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

^*此作品的通讯作者

计算机学院

Beijing Institute of Technology
Sohu, Inc.

科研成果: 书/报告/会议事项章节 › 会议稿件 › 同行评审

3 引用（Scopus）

摘要

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.

源语言	英语
主期刊名	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
编辑	Maosong Sun, Zhiyuan Liu, Yang Liu, Min Zhang
出版商	Springer Verlag
页	215-226
页数	12
ISBN（印刷版）	9783319258157
DOI	https://doi.org/10.1007/978-3-319-25816-4_18
出版状态	已出版 - 2015
活动	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, 中国期限: 13 11月 2015 → 14 11月 2015

出版系列

姓名	Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
卷	9427
ISSN（印刷版）	0302-9743
ISSN（电子版）	1611-3349

会议

会议	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
国家/地区	中国
市	Guangzhou
时期	13/11/15 → 14/11/15

访问文件

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

其它文件与链接

链接到 Scopus 的出版物

引用此

Mao, X. L., Xiao, Y., Zhou, Q., Wang, J., & Huang, H. (2015). EHLLDA: A supervised hierarchical topic model. 在 M. Sun, Z. Liu, Y. Liu, & M. Zhang (编辑), 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 (页码 215-226). (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); 卷 9427). Springer Verlag. https://doi.org/10.1007/978-3-319-25816-4_18

Mao, Xian Ling ; Xiao, Yixuan ; Zhou, Qiang 等. / 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. 编辑 / Maosong Sun ; Zhiyuan Liu ; Yang Liu ; Min Zhang. Springer Verlag, 2015. 页码 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. 在 M Sun, Z Liu, Y Liu & M Zhang (编辑), 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), 卷 9427, Springer Verlag, 页码 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, 中国, 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 等.
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. 编辑 / Maosong Sun; Zhiyuan Liu; Yang Liu; Min Zhang. Springer Verlag, 2015. 页码 215-226 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); 卷 9427).

科研成果: 书/报告/会议事项章节 › 会议稿件 › 同行评审

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 - https://www.scopus.com/pages/publications/84952646774

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. 在 Sun M, Liu Z, Liu Y, Zhang M, 编辑, 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. 页码 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

摘要

出版系列

会议

访问文件

其它文件与链接

指纹

引用此