This repository contains a comprehensive Hotel Recommender System developed using Aspect-Based Sentiment Analysis (ABSA). The project aims to provide insightful hotel recommendations based on customer feedback, identifying hotel-specific strengths and actionable areas for improvement.
Through thorough data cleaning and preprocessing, I improved the frequency count of hotel review aspects by 81%, increasing the total aspect-related mentions from 7791 to 14135.
| Aspect | Before Count | After Count |
|---|---|---|
| Food | 1,864 | 2,990 |
| Staff | 2,141 | 2,969 |
| Service | 1,705 | 3,020 |
| Clean | 456 | 1,028 |
| Room | 1,625 | 4,128 |
The dataset for this project can be found on Kaggle: