book

Computational Techniques for Text Summarization based on Cognitive Intelligence

by V. Priya, K. Umamaheswari

March 2023

Intermediate to advanced

228 pages

6h 9m

English

CRC Press

Read now

Unlock full access

1.1 Introduction1.2 Need for Text Summarization1.3 Approaches to Text Summarization1.3.1 Extractive Summarization1.3.2 Abstractive Summarization1.4 Text Modeling for Extractive Summarization1.4.1 Bag-of-Words Model1.4.2 Vector Space Model1.4.3 Topic Representation Schemes1.4.4 Real-Valued Model1.5 Preprocessing for Extractive Summarization1.6 Emerging Techniques for Summarization1.7 Scope of the BookReferencesSample CodeSample Screenshots
2.1 Scaling to Summarize Large Text2.2 Machine Learning Approaches2.2.1 Different Approaches for Modeling Text Summarization Problem2.2.2 Classification as Text Summarization2.2.2.1 Data Representation2.2.2.3 Classification Techniques2.2.3 Clustering as Text Summarization2.2.4 Deep Learning Approach for Text SummarizationReferencesSample Code

3.1 Introduction3.2 Sentiment Analysis: Overview3.2.1 Sentiment Extraction and Summarization3.2.1.1 Sentiment Extraction from Text3.2.1.2 Classification3.2.1.3 Score Computation3.2.1.4 Summary Generation3.2.2 Sentiment Summarization: An IllustrationSummarized Output3.2.3 Methodologies for Sentiment Summarization3.3 Implications of Sentiments in Text SummarizationCognition-Based Sentiment Analysis and Summarization3.4 SummaryPractical ExamplesExample 1Example 2Sample Code (Run Using GraphLab)Example 3ReferencesSample Code
4.1 IntroductionParallelizing Computational TasksParallelizing for Distributed Data4.2 Parallel Processing Approaches4.2.1 Parallel Algorithms for Text Summarization4.2.2 Parallel Bisection k-Means Method4.3 Parallel Data Processing Algorithms for Large-Scale Summarization4.3.1 Designing MapReduce Algorithm for Text Summarization4.3.2 Key Concepts in Mapper4.3.3 Key Concepts in Reducer4.3.4 Summary GenerationAn Illustrative Example for MapReduceGood Time: Movie Review4.4 Other MR-Based Methods4.5 Summary4.6 ExamplesK-Means Clustering Using MapReduceParallel LDA Example (Using Gensim Package)Sample Code: (Using Gensim Package)Example: Creating an Inverted IndexExample: Relational Algebra (Table JOIN)ReferencesSample Code
5.1 Introduction5.2 Optimization for Summarization5.2.1 Modeling Text Summarization as Optimization Problem5.2.2 Various Approaches for Optimization5.3 Formulation of Various Approaches5.3.1 Sentence Ranking Approach5.3.1.1 Stages and Illustration5.3.2 Evolutionary Approaches5.3.2.1 Stages5.3.2.2 Demonstration5.3.3 MapReduce-Based Approach5.3.3.1 In-Node Optimization Illustration5.3.4 Multi-objective-Based ApproachSummaryExercisesReferencesSample Code
6.1 Evaluation of Summaries6.1.1 CNN Dataset6.1.2 Daily Mail Dataset6.1.3 Description6.2 Methodologies6.2.1 Intrinsic Methods6.2.2 Extrinsic Methods6.3 Intrinsic Methods6.3.1 Text Quality Measures6.3.1.1 Grammaticality6.3.1.2 Non-redundancy6.3.1.3 Reverential Clarity6.3.1.4 Structure and Coherence6.3.2 Co-selection-Based Methods6.3.2.1 Precision, Recall, and F-score6.3.2.2 Relative Utility6.3.3 Content-Based Methods6.3.3.1 Content-Based Measures6.3.3.2 Cosine Similarity6.3.3.3 Unit Overlap6.3.3.4 Longest Common Subsequence6.3.3.5 N-Gram Co-occurrence Statistics: ROUGE6.3.3.6 Pyramids6.3.3.7 LSA-Based Measure6.3.3.8 Main Topic Similarity6.3.3.9 Term Significance Similarity6.4 Extrinsic Methods6.4.1 Document Categorization6.4.1.1 Information Retrieval6.4.1.2 Question Answering6.4.2 Summary6.4.3 ExamplesBibliography
7.1 Possible Directions in Modeling Text Summarization7.2 Scope of Summarization Systems in Different Applications7.3 Healthcare DomainFuture Directions for Medical Document Summarization7.4 Social MediaChallenges in Social Media Text SummarizationDomain Knowledge and Transfer LearningOnline LearningInformation CredibilityApplications of Deep LearningImplicit and Explicit Information for Actionable Insights7.5 Research Directions for Text Summarization7.6 Further Scope of Research on Large-Scale SummarizationConclusionReferences

Content preview from Computational Techniques for Text Summarization based on Cognitive Intelligence

3 Sentiment Analysis Approach to Text Summarization

DOI: 10.1201/9781003371199-3

3.1 Introduction

Sentiment analysis has been widely used for making key decisions in business. Sentiments for summarization have gained importance since they capture opinions about the entities which are of great interest. Sentiment summarization can be viewed as a subfield of text summarization. The goal of a sentiment summarization system is to generate a summary S, representative of the average opinion about the important aspects relevant to an entity. This can be managed to solve using either a machine learning approach or a natural language processing approach, since it learns sentiments from text using natural language features. This performs fine-grained ...