Automatic Keyphrase Extraction describes the process of extracting keywords or keyphrases from the body of a document. To our knowledge until now all algorithms rely on a set of manually crafted statistical features to model word importance. In this paper we propose an end-to-end neural keyphrase extraction algorithm using a siamese LSTM network, eliminating the need for manual feature engineering. We train and evaluate our model on the Inspec [6] dataset for keyphrase extraction and achieve comparable results to state-of-the-art algorithms.
CITATION STYLE
Villmow, J., Wrzalik, M., & Krechel, D. (2018). Automatic keyphrase extraction using recurrent neural networks. In Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics) (Vol. 10935 LNAI, pp. 210–217). Springer Verlag. https://doi.org/10.1007/978-3-319-96133-0_16
Mendeley helps you to discover research relevant for your work.