University of Leicester
Browse
GaussianProduct Hossein.pdf (239.58 kB)

GaussianProductAttributes: Density-Based Distributed Representations for Products

Download (239.58 kB)
Version 2 2022-04-07, 08:03
Version 1 2022-03-31, 09:43
conference contribution
posted on 2022-04-07, 08:02 authored by HG Noushahr, J Levesley, S Ahmadi, E Mirkes
Multivariate Gaussian probability distributions have been used as distributed representations for text. In comparison with traditional vector representations, these density-based representations are able to model uncertainty, inclusion and entailment. We present a model to learn such representations for products based on a public e-commerce dataset. We qualitatively analyse the properties of the proposed model and how the learned representations capture semantic relatedness, similarity and entailment between products and text.

History

Citation

GaussianProductAttributes: Density-Based Distributed Representations for Products. In: Bramer, M., Ellis, R. (eds) Artificial Intelligence XXXVIII. SGAI-AI 2021. Lecture Notes in Computer Science(), vol 13101. Springer, Cham. https://doi.org/10.1007/978-3-030-91100-3_11

Author affiliation

Department of Mathematics

Version

  • AM (Accepted Manuscript)

Published in

Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)

Volume

13101 LNAI

Pagination

139 - 145

Publisher

Springer International Publishing

issn

0302-9743

eissn

1611-3349

isbn

9783030910990

Acceptance date

2021-08-31

Copyright date

2022

Available date

2022-12-06

Language

en

Usage metrics

    University of Leicester Publications

    Categories

    Exports

    RefWorks
    BibTeX
    Ref. manager
    Endnote
    DataCite
    NLM
    DC