Can we build a machine to understand data automatically?

“Understanding” is a common term we use often. What exactly does it mean? What is an “understanding machine” like in terms of artificial intelligence? One possible such machine is a reading machine. This research project is aimed at providing an insight into the process and nature of understanding from AI point of view, with the ultimate goal of definitely (with mathematical precision) answering these questions.
This project involves the following tasks: study the textbook and literature, conduct research on the above title and write a report. The requirements are the following:
• There should be a case study where a piece of data (e.g. a text document or a video or an image) is used to show your machine understanding procedure or to show this is not possible.
• There should be some degree of novelty.
• This is a piece of coursework so you need to demonstrate your knowledge and skill relevant to the module in breadth and depth.

Stage one: focusing on the research plan.
You are expected to come up with some relevant research questions, suggest methodologies about how the questions are answered, and propose hypotheses with initial justification (reasoning). Example questions are the following:
• What is understanding? What do we mean when we say we understand something?
• What kind of knowledge do we use when we try to understand a story (written in English) or a video scene? Can we write down such knowledge in terms of objects, properties and relationships? Is ontology suitable for representing such knowledge?
• How do we actually understand a story? Do we need to represent the story? If so, what is the representation scheme like