| --- |
| license: mit |
| language: |
| - en |
| library_name: bertopic |
| tags: |
| - topic-modeling |
| - nlp |
| - amazon-reviews |
| - clustering |
| --- |
| |
| # Amazon Fine Food Reviews - BERTopic Model |
|
|
| ## Model Description |
| This is a **Topic Modeling** model built using the [BERTopic](https://maartengr.github.io/BERTopic/) library. It was trained on the **Amazon Fine Food Reviews** dataset to automatically extract and cluster the main topics and aspects discussed by customers in their reviews. |
|
|
| This model is part of an academic Web Mining project developed at EMSI Marrakech. |
|
|
| ### Intended Uses & Limitations |
| - **Intended Use:** Automatically extracting topics from English food product reviews (e.g., Taste & Flavor, Packaging, Delivery, Specific foods like Coffee or Tea). |
| - **Language:** English |
| - **Limitation:** The model is optimized for food-related e-commerce reviews and may not perform well on general text or other domains. |
|
|
| ### How to use |
| You can load this model directly in Python using the BERTopic library: |
|
|
| ```python |
| from bertopic import BERTopic |
| |
| # Load the model from Hugging Face |
| topic_model = BERTopic.load("aablaess/amazon-bertopic") |
| |
| # Predict topics for new reviews |
| docs = ["The coffee tastes great but the packaging was damaged.", "My dog absolutely loves this food!"] |
| topics, probs = topic_model.transform(docs) |
