Skip to main navigation Skip to search Skip to main content

Interactive area topics extraction with policy gradient

  • Jingfei Han
  • , Wenge Rong*
  • , Fang Zhang
  • , Yutao Zhang
  • , Jie Tang
  • , Zhang Xiong
  • *Corresponding author for this work
  • Beihang University
  • Tsinghua University

Research output: Chapter in Book/Report/Conference proceedingConference contributionpeer-review

Abstract

Extracting representative topics and improving the extraction performance is rather challenging. In this work, we formulate a novel problem, called Interactive Area Topics Extraction, and propose a learning interactive topics extraction (LITE) model to regard this problem as a sequential decision making process and construct an end-to-end framework to use interaction with users. In particular, we use recurrent neural network (RNN) decoder to address the problem and policy gradient method to tune the model parameters considering user feedback. Experimental result has shown the effectiveness of the proposed framework.

Original languageEnglish
Title of host publicationArtificial Neural Networks and Machine Learning – ICANN 2018 - 27th International Conference on Artificial Neural Networks, 2018, Proceedings
EditorsYannis Manolopoulos, Barbara Hammer, Vera Kurkova, Lazaros Iliadis, Ilias Maglogiannis
PublisherSpringer Verlag
Pages84-93
Number of pages10
ISBN (Print)9783030014230
DOIs
StatePublished - 2018
Event27th International Conference on Artificial Neural Networks, ICANN 2018 - Rhodes, Greece
Duration: 4 Oct 20187 Oct 2018

Publication series

NameLecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Volume11141 LNCS
ISSN (Print)0302-9743
ISSN (Electronic)1611-3349

Conference

Conference27th International Conference on Artificial Neural Networks, ICANN 2018
Country/TerritoryGreece
CityRhodes
Period4/10/187/10/18

Keywords

  • Interactive area topics extraction
  • Policy gradient
  • RNN decoder

Fingerprint

Dive into the research topics of 'Interactive area topics extraction with policy gradient'. Together they form a unique fingerprint.

Cite this