A Joint Model for Multi-Granularity Topics and Sentiments with Fusing Word Order Information

doi:10.7652/xjtuxb201411018

您当前的位置：

首页 >

文章列表页 >

A Joint Model for Multi-Granularity Topics and Sentiments with Fusing Word Order Information

更新时间：2025-07-09

- A Joint Model for Multi-Granularity Topics and Sentiments with Fusing Word Order Information
- Vol. 48, Issue 11, Pages: 103-108(2014)
- 作者机构：
  
  西安建筑科技大学管理学院,西安,710055
- 作者简介：
- 基金信息：
- DOI：10.7652/xjtuxb201411018
  CLC： TP391
- Online First：10 November 2014，
  
  Published：2014
- 稿件说明：
移动端阅览
A Joint Model for Multi-Granularity Topics and Sentiments with Fusing Word Order Information[J]. 2014, 48(11): 103-108.
DOI：

A Joint Model for Multi-Granularity Topics and Sentiments with Fusing Word Order Information[J]. 2014, 48(11): 103-108. DOI： 10.7652/xjtuxb201411018.

摘要

针对基本话题模型只能抽取粗粒度上下文信息的问题

通过对潜在狄里克雷分配(LDA)模型进行扩展

建立了一种利用词序信息的多粒度话题情感联合模型(MTSU-Col)。MTSU-Col模型客观表达了词汇、全局/局部话题、情感标签和词序信息之间的关联关系

使模型中话题和情感的建模更加符合文本的语义表达

有效解决了现有话题、情感分析方法存在的领域依赖问题

从而实现了文本多粒度话题信息和情感倾向信息的同步非监督获取。实验表明:利用MTSU-Col模型对文本进行情感倾向性分类

可使综合评价指标F

值达到84%

整体性能与监督分类方法支持向量机(SVM)类似

均优于未采用词序信息的分析方法。由于挖掘话题集合具有层次化、语义相关的特点

因此MTSU-Col模型对观点挖掘是可行、有效的。

Abstract

A joint model for multi-granularity topics and sentiments(MTSU-Col model)based on an extension to LDA model by incorporating collocation is proposed to solve the problem that the basic topic model captures only coarse-granularity contextual information. The MTSU-Col model objectively expresses the correlative relationship among words

global\local topics

sentiment labels and collocation

allows us to infer topics and sentiment information

and provides a closer match to real semantic representation contained in texts. The MTSU-Col model synchronously realizes an unsupervised mining of multi-granularity topics and sentiment information

and effectively solves the domain dependent problem in existing methods. Experimental results show that the proposed

model achieves F

of 84% for sentiment classification

and its performance is comparable to the performance of SVM methods. Since the mining collection of topics is hierarchy and semantic related

it is feasible and effective to use the proposed model for opinion mining.

关键词

Keywords

references

LIU B, ZHANG L. A survey on opinion mining and sentiment analysis[M]. Berlin, Germany: Springer, 2012: 415-463.

MEI Q, ZHAI C. Discovering evolutionary theme patterns from text-an exploration of temporal text mining[C]∥Proceedings of the 12th ACM SIGKDD International Conference on Knowledge Discovery and Data Mining. New York, USA: ACM, 2005: 198-207.

PANG B, LEE L. Opinion mining and sentiment analysis[J]. Foundations and Trends in Information Retrieval, 2008, 2(1/2): 1-135.

TANG H, TAN S, CHENG X. A survey on sentiment detection of reviews[J]. Expert Systems with Applications, 2009, 36(7): 10760-10773.

CARENINI G, NG R, PAULS A. Multi-document summarization of evaluative text[C]∥Proceedings of the 11th European Chapter of the Association for Computational Linguistics. Trento, Italy: ACL, 2006: 3-7.

HU M, LIU B. Mining and summarizing customer reviews[C]∥The 10th ACM SIGKDD Conference on Knowledge Discovery and Data Mining 2004. New York, USA: ACM, 2004: 168-177.

BLEI D M, NG A Y, JORDAN M I. Latent Dirichlet allocation[J]. Journal of Machine Learning Research, 2003, 3(4/5): 993-1022.

徐戈, 王厚峰. 自然语言处理中主题模型的发展[J]. 计算机学报, 2011, 34(8): 1423-1436.

XU Ge, WANG Houfeng. The development of topics models in natural language processing[J]. Chinese Journal of Computers, 2011, 34(8): 1423-1436.

冯时, 景珊, 杨卓, 等. 基于LDA模型的中文微博话题意见领袖挖掘[J]. 东北大学学报, 2013, 34(4): 490-494.

FENG Shi, JING Shan, YANG Zhuo, et al. Detecting topical opinion leaders based on LDA model in Chinese microblogs[J]. Journal of Northeastern University, 2013, 34(4): 490-494.

LIN C, HE Y. Joint sentiment/topic model for sentiment analysis[C]∥The 18th ACM Conference on Information and Knowledge Management. New York, USA: ACM, 2009: 375-384.[11] TITOV I, MCDONALD R. Modeling online reviews with multi-grain topic models[C]∥The 17th International World Wide Web Conference 2008. New York, USA: ACM, 2008: 111-120.

JO Y, OH A. Aspect and sentiment unification mode for online review analysis[C]∥The 4th ACM International Conference on Web Search and Data Mining. New York, USA: ACM, 2011: 815-824.

GRIFFITHS T, STEYVERS M, TENENBAUM J B. Topics in semantic representation[J]. Psychological Review, 2007, 114(2): 211-244.

GRIFFITHS T, STEYVERS M. Finding scientific topics[C]∥Proceedings of the National Academy of Sciences. New York, USA: United States National Academy of Sciences, 2004: 5228-5235.

Views

下载量

CSCD

Alert me when the article has been cited

提交

Tools

Publicity Resources

Electrothermal Coupling Research on the Evolutionary Law of Anti-Corona Performance for High-Altitude Air-Cooled Generators

Effects of Expansion Angle Crater Film Hole on the Film Cooling Characteristics of a Double-Wall Structure

Integrated and Joint Design Optimization of Centrifugal Compressor Impeller and Casing Treatment at Multiple Working Points

Knowledge Graph Representation Learning Driven by Semantic Discrimination and Subgraph Integration

Sensitivity Analysis of Foil Geometrical Parameters for the Third-Generation CO2 Gas Foil Bearings Based on Design of Experiments Method

Related Author

No data

Related Institution

No data

⁰