Contents I. Introduction
TRADITIONAL document analysis methods often rely on manual review or simplistic keyword-based searches, leading to significant inefficiencies and limitations. As the volume and diversity of documents continue to expand, the need for innovative approaches that can streamline analysis while preserving accuracy and comprehensiveness arises. In this study, we explore different methods for assisting the analysis of selected EdTech providers’ Data Privacy Policy (DPP) documents. On the task at hand, we aim to evaluate the efficacy, consistency, benefits and limitations of various LLMs in assessing DPP documents. The importance for such an assessment is founded on the need for an automated, scalable and reliable way to systematically analyze large bodies of text semantically. We also aim to examine the most optimal way of using LLMs regarding the optimizing factor - whether it is the price, the duration or the format of the answers provided that plays a key role in a technical chore. All of these LLM models are trained on extensive datasets and exhibit remarkable proficiency in generating human-like responses and cognitive reasoning across diverse tasks. LLMs are built to handle various tasks, such as text generation, translation, content summary, chatbot conversations, and more [1].
