https://github.com/abhigyan126/similaritymap
This application calculates document similarities using BERT embeddings. It reads text files from a folder, computes pairwise similarities, and generates a heat-map to visualise the results. The similarity matrix can be saved as a CSV file, and the heat-map as an image.
https://github.com/abhigyan126/similaritymap
bert-embeddings cosine-similarity embeddings plagarism-detection
Last synced: 6 months ago
JSON representation
This application calculates document similarities using BERT embeddings. It reads text files from a folder, computes pairwise similarities, and generates a heat-map to visualise the results. The similarity matrix can be saved as a CSV file, and the heat-map as an image.
- Host: GitHub
- URL: https://github.com/abhigyan126/similaritymap
- Owner: Abhigyan126
- Created: 2024-10-16T11:28:29.000Z (12 months ago)
- Default Branch: main
- Last Pushed: 2024-10-16T11:52:31.000Z (12 months ago)
- Last Synced: 2025-02-05T12:18:00.216Z (8 months ago)
- Topics: bert-embeddings, cosine-similarity, embeddings, plagarism-detection
- Language: Python
- Homepage:
- Size: 18.6 KB
- Stars: 0
- Watchers: 1
- Forks: 0
- Open Issues: 0
-
Metadata Files:
- Readme: README.md
Awesome Lists containing this project
README
### **SimilarityMAP**
This application calculates document similarities using BERT embeddings. It reads text files from a folder, computes pairwise similarities, and generates a heatmap for visualizing the results. The similarity matrix can be saved as a CSV file, and the heatmap as an image.
---
### **Files**:
- **App.py**: A Tkinter application that requires a folder with text files. It creates a similarity heatmap for all the text files.
- **get_map.py**: A non-GUI application that performs the same task as `App.py`.
---
### **Screenshots**
---
### **Usage**
- Can be used to **automatically grade** documents by identifying the best match and comparing similarity.
- **Limitation**: Only provides deviation information, not whether the content is correct.
- Can also be used to **detect copied content and plagiarism**.
- **Limitation**: Only identifies plagiarism within the scope of the provided text context.