Knowledge Bases in Chatbot
Introduction
In recent years, Large Language Models (LLMs) have revolutionized the way we interact with chatbots and AI-powered assistants. These models are trained on vast amounts of data, allowing them to generate human-like responses and engage in meaningful conversations. However, one limitation of LLMs is that they are not trained on user-specific data, which means they may struggle to answer questions or provide information that is unique to an individual user's needs. To overcome this challenge, we have utilized a technique called "Retrieval Augmented Generation" that enables users to create personalized repositories of information for their chatbots.
What is Retrieval-Augmented Generation (RAG)?
Retrieval-Augmented Generation (RAG) is a technique that combines the power of a pre-trained language model with a retrieval system. Instead of relying solely on the model's existing knowledge, RAG fetches relevant information from a user's given data files based on the query. This allows the model to generate more accurate and context-specific responses, even when the required information is not part of its original training data. In simple terms, RAG helps the AI to "look up" and use specific information when needed, much like how a person might consult a reference book.
Using Documents and URLs in Qolaba AI Chatbot
In Qolaba AI chatbot, the feature of using individual or combination of documents and URLs is available. To use this feature, users can upload files or scrap URLs from the + icon available on the prompt box. After uploading files, users can select the relevant files or documents to fetch the data from there to get specific answers using chatbot. However, users need to select the relevant files every time they want to use them. To solve this issue, we have developed the knowledge base feature.
Advantages of Knowledge Bases
One of the key advantages of Knowledge Bases is the ability to organize and group relevant information together. For example, if a student wants to create a Knowledge Base for a particular subject, they can easily integrate all the necessary books, YouTube tutorials, and audio files into a single folder. This way, users can keep their resources organized and easily accessible, eliminating the need to search for relevant PDFs or documents each time they need information on a specific topic. Users can create multiple knowledge bases on different topics and use a particular knowledge base to find answers on a specific topic instead of reselecting the same set of files every time.
Creating Your Knowledge Base
To create a Knowledge Base, follow these simple steps:
Step 1: Accessing the Knowledge Base Feature
Visit our chatbot page.
Look for a folder icon in the bottom right corner.
Clicking this icon will open a pop-up window where you can begin building your Knowledge Base by clicking on "+create knowledge base" option.
Step 2: Adding Diverse Content
Our system is designed to handle various types of content:
Documents: Upload PDFs, DOCs, or TXT files containing your important information.
Web Content: Add URLs of webpages with relevant information.
YouTube Videos: Include links to educational videos or lectures.
Audio Files: Add URLs of podcasts or audio lectures.
Images: Upload up to 20 images that contain important visual information.
To save time and costs, you can reuse previously uploaded content. Just visit the history section in the pop-up to select files or URLs you've used before. This feature is especially handy for frequently referenced materials.
Step 3: Automatic Information Extraction
Don't worry about the technical details! Our system automatically extracts text from your uploaded files, transcribes audio, and even describes images. All this information is then organized and stored in a special database that the AI can quickly access.
Step 4: Organizing Your Knowledge
You can create multiple Knowledge Bases for different topics or subjects. For example, if you're a student, you might have one Knowledge Base for "Biology 101" and another for "World History." This organization helps you keep related information together and easily accessible.
Using Your Knowledge Base
When you're chatting with our AI, you can enable the Knowledge Base option. This tells the AI to consider the information in your selected Knowledge Base when answering your questions. It's like giving the AI a crash course in your specific topic before you start your conversation!
By following these steps and utilizing the knowledge base feature, you can enhance the effectiveness and personalization of your interactions with the Qolaba AI chatbot.
Last updated
