Semantic Similarity vs. Sentiment Analysis – What Suits Better in Identifying Hate Speech on Twitter?

Zhixian Li ()
Additional contact information
Zhixian Li: Vanderbilt University

A chapter in LISS 2024, 2025, pp 1121-1134 from Springer

Abstract: Abstract Due to harmful content, hate speech’s prevalence on social media platforms like twitter poses significant challenges. This paper explores the efficacy of two natural language processing (nlp) methods, semantic similarity and part of speech (pos)-tagging sentiment analysis, in identifying hate speech. Semantic similarity compares the distance between word vectors in a sentence and in a corresponding corpus of a certain type of speech, while sentiment analysis tokenizes particular parts of speech (such as adjectives) in a sentence and identifies how often these emotions appear in the sentence. This study uses a dataset of tweets categorized by crowdflower workers based on whether they contain hate speech, offensive language, or neither, and conducts statistical analyses to compare the effectiveness of these two methods. Results indicate that negative sentiments are most effective in distinguishing disturbing tweets from normal ones, while semantic similarity analysis shows promise in conducting this distinction using a corpus made out of disturbing tweets. However, challenges remain in reliably classifying hate speech, suggesting avenues for future research to refine methods and definitions. Findings from this study contribute to the ongoing discourse on content moderation and online safety, visioning a more accurate system in tagging hate speech/offensive language on online platforms and thereby aiming to foster a more inclusive digital environment.

Keywords: hate speech; social media platforms; semantic similarity; sentiment analysis; natural language processing (search for similar items in EconPapers)
Date: 2025
References: Add references at CitEc
Citations:

There are no downloads for this item, see the EconPapers FAQ for hints about obtaining it.

Related works:
This item may be available elsewhere in EconPapers: Search for items with the same title.

Export reference: BibTeX RIS (EndNote, ProCite, RefMan) HTML/Text

Persistent link: https://EconPapers.repec.org/RePEc:spr:lnopch:978-981-96-9697-0_85

Ordering information: This item can be ordered from
http://www.springer.com/9789819696970

DOI: 10.1007/978-981-96-9697-0_85

Access Statistics for this chapter

More chapters in Lecture Notes in Operations Research from Springer
Bibliographic data for series maintained by Sonal Shukla () and Springer Nature Abstracting and Indexing ().