Skip to main navigation Skip to search Skip to main content

Classification over Clustering: Augmenting Text Representation with Clusters Helps!

  • Xiaoye Tan
  • , Rui Yan*
  • , Chongyang Tao
  • , Mingrui Wu
  • *Corresponding author for this work
  • Peking University
  • Alibaba Group Holding Ltd.

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

Abstract

Considering that words with different characteristic in the text have different importance for classification, grouping them together separately can strengthen the semantic expression of each part. Thus we propose a new text representation scheme by clustering words according to their latent semantics and composing them together to get a set of cluster vectors, which are then concatenated as the final text representation. Evaluation on five classification benchmarks proves the effectiveness of our method. We further conduct visualization analysis showing statistical clustering results and verifying the validity of our motivation.

Original languageEnglish
Title of host publicationNatural Language Processing and Chinese Computing - 8th CCF International Conference, NLPCC 2019, Proceedings
EditorsJie Tang, Min-Yen Kan, Dongyan Zhao, Sujian Li, Hongying Zan
PublisherSpringer
Pages28-40
Number of pages13
ISBN (Print)9783030322359
DOIs
StatePublished - 2019
Externally publishedYes
Event8th CCF International Conference on Natural Language Processing and Chinese Computing, NLPCC 2019 - Dunhuang, China
Duration: 9 Oct 201914 Oct 2019

Publication series

NameLecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Volume11839 LNAI
ISSN (Print)0302-9743
ISSN (Electronic)1611-3349

Conference

Conference8th CCF International Conference on Natural Language Processing and Chinese Computing, NLPCC 2019
Country/TerritoryChina
CityDunhuang
Period9/10/1914/10/19

Keywords

  • Classification
  • Clustering
  • Semantic
  • Visualization

Fingerprint

Dive into the research topics of 'Classification over Clustering: Augmenting Text Representation with Clusters Helps!'. Together they form a unique fingerprint.

Cite this