Staff Profile Pages

Novel Deep Learning Approach for Scientific Literature Classification

Authors: Ahanger, M.M. and Wani, M.A.

Journal: Proceedings of the 2022 9th International Conference on Computing for Sustainable Global Development, INDIACom 2022

Pages: 249-254

ISBN: 9789380544441

DOI: 10.23919/INDIACom54597.2022.9763218

Abstract:

Millions of research articles published each year require enhanced methods for both access and management. Text classification methods based on classical machine learning algorithms perform poorly for tasks involving hundreds of possible categories of research articles. Text classification has been revolutionized by the use of deep learning models such as convolution neural networks and recurrent neural networks. This paper demonstrates the applicability of hybrid approaches to the task of classifying scientific literature. A new hybrid approach that uses multiple word embeddings and a hybrid architecture is proposed. The proposed approach is evaluated on seven datasets. The proposed approach outperforms other methods on the task.

Source: Scopus