Akshay Kumar Dhar Email & Phone Number

Belfast, Northern Ireland, United Kingdom

Belfast, Northern Ireland, United Kingdom

• Created an end-to-end NLP - machine learning pipeline to analyze artist-fan sentiment on social media to help unsigned musicians understand their audience better.• Scraped over 195k social media comments for 6 target music artists and pre-processed data for modeling. • Developed LSTM and RoBERTa models, optimizing RoBERTa using contextual embedding to improve accuracy by 5% to 78%. • Performed K-Means clustering to categorize artists' fanbases and quantify distribution across super, engaged and ambient fans types. • Used BerTopic for topic modeling to automatically discover relationships in artists' content.• Built interactive visualizations to monitor sentiment changes over time across platforms.• Presented sentiment analysis results and visualizations to stakeholders.• Recommended artists tailor content types and topics to specific fan types based on sentiment analysis and fan categorization.• This enabled data-driven recommendations on how artists can better resonate with fans through tailored content.Technologies used -1. Python2. NLP3. Web Scraping4. LSTM5. BERT and RoBERTa Models:6. Topic modelling: BerTopic7. Unsupervised learning - k-Means8. Supervised learning - Time series analysis.9. Topic modelling.10. Emotion Classification11. Microsoft 365 - For documentation and Presentation

India

• Conducted extensive EDA on a dataset of 950,000+ students using Python's pandas and NumPy, uncovering learning patterns and engagement trends.• Applied statistical measures and visualizations to unveil previously undiscovered patterns, contributing to a deeper understanding of student behavior• Developed and optimized Python scripts for data ETL processes, utilizing pandas and PySpark libraries.• Implemented parallel processing techniques, reducing data processing time by 20% and enhancing content delivery and assessment analytics efficiency.• Executed advanced SQL queries, including JOINs and subqueries, to extract and manipulate complex datasets.• Introduced indexing strategies, resulting in a 15% improvement in query performance for personalized learning recommendations.• Collaborated with instructional design and content development teams to align data analytics outputs with educational content enhancements.• Implemented machine learning algorithms for personalized learning path recommendations, resulting in a 15% improvement in accuracy for student segmentation and academic intervention strategies.• Created dynamic Power BI dashboards with interactive features, effectively communicating findings to stakeholders and presenting results to a diverse audience

Master'S Degree, Data Analytics, 2:2

Activities and Societies: QUB mountaineering• Data Analytics Fundamentals • Database & Programming Fundamentals • Data Mining • Machine.

Gurucharan College ,Bachelor'S Degree, Computer Science, 2:1

Activities and Societies: College Table tenis Team, Debate club.• Programming with C. • Digital Electronics & Fundamental Of Computers. •.

A Level, 2:1

Activities and Societies: School Table tennis team.

Gcse, 2:1

Activities and Societies: School Table tennis team.