Fake Financial Content Identification using Machine Learning
Loading...
Date
2024
Journal Title
Journal ISSN
Volume Title
Publisher
UMT, Lahore
Abstract
This research explores the realm of finance-related tweets on Twitter, aiming to recognize the origin of content by distinguishing between tweets created by humans, ChatGPT, and Quillbot. Leveraging preprocessing techniques in Natural Language Processing (NLP) and Machine Learning (ML), the research undertakes the compilation of a diverse dataset comprising finance-related tweets sourced from the Twitter platform. Here, the preprocessing steps including text cleaning process, converting the text into tokens, and finally feature selection which are equally important for fine-tuning and preparing the data for the classification process are restricted. To fulfill the study's objectives, two distinct machine learning models are employed: One was trained only by the human-generated Tweets whereas the other was trained on the Tweets generated by both the ChatGPT and Quillbot. During this assessment, these aforementioned models will undergo the evaluation process centered on the finance tweets set and examine their classification techniques. Thus, the primary focus of this research to outline the values of NLP and ML approaches as a means of understanding and distinguishing between the texts penned by people and the texts created by AI, where the landscape of the finance discussions on social media. From this study, the following outcomes are expected to accomplish the research aim and objectives, as well as make significant findings regarding the applicability and legitimacy of using complicated preprocessing techniques and machine learning algorithms to analyze the origin of financial-relevant messages published on Twitter and provide insights into developing human and artificial intelligence interactions in computing