Objective:
To develop and validate a methodology for detecting and analyzing disinformation campaigns by leveraging sentiment analysis and large language models.
Background and Significance:
- Problem Statement: Disinformation campaigns manipulate public opinion, affecting democratic processes and social harmony.
- Gap in Literature: Limited research combining sentiment analysis and language models to systematically identify and analyze disinformation campaigns.
- Significance: Enhancing the understanding and detection of disinformation, thus contributing to the mitigation of its adverse effects on society.
Research Questions:
- How effective are sentiment analysis and large language models in identifying disinformation campaigns?
- What are the characteristic features of the language used in disinformation campaigns, and how do they differ from legitimate information dissemination?
Expected Outcomes:
- Development of a reliable methodology/framework for disinformation detection.
- Insights into the linguistic features and sentiment patterns characteristic of disinformation.
- Enhanced understanding of the limitations and potentials of using sentiment analysis and large language models in this domain.