will discuss an innovative way to solve the citation recommendation task. To begin, we employed a probabilistic model known as the Gaussian mixture model, to minimise the search space, which we subsequently integrated with other neural architectures such as ANN and a hybrid model, CNN-LSTM. Finally, we performed a thorough experimental investigation employing a variety of evaluation metrics.