Informational Query Detection on Social Media Posts in Bengali Language Using Machine Learning And Transfer Learning Techniques

Informational Queries are those queries or questions where a user's goal is to get an answer or fact to a specific question or to find a relevant piece of information. In recent years, social media has become an important platform for people to seek knowledge, data, and information from others. However, the vast amount of information on social media makes it difficult to identify and prioritize posts that are specifically informational queries. This paper presents machine learning and transfer learning methods for detecting informational queries on social media in the Bengali language using natural language processing and machine learning techniques. The approach involves analyzing the text of posts to identify keywords and phrases commonly used in informational queries and using this information to train machine learning classifiers, including transfer learning models. The method was evaluated on a dataset of real-world Bengali social media posts and was effective in accurately identifying posts that are informational queries with high precision and recall. This work aims to assist organizations and individuals in providing support to individuals seeking information, data, and knowledge on social media, particularly in low-resource languages.