Hidden Markov Models for Sentiment Analysis in Social Media

Perikos, Isidoros; Kardakis, Spyridon; Paraskevas, Michael; Hatzilygeroudis, Ioannis

doi:10.1109/bcd.2019.8885272

Cited by 7 publications

(2 citation statements)

References 20 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Perikos et al [23] treat sentiment analysis as a classification problem and propose a HMM model to utilize the sequential nature of text data. Separate HMM is trained for each sentiment label that uses and compares the Baum-Welch algorithm, Viterbi algorithm and simple Maximum Likelihood Estimation to estimate the HMM parameters.…”

Section: Related Workmentioning

confidence: 99%

MDFBEAM: Multi-Dimensional Fuzzy-Based Emotion Analysis Framework for Micro-Blog News Sub-Events

Vijayakumar,

2023

Preprint

View full text Add to dashboard Cite

Social media has replaced traditional journalism, such as television and newspapers, by a large proportion. Social media data presents an incredible opportunity for governments and organizations to understand the public's response to a news event, which is critical for government organizations to make informed decisions. News events are made up of a series of related sub-events. These sub-events determine the flow and dynamics of the entire news event. Hence, monitoring the public's response to each sub-event is crucial. This work proposes a framework to classify microblogs that discuss news sub-events into eight dimensions of emotion: Happy, Sad, Fear, Anger, Cynical, Neutral, Positive and Negative. This work introduces a cynical emotion that attempts to catch skepticism and sarcastic emotions, which are highly common in news-based events. Not limited to news events and microblogs, this framework can be utilized on any public response monitoring system without much changes. The proposed framework employs a bi-layered composite fuzzy mechanism. The emotions are represented as fuzzy sets. The fuzzy membership score of a microblog towards each fuzzy set is generated by integrating neural-network-based pre-trained word vector similarity, thesaurus-based similarity, and linguistic hedge information for each word. Fuzzy inference is then employed to identify the predominant emotion expressed in a microblog. This work is compared against state-of-the-art BERT models and demonstrates a remarkably high level of precision.

show abstract

Section: Related Workmentioning

confidence: 99%

MDFBEAM: Multi-Dimensional Fuzzy-Based Emotion Analysis Framework for Micro-Blog News Sub-Events

Vijayakumar,

2023

Preprint

View full text Add to dashboard Cite

show abstract

“…In particular, sentiment analysis is the computational study of people's opinions, emotions and attitudes towards entities, such as products, services and events [25]; it is highly sought after by businesses and service providers. However, automatic knowledge extraction about people's opinions and emotional states can be a very challenging task [26]. As a result, achieving performance gains, even on a small scale is of vital importance in natural language processing and sentiment analysis [27].…”

Section: Figurementioning

confidence: 99%

Examining Attention Mechanisms in Deep Learning Models for Sentiment Analysis

et al. 2021

Self Cite

View full text Add to dashboard Cite

Attention-based methods for deep neural networks constitute a technique that has attracted increased interest in recent years. Attention mechanisms can focus on important parts of a sequence and, as a result, enhance the performance of neural networks in a variety of tasks, including sentiment analysis, emotion recognition, machine translation and speech recognition. In this work, we study attention-based models built on recurrent neural networks (RNNs) and examine their performance in various contexts of sentiment analysis. Self-attention, global-attention and hierarchical-attention methods are examined under various deep neural models, training methods and hyperparameters. Even though attention mechanisms are a powerful recent concept in the field of deep learning, their exact effectiveness in sentiment analysis is yet to be thoroughly assessed. A comparative analysis is performed in a text sentiment classification task where baseline models are compared with and without the use of attention for every experiment. The experimental study additionally examines the proposed models’ ability in recognizing opinions and emotions in movie reviews. The results indicate that attention-based models lead to great improvements in the performance of deep neural models showcasing up to a 3.5% improvement in their accuracy.

show abstract