Data Fusion

• If your data comes from different sources, you usually have to put it together to get the benefits of all the sources.
• This putting together is called data fusion or data integration.
• This is a long standing problem with databases, but is now even more important.
• With databases it involved data cleansing.
• Now, with distributed databases, you have problems with maintaining consistency.
• Moreover, there is the ongoing frame problem. How does the data change with time?
• Moreover, you still have the problems of putting different data sources together.
• If there is one database that says Chris Huyck is the Reader in AI at Middlesex, a second says he's the Professor in AI, and a third says Christian Huyck is the Professor in AI, which is right? Are there three Chris Huycks? What about John Smith?