How to conduct sentiment analysis / Mike Thelwall.

Format:

Book

Author/Creator:

Thelwall, Mike, author.

Series:

SAGE research methods: doing research online

Language:

English

Subjects (All):

Sentiment analysis.

Physical Description:

1 online resource.

Place of Publication:

London : SAGE Publications, Ltd., 2022.

Summary:

This How-to Guide describes what sentiment analysis is, when it might be used, how sentiment analysis software works, and how sentiment analysis can be applied in academic research projects. Essentially, the task of sentiment analysis software is to guess the emotions or opinions expressed in (usually) texts. It does this by using machine learning or linguistic rules and, in some contexts, can be as accurate as humans. This software is much faster than humans, allowing sentiment detection in huge sets of texts. Nevertheless, automatic sentiment analysis needs human face-validity checks to guard against algorithmic bias. Sentiment analysis can be used for tasks such as detecting topics that elicit strong opinions or analysing the evolution of emotional reactions over time for an issue. It works best on huge collections of text (e.g., hundreds of thousands of tweets) on topics where people express sentiment directly with clear statements rather than indirectly, such as through implication, sarcasm, or irony. This guide finishes with brief information about how the free software SentiStrength can be used for sentiment analysis and how the free software Mozdeh can be used to detect sentiment-related patterns in large sets of text.

Notes:

Description based on publisher supplied metadata and other sources.

ISBN:

1-5296-0740-X

9781529607406

OCLC:

1312917658