Intelligent Travel Recommendation System

Amruta Habbu; Bach Hoang Pham; Dang Khoa Tran; Khyati Sharma; Ragini Jakkan

2023-10-24

PROJECT OVERVIEW

Objective: Develop an image classification ML model for diverse categories such as sunsets, mountains, and beaches.
Functionality: Categorize and identify images with high accuracy.
Added Feature: Offer travel recommendations based on user preferences.
Purpose: Address the need for a dependable image classifier in travel planning.
Impact: Enhance the overall travel planning experience.

THE MODELS:

Tag-Based Image Classification:

- Model Purpose: Classify images into categories like “sunset” and “scenery”.
- Training Data: Use image data and tags.
- Objective: Predict best-fitting image tags.
- Technique: Employ CNNs and pre-trained models.

Recommendation Model:

- Recommendation Focus: Suggestion based on preferences, location, and queries.
- Features: Utilize user profiles, ratings, location, and more.
- Alignment: Match images with user interests and context.
- Technique: Content-based filtering over collaborative methods.

EVALUATION METRICS

For the tag prediction model:

Confusion matrix.

For the recommendation system:

Mean Average Precision (MAP) or Root Mean Square Error (RMSE).

USER INTERFACE

Available Interaction using R Shiny Dashboard:

Search for keywords like “beaches.”
See photos related to their search.
View recommended photos based on their locations.

Conclusion

Collaborative Models:

Tag-based model: Image classification & tagging.
Recommendation model: Personalized image suggestions.

Integration:

Recommendation model harnesses user data & tag-based model results.

End Result:

Tailored image recommendations enhancing user experience.