Date of Award

Spring 4-2022

Document Type

Dissertation

Degree Name

Doctor of Philosophy in Information Systems (PhDIS)

Department

Business and Information Systems

First Advisor

Jun Liu

Second Advisor

Cherie Noteboom

Third Advisor

Shengjie Xu

Abstract

The acceptance and popularity of social media platforms for the dispersion and proliferation of news articles have led to the spread of questionable and untrusted information (in part) due to the ease by which misleading content can be created and shared among the communities. While prior research has attempted to automatically classify news articles and tweets as credible and non-credible. This work complements such research by proposing an approach that utilizes the amalgamation of Natural Language Processing (NLP), and Deep Learning techniques such as Long Short-Term Memory (LSTM).

Moreover, in Information System’s paradigm, design science research methodology (DSRM) has become the major stream that focuses on building and evaluating an artifact to solve emerging problems. Hence, DSRM can accommodate deep learning-based models with the availability of adequate datasets. Two publicly available datasets that contain labeled news articles and tweets have been used to validate the proposed model’s effectiveness. This work presents two distinct experiments, and the results demonstrate that the proposed model works well for both long sequence news articles and short-sequence texts such as tweets. Finally, the findings suggest that the sentiments, tagging, linguistics, syntactic, and text embeddings are the features that have the potential to foster fake news detection through training the proposed model on various dimensionality to learn the contextual meaning of the news content.

Download

Share

COinS