Skip to main content

Table 7 Summary of the lexicon-based approach (dictionary based approach) used for opinion mining

From: Opinion mining for national security: techniques, domain applications, challenges and research opportunities

Reference

Objectives

Lexicon type

Materials

Output

[35]

To predict whether an online text expresses positive, negative or neutral sentiments without the need for supervision

Dictionary-based approach

The Cornell Movie Review dataset, The Obama-McCain Debate dataset, the SemEval-2015 dataset

Positive, negative or neutral

[82]

To improve the SWN performance by building a new lexical resource named SentiMI

SentiMI based classification, SentiWordNet

Movie review dataset

Positive, negative and objective

[85]

To present a web-based system known "TweeSent" that can estimate the polarity and emotion of tweets based on their input data from Twitter

NRC emotion lexicon

Tweets from Twitter

Joy, happiness, sadness, anger, trust, surprise, anticipation, fear, positive and negative

[87]

To classify movie reviews into positives, negatives and neutral polarity

The lexicon that has been published by Hu and Liu (2004)

Twitter data

Positives, negatives and neutral

[88]

To improve SentiWordNet performance and propose a complete sentiment analysis and classification framework according to SentiWordNet based vocabulary

SentiWordNet based classification

Large movie review dataset, Cornell movie review dataset, multi-domain sentiment datasets

Positive, negative or neutral

[89]

To investigate Alaskans’ perceptions and opinions on various energy sources and, in particular, clean energy sources

Subjectivity lexicon of English adjectives called ADJLex

Twitter data (Alaskans’ review) on energy consumption

Positive, neutral and negative

[80]

To recognise the emotional segmentation of a movie reviewer by extracting the sentiments from a given text and classifying them

Dictionary-based methods

Text movie review (IMDB)

Positive and negative

[90]

To automatically analyse student feedbacks (known as OMFeedback)

Vader Sentiment Intensity Analyser database of English sentiment words (Vader Lexicon)

Feedback

Positive, negative and neutral

[91]

To extract and classify sentiments and emotions from 141,208 headlines of global English news sources regarding the coronavirus disease (COVID-19)

NRC emotion lexicon, R package “sentiment”

English Headlines news sources

Positive, negative and neutral

[92]

To identify the public opinion of Filipino Twitter users concerning COVID-19 in three different timelines

Lexicon-based Approach

R package “sentiment dictionary”

Twitter textual (COVID-19)

Positive, negative, joy, sadness, fear, anticipation, anger, trust, surprise, disgust

[93]

To classify user reviews and use co-occurrence analysis to identify passengers’ concerns on different aspects of service in the aviation industry

Vader and Pattern lexicons

Reviews on SKYTRAX

Positive, negative and neutral

[94]

To study people’s reactions and emotions regarding Trump’s primary debates

R package “sentiment dictionary”

Tweets regarding the Trump Republican primary debate

Negative or positive

[95]

To illustrate and analyse the emotional sentiment of the campaign speeches of the two main candidates of 2016 US presidential elections

Word-Emotion Association Lexicon

Text files of American Presidency Project website

Negative and Positive

[96]

To estimate the reputation polarity of tweets

RepLab 2013 collection

Twitter data in English and Spanish

Positive, negative or neutral

[83]

To categorise YouTube comments based on content relevance

Wordnet

Keenformatics

Relevant, irrelevant, positive and negative

[97]

To correlate the distinct twitter comments of statesmen of distinct countries for having concrete knowledge on the application of drugs to patients attacked by COVID-19

TextBlob lexicon

Twitter

Positive and negative