Projects

The Etymology of the word Car is derived from the Latin word Carrus. In the modern world, Cars compete for an important role in the place of Luxury and elegance; choosing the car whether economy or luxury is a big problem, even when we choose the brand, people are confused to choose the perfect model. So Here comes the Car Evaluation prediction based on the Decision Tree Algorithm using Gini impurities in OOPS.

MedGPT Bot was developed using Large Language models with Prompt Engineering and LangChain technologies to summarize the medical files with Retrieval Augment Generation (RAG) as a principle, to avoid hallucinations to a feasible extent. A notable aspect of the MedGPT bot is its ability to process diverse medical data types, providing a text-based output, EN-FR language translation facilities, and voice synthesis, all of these will greatly enrich the user experience and user accessibility. The MedGPT Bot’s performance has been affected due to the integration of multiple LLMs and it developed with limited resources, resulting in some issues such as slow response time, delay in voice generation response, and restricted language translation capability. In the future, we plan to address these limitations by enhancing the MedGPT Bot by expanding the capabilities of the bot to process diverse data types including multimedia files, implementing a robust database infrastructure capable of storing various files, and training the bot with advanced multilingual and multimodal LLMs. This will empower the MedGPT Bot to serve as a highly efficient and trustworthy virtual assistant for the medical academia.

Despite the wealth of data available in YouTube video comments, manually analyzing sentiment across large volumes of comments is time-consuming and resource-intensive. ⏳ Additionally, accurately interpreting the nuances of sentiment expressed by users can be challenging. 🤔
To address this, our project seeks to automate the sentiment analysis process using advanced natural language processing techniques. 💻 By integrating the YouTube API with AWS DynamoDB for data storage and manipulation and utilizing HuggingFace Library’s powerful LLM sentiment models, I aim to develop a robust solution for sentiment analysis of YouTube video comments. 🛠️ My goal is to provide content creators and marketers with actionable insights into audience sentiment, facilitating informed decision-making and enhancing user engagement strategies.