Sentiment Analysis

Inspiration

I just of projects that would require fine tunning AI models and this was stood out to me.

What it does

The tool leverages web scraping (via Selenium), natural language processing (NLP), and machine learning models to classify customer sentiment as positive, negative, or neutral with a 90% accuracy.

How we built it

I used python3 as the main language for this project, PyQt5 for the GUI, Selenium & BeautifulSoup for scrapping static & dynamic web, Transformers (bert-base-uncased) for the AI and its fine-tunning and Matplotlib and WordCloud to visualize the final results.

Challenges we ran into

Fine-tunning the AI model with correct and relevant data and training the AI based on the data took a lot of resources.

Accomplishments that we're proud of

I am proud of this project overall, since i did everything by myself from scratch and overcame the challenges and the time required to complete this project.

What we learned

I have learned to have patience and be calm even when things aren't going my way. I also learned how to divide the project into smaller parts that make up the final code.

What's next for Sentiment Analysis with Web Scraper

I would like to train the AI model further with better data and make it more accurate. I am looking for people to collaborate with to further develop the idea and its functionality.