Labelling topics in Weibo using word embedding and graph-based method

Zhipeng Jin, Qiudan Li, Can Wang, Dajun Zeng, Lei Wang

Research output: Chapter in Book/Report/Conference proceedingConference contribution

1 Citation (Scopus)

Abstract

Nowadays, in China, Weibo is becoming an increasingly popular way for people to know what is happening in the world. Labelling topics is of much importance for better understanding the semantics of topics. Existing works mainly focus on deriving candidate labels by exploring the use of external knowledge, which may be more appropriate for well formatted and static documents. Recently, it has been a new trend to generate labels for sparse and dynamic microblogging environment using summarization method. The challenges of labelling topics are how to obtain coherent candidate labels and how to rank the labels. In this paper, based on the latest research work in deep learning, we propose a novel and unified model for labelling topics in Weibo, which firstly adopts word embedding and clustering method to learn dense semantic representation of topic words and mine the coherent candidate topic labels, then, generates interpretable labels using a graph-based model. Experimental results show that topics labels discovered by our model not only have high topic coherence, but also are meaningful and interpretable.

Original languageEnglish (US)
Title of host publicationProceedings - 2016 International Conference on Information Systems Engineering, ICISE 2016
PublisherInstitute of Electrical and Electronics Engineers Inc.
Pages34-37
Number of pages4
ISBN (Electronic)9781509022885
DOIs
StatePublished - Jun 7 2016
Event2016 International Conference on Information Systems Engineering, ICISE 2016 - Los Angeles, United States
Duration: Apr 20 2016Apr 22 2016

Other

Other2016 International Conference on Information Systems Engineering, ICISE 2016
CountryUnited States
CityLos Angeles
Period4/20/164/22/16

Fingerprint

Labeling
Labels
Semantics

Keywords

  • Deep learning
  • Graph
  • Labelling topics
  • Microblogs
  • Weibo

ASJC Scopus subject areas

  • Computer Science (miscellaneous)
  • Control and Systems Engineering

Cite this

Jin, Z., Li, Q., Wang, C., Zeng, D., & Wang, L. (2016). Labelling topics in Weibo using word embedding and graph-based method. In Proceedings - 2016 International Conference on Information Systems Engineering, ICISE 2016 (pp. 34-37). [7486210] Institute of Electrical and Electronics Engineers Inc.. https://doi.org/10.1109/ICISE.2016.15

Labelling topics in Weibo using word embedding and graph-based method. / Jin, Zhipeng; Li, Qiudan; Wang, Can; Zeng, Dajun; Wang, Lei.

Proceedings - 2016 International Conference on Information Systems Engineering, ICISE 2016. Institute of Electrical and Electronics Engineers Inc., 2016. p. 34-37 7486210.

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Jin, Z, Li, Q, Wang, C, Zeng, D & Wang, L 2016, Labelling topics in Weibo using word embedding and graph-based method. in Proceedings - 2016 International Conference on Information Systems Engineering, ICISE 2016., 7486210, Institute of Electrical and Electronics Engineers Inc., pp. 34-37, 2016 International Conference on Information Systems Engineering, ICISE 2016, Los Angeles, United States, 4/20/16. https://doi.org/10.1109/ICISE.2016.15
Jin Z, Li Q, Wang C, Zeng D, Wang L. Labelling topics in Weibo using word embedding and graph-based method. In Proceedings - 2016 International Conference on Information Systems Engineering, ICISE 2016. Institute of Electrical and Electronics Engineers Inc. 2016. p. 34-37. 7486210 https://doi.org/10.1109/ICISE.2016.15
Jin, Zhipeng ; Li, Qiudan ; Wang, Can ; Zeng, Dajun ; Wang, Lei. / Labelling topics in Weibo using word embedding and graph-based method. Proceedings - 2016 International Conference on Information Systems Engineering, ICISE 2016. Institute of Electrical and Electronics Engineers Inc., 2016. pp. 34-37
@inproceedings{da19451123874eef945c8c7f3538c389,
title = "Labelling topics in Weibo using word embedding and graph-based method",
abstract = "Nowadays, in China, Weibo is becoming an increasingly popular way for people to know what is happening in the world. Labelling topics is of much importance for better understanding the semantics of topics. Existing works mainly focus on deriving candidate labels by exploring the use of external knowledge, which may be more appropriate for well formatted and static documents. Recently, it has been a new trend to generate labels for sparse and dynamic microblogging environment using summarization method. The challenges of labelling topics are how to obtain coherent candidate labels and how to rank the labels. In this paper, based on the latest research work in deep learning, we propose a novel and unified model for labelling topics in Weibo, which firstly adopts word embedding and clustering method to learn dense semantic representation of topic words and mine the coherent candidate topic labels, then, generates interpretable labels using a graph-based model. Experimental results show that topics labels discovered by our model not only have high topic coherence, but also are meaningful and interpretable.",
keywords = "Deep learning, Graph, Labelling topics, Microblogs, Weibo",
author = "Zhipeng Jin and Qiudan Li and Can Wang and Dajun Zeng and Lei Wang",
year = "2016",
month = "6",
day = "7",
doi = "10.1109/ICISE.2016.15",
language = "English (US)",
pages = "34--37",
booktitle = "Proceedings - 2016 International Conference on Information Systems Engineering, ICISE 2016",
publisher = "Institute of Electrical and Electronics Engineers Inc.",
address = "United States",

}

TY - GEN

T1 - Labelling topics in Weibo using word embedding and graph-based method

AU - Jin, Zhipeng

AU - Li, Qiudan

AU - Wang, Can

AU - Zeng, Dajun

AU - Wang, Lei

PY - 2016/6/7

Y1 - 2016/6/7

N2 - Nowadays, in China, Weibo is becoming an increasingly popular way for people to know what is happening in the world. Labelling topics is of much importance for better understanding the semantics of topics. Existing works mainly focus on deriving candidate labels by exploring the use of external knowledge, which may be more appropriate for well formatted and static documents. Recently, it has been a new trend to generate labels for sparse and dynamic microblogging environment using summarization method. The challenges of labelling topics are how to obtain coherent candidate labels and how to rank the labels. In this paper, based on the latest research work in deep learning, we propose a novel and unified model for labelling topics in Weibo, which firstly adopts word embedding and clustering method to learn dense semantic representation of topic words and mine the coherent candidate topic labels, then, generates interpretable labels using a graph-based model. Experimental results show that topics labels discovered by our model not only have high topic coherence, but also are meaningful and interpretable.

AB - Nowadays, in China, Weibo is becoming an increasingly popular way for people to know what is happening in the world. Labelling topics is of much importance for better understanding the semantics of topics. Existing works mainly focus on deriving candidate labels by exploring the use of external knowledge, which may be more appropriate for well formatted and static documents. Recently, it has been a new trend to generate labels for sparse and dynamic microblogging environment using summarization method. The challenges of labelling topics are how to obtain coherent candidate labels and how to rank the labels. In this paper, based on the latest research work in deep learning, we propose a novel and unified model for labelling topics in Weibo, which firstly adopts word embedding and clustering method to learn dense semantic representation of topic words and mine the coherent candidate topic labels, then, generates interpretable labels using a graph-based model. Experimental results show that topics labels discovered by our model not only have high topic coherence, but also are meaningful and interpretable.

KW - Deep learning

KW - Graph

KW - Labelling topics

KW - Microblogs

KW - Weibo

UR - http://www.scopus.com/inward/record.url?scp=84978542832&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84978542832&partnerID=8YFLogxK

U2 - 10.1109/ICISE.2016.15

DO - 10.1109/ICISE.2016.15

M3 - Conference contribution

AN - SCOPUS:84978542832

SP - 34

EP - 37

BT - Proceedings - 2016 International Conference on Information Systems Engineering, ICISE 2016

PB - Institute of Electrical and Electronics Engineers Inc.

ER -