What is a Knowledge Base? 📚🤖

A knowledge base is like an information library that is classified and stored in a machine-readable form. It serves as an encyclopedia 🧠 that can be narrowed down and deepened based on a user’s query.

📌 How is a Knowledge Base Formed?

• 💡 Propositions: Fundamental statements or assertions.
• 🔍 Fact Extraction: Identifying and retrieving relevant information.
• 📋 Information Extraction: Organizing data for structured storage.

🌐 Relationship Between a Semantic Network and a Knowledge Base

A semantic network 🌐 organizes information into connected concepts and relationships. Everything in the semantic network is extracted and placed into the knowledge base, ensuring that facts and their relationships are structured for easy retrieval. 🧩

🚀 The Power of a Knowledge Base

In essence, a knowledge base transforms raw information into a usable resource for answering questions and retrieving specific facts efficiently! 🔥

Google’s Knowledge Bases

Google has knowledge bases with billions of facts. They use “data-to-text” to verbalize them. If you publish everything at once without brand authority or historical data, you’ll get stuck.

💡 Inference Example

Example: “We paid 20 at the Buckingham Palace gift shop” → Google might interpret it as 20 pounds.

💸 Cost of Ambiguity

This is expensive for them, which is why you want to be explicit and cheaper for Google, so they don’t have to run advanced inference. If you’re ambiguous, it costs them more.