Labeling topics with images using a neural network

Aletras, N. orcid.org/0000-0003-4285-1965 and Mittal, A. (2017) Labeling topics with images using a neural network. In: Jose, J.M., Hauff, C., Altıngovde, I.S., Song, D., Albakour, D., Watt, S. and Tait, J., (eds.) Advances in Information Retrieval : 39th European Conference on IR Research, ECIR 2017, Aberdeen, UK, April 8-13, 2017, Proceedings. 39th European Conference on IR Research (ECIR 2017), 08-13 Apr 2017, Aberdeen, UK. Lecture Notes in Computer Science (10193). Springer International Publishing , pp. 500-505. ISBN 9783319566078

Abstract

Topics generated by topic models are usually represented by lists of t terms or alternatively using short phrases or images. The current state-of-the-art work on labeling topics using images selects images by re-ranking a small set of candidates for a given topic. In this paper, we present a more generic method that can estimate the degree of association between any arbitrary pair of an unseen topic and image using a deep neural network. Our method achieves better runtime performance O(n) compared to O(n2) for the current state-of-the-art method, and is also significantly more accurate.

Metadata

Authors/Creators:	Aletras, N. https://orcid.org/0000-0003-4285-1965 Mittal, A.
Copyright, Publisher and Additional Information:	© 2017 Springer International Publishing. This is an author-produced version of a paper subsequently published in ECIR 2017 Proceedings. Uploaded in accordance with the publisher's self-archiving policy.
Keywords:	Topic models; Deep neural networks; Topic representation
Dates:	Published (online): 8 April 2017 Published: 2017
Institution:	The University of Sheffield
Academic Units:	The University of Sheffield > Faculty of Engineering (Sheffield) > Department of Computer Science (Sheffield)
Depositing User:	Symplectic Sheffield
Date Deposited:	14 Oct 2020 13:02
Last Modified:	15 Oct 2020 10:23
Status:	Published
Publisher:	Springer International Publishing
Series Name:	Lecture Notes in Computer Science
Refereed:	Yes
Identification Number:	https://doi.org/10.1007/978-3-319-56608-5_40
Related URLs:	Author

CORE (COnnecting REpositories)

Labeling topics with images using a neural network

Abstract

Metadata

Download

Accepted Version

Export

Statistics