Introduction to Knowledge Representation

2024/2025

Academic Year

ENG

Instruction in English

6

ECTS credits

Delivered at:

School of Data Analysis and Artificial Intelligence

Course type:

Elective course

When:

2 year, 1, 2 module

Course Syllabus

Abstract

You may have heard that the Turing Test was formally passed by a chat-bot named Eugene Gustman in 2014. He fooled one-third of judges that he is a person. But does the chat-bot think? Many believe it does not. It is just a clever piece of software that uses verbal acrobatics and trickery to hide the fact that it does not understand a word. But what is understanding? How can we measure it? Another test called Winograd Schema Challenge was proposed in 2011, which does not require conversation at all. It is technically much simpler and is based on the number of multiple-choice questions of a special form. Here are a couple of examples: 1. The trophy doesn't fit into the brown suitcase because it's too small. What is too small, the suitcase or the trophy? 2. Frank felt vindicated when his longtime rival Bill revealed that he was the winner of the competition. Who was the winner of the competition, Frank or Bill? These are very simple questions for humans but computers have a hard time answering them. A computer would need to possess so-called commonsense knowledge and perform commonsense reasoning to pass the test. In this course, we will explore how such kind of knowledge can be represented in formal systems and how reasoning is performed in such systems.