Wikipedia-based Hybrid RAG

Inspiration

We aimed to create a better algorithm for recommending information and related content. Our goal was to improve the accuracy and relevance of recommendations.

What It Does

The project integrates Graph RAG and normal RAG to enhance Wikipedia article recommendations. It showcases how these techniques can work together for better information retrieval.

How We Built It

We used a notebook template and customized both Graph RAG and RAG. ArangoDB was chosen as the database to manage and store the data efficiently.

Challenges We Ran Into

There was limited information available about Graph RAG. Understanding its implementation and optimization required additional research.

Accomplishments That We're Proud Of

The project successfully demonstrated the integration of Graph RAG and RAG. It provided a strong showcase of their combined potential in information retrieval.

What We Learned

We gained valuable insights into Graph RAG and its applications. The project deepened our understanding of its role in improving recommendation systems.

What's Next for Wikipedia-Based Hybrid RAG

We plan to refine the algorithm to enhance accuracy and efficiency. Future improvements may include expanding the dataset and optimizing the retrieval process.

Built With

• google-notebook
• graph
• jupyter
• notebook
• python
• rag
• wikipedia