-
Description: An intelligent web interaction tool that harnesses the power of RAG (Retriever, Aggregator, Generator) flow to facilitate seamless communication with websites.
-
Functionality:
- Upon receiving a webpage link, WebChatter swiftly scrapes the webpage, extracts its textual content, and divides it into manageable chunks.
- These chunks are then embedded with metadata and stored in a Pinecone vector database.
-
User Interaction:
- Once the index is established, users can engage in conversational queries.
- By inputting a query, WebChatter generates embeddings for the query and calculates cosine similarity scores between these embeddings and the embeddings of the webpage chunks.
- The chunks exhibiting the highest similarity scores are selected and passed to the Language Model (LLM) along with the query, facilitating accurate and targeted responses.
-
**UI Demo
-
- Above is the image where User have to input the Webpage link.
-
- Above is the image where User will input his query and chat with the Webpage
-
- Above is the image where User receives his response from the bot.
-
Notifications
You must be signed in to change notification settings - Fork 0
"WebChatter: Utilizes RAG flow to scrape, extract, chunk, and embed webpage text into Pinecone database. Enables query-based interactions by calculating cosine similarity scores between user queries and webpage chunks, directing them to the Language Model for precise responses."
License
omkar806/WebChatter
Folders and files
| Name | Name | Last commit message | Last commit date | |
|---|---|---|---|---|
Repository files navigation
About
"WebChatter: Utilizes RAG flow to scrape, extract, chunk, and embed webpage text into Pinecone database. Enables query-based interactions by calculating cosine similarity scores between user queries and webpage chunks, directing them to the Language Model for precise responses."
Resources
License
Stars
Watchers
Forks
Releases
No releases published
Packages 0
No packages published