WhatsApp Forensic Analysis for Group Chats Using Exploratory Data Analysis and Natural Language Processing
摘要
WhatsApp has become an integral part of our lives. People have started using WhatsApp for various purposes from business communication to leisure chat, it connects people through functions like text and voice messages, images, gifs, emojis, voice and video calls, share statuses, and is even used for payments. WhatsApp forensics is a globally hot topic due to its active involvement in our daily lives as it is an excellent source of data extraction that can be visualized to benefit society in different domains such as law enforcement, business studies, psychological studies, healthcare, educational research, and much more. The WhatsApp forensic analyzer proposed in this paper deals with an evidence-based study of the data which helps in the behavior and pattern analysis of the members of the group. This is done using techniques like Exploratory Data Analysis, Sentiment Analysis, and Forensic Algorithms. Exploratory Data Analysis gives a visual view of the data using various tools like bar plots, spider plots, line plots, histograms, and word clouds. Sentiment Analysis utilizes Natural Language Processing to study the emotions of group members. Forensic Algorithms, such as the extraction of potential criminal keywords, translation of messages containing potential criminal keywords, extraction of links, and knowledge graphs, help forensic analysts in the forensic study of group chat data. It also tries to give results in visual form to enhance readability to the normal person.