Very basically we have created a technology which can represent the world’s knowledge in a form that is clear and accessible to humans, as well as being comprehensible to computers. This is different from the knowledge stored in websites and books, which is written in natural language that is good for humans, but incomprehensible to computers.
Interesting initiative from True Knowledge. This semantic search engine is semantic not in the way we generally think about (~data that are linked by semantic classes) but if the demo is true it is a nice work around.
To simplify what I understood so far:
- the user type a query like “When was the Eiffel Tower built”
- the query is translated from Natural language to an understandable query for the system
- the system then aggregate answer from a knowledge database and external feeds
- finally the user see an answer to his question
Some screenshots here.
This type of answer can already be extracted with ask.com. However the answer isn’t as clear as the one provided by True Knowledge (IMHO).
In the title of this post you can also read “User Input”, it comes from the fact that when the system cannot provide an answer to your query it will ask you to provide the answer (would I really asked the question if I already knew the answer?). I like the integration of Users knowledge to increase there Knowledge database, now they will have the questions of data reliability.
So far Google isn’t dead but we see more and more “semantic search engine” in beta version and Google didn’t yet made an alternative proposal. Maybe they are just waiting for the better one to win the race, then they will buy the company…
In terms of advertisement, semantic understanding of natural language can be a real shift in pertinence of advertisement.
If you are interested to test there technology they have launched the private beta.
I’ve read about this via Read/WriteWeb in an article written by Marshall Kirkpatrick
Ahmet