Comparative analysis of lexicon-based sentiment analysis methods

BALDWIN, James, BRUNSDON, Teresa, GAUDOIN, Jotham and HIRSCH, Laurence (2023). Comparative analysis of lexicon-based sentiment analysis methods. [Pre-print] (Unpublished)

Preprints have not been peer-reviewed. They should not be relied on to guide clinical practice or health related behaviour and should not be regarded as conclusive or be reported in news media as established information.
Baldwin-ComparativeAnalysisOf(Pre-print).pdf - Pre-print
All rights reserved.

Download (677kB) | Preview
Official URL:


Sentiment Analysis studies the opinions, sentiments and emotions expressed at sentence or document level. Machine learning and lexicon-based approaches have been successfully used to achieve this. This paper will focus on the lexicon-based approach. In contrast to most existing research, we compare the effectiveness of multiple dictionaries across a series of datasets related to public order events. The comparison will look to understand the possible benefits and limitations of sentiment analysis methods, which will bench marked against each other in their evaluation of results. The evaluation will be based four labelled datasets, covering messages on posts related to public order events. The results will highlight the extent of how well each of these methods perform across the datasets comparing sentence level analysis with range of sentiment analysis techniques.

Item Type: Pre-print
Identification Number:
SWORD Depositor: Symplectic Elements
Depositing User: Symplectic Elements
Date Deposited: 25 Aug 2023 09:59
Last Modified: 11 Sep 2023 08:59

Actions (login required)

View Item View Item


Downloads per month over past year

View more statistics