Question

You are tasked with building a Retrieval-Augmented Generation (RAG) system using Elasticsearch for document storage, Watson ML for model hosting, and LangChain for orchestration. The chatbot is supposed to query a large database of medical records and generate responses based on the retrieved information. During testing, you notice that irrelevant documents are often retrieved, leading to low- quality responses. What would be the best approach to improve document relevance in this RAG setup?

Accepted Answer

Update the Elasticsearch index by re-indexing documents using embeddings from the Watson ML model.

Answer

Fine-tune the Watson ML model on a dataset containing both queries and relevant documents, then update Elasticsearch with the model outputs.

Answer

Implement a BM25 similarity algorithm in Elasticsearch to improve document retrieval.

Answer

Pre-process the medical records in LangChain to remove irrelevant information before sending the query to Elasticsearch.

You are tasked with building a Retrieval-Augmented Generation (RAG) system using Elasticsearch for document storage, Watson ML for model hosting, and…

Take the free C1000-185 practice test

Other C1000-185 practice questions