9th International Conference on ICT for Sustainable Development

Authors - Prakruthi Rajendra, R.K. Sneha, Chahak Jain, Tarun Kumar
Abstract - An important cause of carbon emission that goes unnoticed is that which is released from mass digital consumption. Digital carbon footprints amount to the usage of every device, server, and data centre, and produce an alarming rate of carbon emissions each year. An increase in digitalisation has led to the frequent use of smartphones amongst users, leading to an increase in internet usage and social interactions. This study explores the digital carbon footprints and their slow impact on climate change. Climate change is an advancing problem that is caused by the emission of greenhouse gases like carbon dioxide (CO2) into the environment, leading to an increase in atmospheric temperatures. The proposed framework helps address gaps in research and limitations of existing products while building awareness of digital carbon footprints. Therefore, the solution is a smartphone application called “Monochrome,” which helps in tracking and monitoring an individual's digital carbon footprint. The application follows sustainable design principles and ethical guidelines. It aims to help reduce overall digital consumption, having a positive impact on the health of individuals and the environment.

Authors - G Sornavalli, B Mitul Krishna, Mythreya Kesavan
Abstract - Understanding and visualizing data, a fundamental concept in statistics and machine learning, can often pose significant challenges due to its abstract nature and multidimensional complexity. In this paper, we introduce a groundbreaking approach to address this challenge through the utilization of Virtual Reality (VR) technology. Leveraging Unity, a versatile game development platform, we construct an immersive VR environment where users, especially students, can witness the dynamic adjustment of a regression plane to accurately fit the data points. This innovative method offers a dynamic, interactive way to comprehend the complexities of multiple linear regression, rendering it more accessible and intuitive. The VR simulation addresses the challenge of visualizing and understanding high-dimensional data by providing a tangible, three-dimensional representation, thus enhancing the learning experience for students Our research revolutionizes statistical education by bridging abstract concepts with tangible visualizations, transforming interpretations of multiple linear regression and opening new avenues in data science methodologies.

Authors - Supriya S. Gorde
Abstract - The present study endeavors to improve the relationship between educators and engineering students in India, since it is a crucial element that impacts the job marketability, personal development, and progress of graduates. Due to a lack of customization and efficient communication processes, traditional educational models frequently fail to create strong, lasting connections. Tailored learning as well as effective guidance levels is limited by the general nature of current methods to engineering pedagogy which ignores the requirements and career goals of individual students. The proposed study aims to close these gaps by introducing a complete framework based on machine learning (ML) that would transform the landscape of pedagogical relationships in an academic scenario. We make use of Support Vector Machine (SVM), Logistic Regression (LR) and Extended Gradient Boosting (XGB) as these approaches can improve student performance analysis for custommade learning routes. Leveraging these particular machine learning techniques makes sense because of their demonstrated capacity to analyze large, complicated datasets and derive actionable insights. These insights are essential for customizing education to student profiles, forecasting employment trends, and enhancing communication efficacy. When evaluated across numerous datasets, the suggested model significantly improved precision, accuracy, recall, and other important performance characteristics, demonstrating its profound influence. It makes sense to utilize these specific machine learning approaches since they have a track record of successfully analyzing vast, complex datasets and producing insightful results. These insights are critical for predicting job trends, personalizing instruction to each student's profile, and improving communication effectiveness. The proposed model had a remarkable effect on precision, accuracy, recall, and other key performance parameters when tested on many datasets.

Authors - Tadikamalla Prasrita, Tarun Kumar
Abstract - The study explores the junction between technology and personal safety for women through the creation of a multipurpose jewelry pendant. This wearable device, which includes an SOS alert system, a discreet camera, and GPS tracking capabilities, intends to solve the serious issue of women's safety. The research approach included a thorough evaluation of the present literature, in-depth interviews, and questionnaires performed on college students to learn about their safety concerns and preferences. Based on the findings of the research phase, the proposed framework and design concepts were created. The pendant aims to easily fit into women's lifestyles while offering an added level of security by utilizing user-centered design methodologies. Iterative variations in design were used during the evaluation phase to ensure that the final product met both functional and aesthetic objectives. In the end, the project aims to provide a complete safety solution, which will increase women's security and confidence. The wearable pendant aspires to enable women to overcome their everyday lives with an increased feeling of safety by blending technology and user-centric design, producing a more secure and inclusive environment.

Authors - Smriti Jain, Shreya Agarwal
Abstract - This research investigates the environmental impact of disposable footwear choices within the context of throwaway culture. The study employs a mixed-methods approach to bridge the gap between consumer footwear preferences and ecological consequences. A consumer survey revealed a concerning disconnect between these two aspects, highlighting the significant ecological burden of the footwear industry, particularly regarding landfill waste and carbon footprint. The research acknowledges the nascent market for sustainable footwear options. However, it emphasizes the need for user-centered and ecofriendly solutions. The methodology leverages a comprehensive approach, including interviews, surveys, mind mapping, brainstorming, and prototyping. This iterative design process aims to develop solutions tailored to the target audience's needs and preferences, ultimately enhancing user engagement and a sustainable future. Furthermore, the research also calls for the creation of better ways to handle and reuse the parts of old shoes, with a focus on efficiency and minimizing environmental impact

Authors - Harsh Chinchakar, Gaurav Ratnaparkhi, Atharva Tanawade, Saloni Raj Singh, Divyansh Modi, Amaan Shaikh, Vidya Patil
Abstract - Generating informative descriptions for photographs automatically has become an interesting but challenging endeavor. This paper introduces AI Image Captioning using ResNet-50 and LSTM (AICRL), a unified model that uses LSTM with ResNet50 for automatic image captioning. AICRL combines an encoder that uses ResNet50 to build a full image representation with a decoder that uses LSTM and a soft attention mechanism to predict the next phrase while highlighting particular features of the image. AICRL is evaluated using metrics like BLEU. It was trained on the Flickr8k and COCO 2017 dataset individually and optimizes the likelihood of the target description sentence given training photos. The results highlight the effectiveness of AICRL in creating image descriptions. Additionally, the PERSONALITY-CAPTIONS challenge is presented, to generate engaging captions through the integration of configurable characteristics of personality and style. A large dataset is applied to the models created by combining state-of-the-art methods in sentence and image representations. The proposed models demonstrate both robust performance on the unique PERSONALITY-CAPTIONS task and state-of-the-art performance on well-known datasets such as Flickr8k and COCO 2017. Online assessments confirm the best performance of the proposed model close to already existing models.

Authors - Ashwani Kumar, Prathamesh Kusalkar, Sanket Naitam, Harsh Pande, S. N. Ghotkar, S. S. Kumbhar, Shanu Kumar, Sanket Jiwane
Abstract - The goal of this paper is to create a machine learning based system for seasonal pricing patterns and dynamic pricing [1] of vegetables in wholesale markets across the country. Using machine learning algorithms, the system optimizes prices by taking advantage of quantity, seasonal, and regional aspects. Regression analysis and tree based machine learning techniques are used by the system to create customized predictions for various crops, analyze historical data, and use models for price forecasting [2]. Planning and profit-maximizing techniques work together to maximize resources, cut waste, and promote sustainable farming methods