Text data management and analysis : a practical introduction to information retrieval and text mining / ChengXiang Zhai, Sean Massung.

Format:

Book

Author/Creator:

Zhai, ChengXiang, author.

Massung, Sean, author.

Series:

ACM books ; 2374-6777 #12.

ACM books, 2374-6777 ; #12

Language:

English

Subjects (All):

Data mining.

Natural language processing (Computer science).

Computational linguistics--Statistical methods.

Computational linguistics.

Physical Description:

1 online resource (xx, 510 pages) : illustrations.

Edition:

First edition.

Place of Publication:

[New York] : Association for Computing Machinery ; [San Rafael, California] : Morgan & Claypool, 2016.

System Details:

Mode of access: World Wide Web.

System requirements: Adobe Acrobat Reader.

Summary:

The growth of "big data" created unprecedented opportunities to leverage computational and statistical approaches to turn raw data into actionable knowledge that can support various application tasks. This is especially true for the optimization of decision making in virtually all application domains such as health and medicine, security and safety, learning and education, scientific discovery, and business intelligence. Just as a microscope enables us to see things in the "micro world" and a telescope allows us to see things far away, one can imagine a "big data scope" would enable us to extend our perception ability to "see" useful hidden information and knowledge buried in the data, which can help make predictions and improve the optimality of a chosen decision. This book covers general computational techniques for managing and analyzing large amounts of text data that can help users manage and make use of text data in all kinds of applications.

Contents:

Part I. Overview and background

1. Introduction

1.1 Functions of text information systems

1.2 Conceptual framework for text information systems

1.3 Organization of the book

1.4 How to use this book

Bibliographic notes and further reading

2. Background

2.1 Basics of probability and statistics

2.2 Information theory

2.3 Machine learning

Exercises

3. Text data understanding

3.1 History and state of the art in NLP

3.2 NLP and text information systems

3.3 Text representation

3.4 Statistical language models

4. MeTA: a unified toolkit for text data management and analysis

4.1 Design philosophy

4.2 Setting up MeTA

4.3 Architecture

4.4 Tokenization with MeTA

4.5 Related toolkits

Part II. Text data access

5. Overview of text data access

5.1 Access mode: pull vs. push

5.2 Multimode interactive access

5.3 Text retrieval

5.4 Text retrieval vs. database retrieval

5.5 Document selection vs. document ranking

6. Retrieval models

6.1 Overview

6.2 Common form of a retrieval function

6.3 Vector space retrieval models

6.4 Probabilistic retrieval models

7. Feedback

7.1 Feedback in the vector space model

7.2 Feedback in language models

8. Sarch engine implementation

8.1 Tokenizer

8.2 Indexer

8.3 Scorer

8.4 Feedback implementation

8.5 Compression

8.6 Caching

9. Search engine evaluation

9.1 Introduction

9.2 Evaluation of set retrieval

9.3 Evaluation of a ranked list

9.4 Evaluation with multi-level judgements

9.5 Practical issues in evaluation

10. Web search

10.1 Web crawling

10.2 Web indexing

10.3 Link analysis

10.4 Learning to rank

10.5 The future of web search

11. Recommender systems

11.1 Content-based recommendation

11.2 Collaborative filtering

11.3 Evaluation of recommender systems

Part III. Text data analysis

12. Overview of text data analysis

12.1 Motivation: applications of text data analysis

12.2 Text vs. non-text data: humans as subjective sensors

12.3 Landscape of text mining tasks

13. Word association mining

13.1 General idea of word association mining

13.2 Discovery of paradigmatic relations

13.3 Discovery of syntagmatic relations

13.4 Evaluation of word association mining

14. Text clustering

14.1 Overview of clustering techniques

14.2 Document clustering

14.3 Term clustering

14.4 Evaluation of text clustering

15. Text categorization

15.1 Introduction

15.2 Overview of text categorization methods

15.3 Text categorization problem

15.4 Features for text categorization

15.5 Classification algorithms

15.6 Evaluation of text categorization

16. Text summarization

16.1 Overview of text summarization techniques

16.2 Extractive text summarization

16.3 Abstractive text summarization

16.4 Evaluation of text summarization

16.5 Applications of text summarization

17. Topic analysis

17.1 Topics as terms

17.2 Topics as word distributions

17.3 Mining one topic from text

17.4 Probabilistic latent semantic analysis

17.5 Extension of PLSA and latent Dirichlet allocation

17.6 Evaluating topic analysis

17.7 Summary of topic models

18. Opinion mining and sentiment analysis

18.1 Sentiment classification

18.2 Ordinal regression

18.3 Latent aspect rating analysis

18.4 Evaluation of opinion mining and sentiment analysis

19. Joint analysis of text and structured data

19.1 Introduction

19.2 Contextual text mining

19.3 Contextual probabilistic latent semantic analysis

19.4 Topic analysis with social networks as context

19.5 Topic analysis with time series context

19.6 Summary

Part IV. Unified text data management analysis system

20. Toward a unified system for text management and analysis

20.1 Text analysis operators

20.2 System architecture

20.3 MeTA as a unified system

Appendix A. Bayesian statistics

Binomial estimation and the beta distribution

Pseudo counts, smoothing, and setting hyperparameters

Generalizing to a multinomial distribution

The Dirichlet distribution

Bayesian estimate of multinomial parameters

Conclusion

Appendix B. Expectation- maximization

A simple mixture Unigram language model

Maximum likelihood estimation

Incomplete vs. complete data

A lower bound of likelihood

The general procedure of EM

Appendix C. KL-divergence and Dirichlet prior smoothing

Using KL-divergence for retrieval

Using Dirichlet prior smoothing

Computing the query model p(w [theta]q)

References

Index

Authors' biographies.

Notes:

Includes bibliographical references and index.

Title from PDF title page (viewed on August 3, 2016).

Other Format:

Print version:

ISBN:

9781970001174

OCLC:

954593347

Access Restriction:

Restricted for use by site license.