- Joined
- Oct 18, 2004
- Messages
- 391
- Reaction score
- 29
- Points
- 4,671
Advertisement - Members don't see this ad
http://www.v3.co.uk/v3-uk/the-frontline-blog/2035264/ibms-watson-moves-jeopardy-pharmacy-assistant
Watson can understand human language syntax and data mine not only from preexisting databases but also unorganized data through context cues. Very different from a search engine or an automated telephone system.
Here's the Watson architecture:
For the Jeopardy match, the sources of information for Watson include encyclopedias, dictionaries, thesauri, newswire articles, and literary works. Watson also used databases, taxonomies, and ontologies. Specifically, DBPedia, WordNet, and Yago were used. Although Watson was not connected to the Internet during the game, it contained 200 million pages of structured and unstructured content consuming four terabytes of disk storage, including the full text of Wikipedia.
Watson can understand human language syntax and data mine not only from preexisting databases but also unorganized data through context cues. Very different from a search engine or an automated telephone system.
Watson took on two Jeopardy champions and came out victorious, helped along by its 3,000 core processors, 15TB of data and 80 trillion instructions per second.
IBM's Mills gave the example of a pharmacist who could use the data crunching technology behind Watson to offer a better patient service, running analysis to recommend a generic cheaper medicine than the normal prescription and checking this against other medicines being taken.
"We see what we did with Watson as an obvious way to create a pharmacist assistant," Mills said. "An assistant that doesn't get tired or emotional, and doesn't walk out the door and quit."
Mills added that IBM was already embarking on healthcare projects based on Watson, but declined to comment on any costs involved. Based on the computing power behind the machine, it's fair to assume this would be a hefty outlay, so CIOs in most organisations will likely have a long wait before they get their hands on the technology.
Here's the Watson architecture:
For the Jeopardy match, the sources of information for Watson include encyclopedias, dictionaries, thesauri, newswire articles, and literary works. Watson also used databases, taxonomies, and ontologies. Specifically, DBPedia, WordNet, and Yago were used. Although Watson was not connected to the Internet during the game, it contained 200 million pages of structured and unstructured content consuming four terabytes of disk storage, including the full text of Wikipedia.
Last edited: