EHLLDA: A supervised hierarchical topic model

Xian Ling Mao; Yixuan Xiao; Qiang Zhou; Jun Wang; Heyan Huang

doi:10.1007/978-3-319-25816-4_18

EHLLDA: A supervised hierarchical topic model

Xian Ling Mao^*, Yixuan Xiao, Qiang Zhou, Jun Wang, Heyan Huang

^*Corresponding author for this work

School of Computer Science and Technology

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution › peer-review

2 Citations (Scopus)

Abstract

In this paper, we consider the problem of modeling hierarchical labeled data – such as Web pages and their placement in hierarchical directories. The state-of-the-art model, hierarchical Labeled LDA (hLLDA), assumes that each child of a non-leaf label has equal importance, and that a document in the corpus cannot locate in a non-leaf node. However, in most cases, these assumptions do not meet the actual situation. Thus, in this paper, we introduce a supervised hierarchical topic models: Extended Hierarchical Labeled Latent Dirichlet Allocation (EHLLDA), which aim to relax the assumptions of hLLDA by incorporating prior information of labels into hLLDA. The experimental results show that the perplexity performance of EHLLDA is always better than that of LLDA and hLLDA on all four datasets; and our proposed model is also superior to hLLDA in terms of p@n.

Original language	English
Title of host publication	Chinese Computational Linguistics and Natural Language Processing Based on Naturally Annotated Big Data - 14th China National Conference, CCL 2015 and 3rd International Symposium, NLP-NABD 2015, Proceedings
Editors	Maosong Sun, Zhiyuan Liu, Yang Liu, Min Zhang
Publisher	Springer Verlag
Pages	215-226
Number of pages	12
ISBN (Print)	9783319258157
DOIs	https://doi.org/10.1007/978-3-319-25816-4_18
Publication status	Published - 2015
Event	14th China National Conference on Chinese Computational Linguistics, CCL 2015 and 3rd International Symposium on Natural Language Processing Based on Naturally Annotated Big Data, NLP-NABD 2015 - Guangzhou, China Duration: 13 Nov 2015 → 14 Nov 2015

Publication series

Name	Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Volume	9427
ISSN (Print)	0302-9743
ISSN (Electronic)	1611-3349

Conference

Conference	14th China National Conference on Chinese Computational Linguistics, CCL 2015 and 3rd International Symposium on Natural Language Processing Based on Naturally Annotated Big Data, NLP-NABD 2015
Country/Territory	China
City	Guangzhou
Period	13/11/15 → 14/11/15

Keywords

Hierarchical topic modeling
Supervised learning
Topic modeling

Access to Document

10.1007/978-3-319-25816-4_18

Cite this

Mao, X. L., Xiao, Y., Zhou, Q., Wang, J., & Huang, H. (2015). EHLLDA: A supervised hierarchical topic model. In M. Sun, Z. Liu, Y. Liu, & M. Zhang (Eds.), Chinese Computational Linguistics and Natural Language Processing Based on Naturally Annotated Big Data - 14th China National Conference, CCL 2015 and 3rd International Symposium, NLP-NABD 2015, Proceedings (pp. 215-226). (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 9427). Springer Verlag. https://doi.org/10.1007/978-3-319-25816-4_18

Mao, Xian Ling ; Xiao, Yixuan ; Zhou, Qiang et al. / EHLLDA : A supervised hierarchical topic model. Chinese Computational Linguistics and Natural Language Processing Based on Naturally Annotated Big Data - 14th China National Conference, CCL 2015 and 3rd International Symposium, NLP-NABD 2015, Proceedings. editor / Maosong Sun ; Zhiyuan Liu ; Yang Liu ; Min Zhang. Springer Verlag, 2015. pp. 215-226 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)).

@inproceedings{75fbad6145064b2090075cbe712aa3b5,

title = "EHLLDA: A supervised hierarchical topic model",

abstract = "In this paper, we consider the problem of modeling hierarchical labeled data – such as Web pages and their placement in hierarchical directories. The state-of-the-art model, hierarchical Labeled LDA (hLLDA), assumes that each child of a non-leaf label has equal importance, and that a document in the corpus cannot locate in a non-leaf node. However, in most cases, these assumptions do not meet the actual situation. Thus, in this paper, we introduce a supervised hierarchical topic models: Extended Hierarchical Labeled Latent Dirichlet Allocation (EHLLDA), which aim to relax the assumptions of hLLDA by incorporating prior information of labels into hLLDA. The experimental results show that the perplexity performance of EHLLDA is always better than that of LLDA and hLLDA on all four datasets; and our proposed model is also superior to hLLDA in terms of p@n.",

keywords = "Hierarchical topic modeling, Supervised learning, Topic modeling",

author = "Mao, {Xian Ling} and Yixuan Xiao and Qiang Zhou and Jun Wang and Heyan Huang",

note = "Publisher Copyright: {\textcopyright} Springer International Publishing Switzerland 2015.; 14th China National Conference on Chinese Computational Linguistics, CCL 2015 and 3rd International Symposium on Natural Language Processing Based on Naturally Annotated Big Data, NLP-NABD 2015 ; Conference date: 13-11-2015 Through 14-11-2015",

year = "2015",

doi = "10.1007/978-3-319-25816-4_18",

language = "English",

isbn = "9783319258157",

series = "Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)",

publisher = "Springer Verlag",

pages = "215--226",

editor = "Maosong Sun and Zhiyuan Liu and Yang Liu and Min Zhang",

booktitle = "Chinese Computational Linguistics and Natural Language Processing Based on Naturally Annotated Big Data - 14th China National Conference, CCL 2015 and 3rd International Symposium, NLP-NABD 2015, Proceedings",

address = "Germany",

}

Mao, XL, Xiao, Y, Zhou, Q, Wang, J & Huang, H 2015, EHLLDA: A supervised hierarchical topic model. in M Sun, Z Liu, Y Liu & M Zhang (eds), Chinese Computational Linguistics and Natural Language Processing Based on Naturally Annotated Big Data - 14th China National Conference, CCL 2015 and 3rd International Symposium, NLP-NABD 2015, Proceedings. Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), vol. 9427, Springer Verlag, pp. 215-226, 14th China National Conference on Chinese Computational Linguistics, CCL 2015 and 3rd International Symposium on Natural Language Processing Based on Naturally Annotated Big Data, NLP-NABD 2015, Guangzhou, China, 13/11/15. https://doi.org/10.1007/978-3-319-25816-4_18

EHLLDA: A supervised hierarchical topic model. / Mao, Xian Ling; Xiao, Yixuan; Zhou, Qiang et al.
Chinese Computational Linguistics and Natural Language Processing Based on Naturally Annotated Big Data - 14th China National Conference, CCL 2015 and 3rd International Symposium, NLP-NABD 2015, Proceedings. ed. / Maosong Sun; Zhiyuan Liu; Yang Liu; Min Zhang. Springer Verlag, 2015. p. 215-226 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 9427).

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution › peer-review

TY - GEN

T1 - EHLLDA

T2 - 14th China National Conference on Chinese Computational Linguistics, CCL 2015 and 3rd International Symposium on Natural Language Processing Based on Naturally Annotated Big Data, NLP-NABD 2015

AU - Mao, Xian Ling

AU - Xiao, Yixuan

AU - Zhou, Qiang

AU - Wang, Jun

AU - Huang, Heyan

N1 - Publisher Copyright: © Springer International Publishing Switzerland 2015.

PY - 2015

Y1 - 2015

N2 - In this paper, we consider the problem of modeling hierarchical labeled data – such as Web pages and their placement in hierarchical directories. The state-of-the-art model, hierarchical Labeled LDA (hLLDA), assumes that each child of a non-leaf label has equal importance, and that a document in the corpus cannot locate in a non-leaf node. However, in most cases, these assumptions do not meet the actual situation. Thus, in this paper, we introduce a supervised hierarchical topic models: Extended Hierarchical Labeled Latent Dirichlet Allocation (EHLLDA), which aim to relax the assumptions of hLLDA by incorporating prior information of labels into hLLDA. The experimental results show that the perplexity performance of EHLLDA is always better than that of LLDA and hLLDA on all four datasets; and our proposed model is also superior to hLLDA in terms of p@n.

AB - In this paper, we consider the problem of modeling hierarchical labeled data – such as Web pages and their placement in hierarchical directories. The state-of-the-art model, hierarchical Labeled LDA (hLLDA), assumes that each child of a non-leaf label has equal importance, and that a document in the corpus cannot locate in a non-leaf node. However, in most cases, these assumptions do not meet the actual situation. Thus, in this paper, we introduce a supervised hierarchical topic models: Extended Hierarchical Labeled Latent Dirichlet Allocation (EHLLDA), which aim to relax the assumptions of hLLDA by incorporating prior information of labels into hLLDA. The experimental results show that the perplexity performance of EHLLDA is always better than that of LLDA and hLLDA on all four datasets; and our proposed model is also superior to hLLDA in terms of p@n.

KW - Hierarchical topic modeling

KW - Supervised learning

KW - Topic modeling

UR - http://www.scopus.com/inward/record.url?scp=84952646774&partnerID=8YFLogxK

U2 - 10.1007/978-3-319-25816-4_18

DO - 10.1007/978-3-319-25816-4_18

M3 - Conference contribution

AN - SCOPUS:84952646774

SN - 9783319258157

T3 - Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)

SP - 215

EP - 226

BT - Chinese Computational Linguistics and Natural Language Processing Based on Naturally Annotated Big Data - 14th China National Conference, CCL 2015 and 3rd International Symposium, NLP-NABD 2015, Proceedings

A2 - Sun, Maosong

A2 - Liu, Zhiyuan

A2 - Liu, Yang

A2 - Zhang, Min

PB - Springer Verlag

Y2 - 13 November 2015 through 14 November 2015

ER -

Mao XL, Xiao Y, Zhou Q, Wang J, Huang H. EHLLDA: A supervised hierarchical topic model. In Sun M, Liu Z, Liu Y, Zhang M, editors, Chinese Computational Linguistics and Natural Language Processing Based on Naturally Annotated Big Data - 14th China National Conference, CCL 2015 and 3rd International Symposium, NLP-NABD 2015, Proceedings. Springer Verlag. 2015. p. 215-226. (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)). doi: 10.1007/978-3-319-25816-4_18

EHLLDA: A supervised hierarchical topic model

Abstract

Publication series

Conference

Keywords

Access to Document

Other files and links

Fingerprint

Cite this