Magnetics

Text summarization is a method for generating a summary of long texts by focusing on the sections that contain essential information while keeping the overall meaning intact. Its goal is to reduce the size of long documents, which would be difficult and expensive to process manually. With the current explosion of data circulating in digital space, particularly unstructured textual data, there is a need to build tools that allow people to extract insights from it. Taking notes is a popular practice for many people employed in situations where it is essential to keep track of what is said, such as during an online lecture. The art of note-taking does not entail taking down every single word stated but rather broad summaries of what is covered. Making succinct yet informative summaries is the key to successful note-taking. In this seminar, we will be discussing how we have used visualization and data storytelling techniques in our project to make informed decisions. This project aims to address the difficulties of note-taking by building an application that produces notes based on the transcripts generated by the Automatic Speech Recognition (ASR) technology of the meeting platforms. We relied on visualization concepts for three major decisions that would define our project as a whole. These decisions are the choice of online meeting platform, preference of text summarization model and the messaging platform choice. Speaker(s): Manoj Varma Alluri, Navaneeth Jawahar, Sharath Kumar Prabhu, Jeel Jani Register: https://events.vtools.ieee.org/m/313207

Credit card fraud detection is an ever-growing problem in today’s financial market with a rapid increase in plastic card usage worldwide. According to the Nelson report, by 2027, financial service providers are expected to take a $40 billion hit globally in credit card losses, a significant increase compared to previous years. Hence, data-driven decisions can largely help in mitigating that risk. We chose this topic to deep dive into different aspects of Fraud Data Analysis & Exploration using Interactive Tableau Dashboard. Tableau dashboards can be very powerful in driving data-driven decisions. We created an interactive dashboard to help stakeholders or less technical people to drive insights, understand data better, and help in business decisions making. The interactive feature helps users to add filters as per their needs and understand the data in a way they want to analyze. Our dashboard is dynamic and would be updated when several filters are applied together. Multiple filters can be added to multiple charts at the same time. These charts are intuitive which makes even new users easily interact and understand the data. Speaker(s): Priyanka Singh & Devy Ratnasari Register: https://events.vtools.ieee.org/m/313201

As a result of the fast development and spread of the COVID-19 pandemic throughout the world, people's everyday lives have been severely disrupted in recent times. One proposal for controlling the epidemic is to make individuals wear face masks in public. As a result, we require face detection systems that are both automated and efficient for such enforcement. We propose a face mask identification model for static and real-time videos in this research, and the pictures are classified as "with mask" or "without a mask." The model uses a Kaggle dataset to train and test. The collected data set contains over 10,000 images (considering 5,000 with mask and similarly 5,000 without) and has a 98 percent performance accuracy rate. The proposed model is computationally efficient and precise compared to Haar-Cascade & ANN. The application of this research are various, including digitized scanning tool in schools, hospitals, banks, airports, and many other public or commercial locations. Speaker(s): Henil Shah, Neenu Markose Register: https://events.vtools.ieee.org/m/312341