TourBot

Inspiration

The inspiration for TOURBOT came from the increasing demand for personalized travel planning experiences. Current tools often fail to adapt to individual user preferences, leaving a gap in tailored recommendations. With the massive Yelp dataset and advanced AI technologies available, the goal was to create a tool that simplifies travel planning while providing personalized and efficient recommendations.

What it does

TOURBOT is an AI-driven chatbot designed to transform travel planning by offering personalized recommendations. It leverages user queries, preferences, and Yelp data to generate travel itineraries and suggestions for restaurants, attractions, and accommodations, ensuring a highly customized experience.

How I built it

TOURBOT was developed using the Yelp Open Dataset, focusing on business and review data relevant to tourism. After extensive data cleaning and preprocessing, advanced deep learning models like Bi-LSTM, GRU, T5, and GPT-2 were fine-tuned to process user queries and generate contextually relevant recommendations. The frontend was built with HTML5, CSS3, and JavaScript, while the backend utilized Flask to integrate the models and manage user interactions.

Challenges I ran into

Key challenges included handling the vast size of the Yelp dataset, ensuring efficient preprocessing, and optimizing the performance of deep learning models. Balancing the trade-off between computational efficiency and model accuracy, particularly for Bi-LSTM and GRU, was a significant hurdle. Additionally, fine-tuning GPT-2 to produce contextually accurate recommendations while avoiding overfitting required extensive experimentation.

Accomplishments that I'm proud of

The successful integration of multiple deep learning models to provide personalized travel recommendations is a significant achievement. Fine-tuning GPT-2 to handle diverse and dynamic queries was particularly rewarding. TOURBOT's ability to deliver accurate and contextually relevant suggestions represents a significant leap in AI-driven travel planning.

What I learned

This project deepened my understanding of natural language processing, deep learning models, and data preprocessing techniques. I also learned about the importance of model fine-tuning, hyperparameter optimization, and the trade-offs between model complexity and computational efficiency. Additionally, I gained valuable insights into deploying AI-driven solutions in real-world scenarios.

Built With

• python