August 3, 2006
Principles of enterprise text technology architecture
My August Computerworld column starts where July’s left off, and suggests principles for enterprise text technology architecture. This will not run Monday, August 7, as I was originally led to believe, but rather in my usual second-Monday slot, namely August 14. Thus, I finished it a week earlier than necessary, and I apologize to those of you I inconvenienced with the unnecessary rush to meet that deadline.
The principles I came up with are:
- Deploy search widely across the enterprise.
- It’s OK for your text data to be distributed across a range of silos.
- Integrate fact extraction/text mining aggressively into your predictive analytics and dashboards.
- Having a preferred enterprise text technology tool suite is nice, but accept that there will probably be lots of departmental exceptions.
- Reinvent your customer communication (and other) processes to exploit text technologies.
- Integrate your taxonomies.
I’ll provide a link when the column is actually posted.
Categories: Enterprise search, Ontologies, Search engines, Text mining
Subscribe to our complete feed!
Comments
One Response to “Principles of enterprise text technology architecture”
Leave a Reply
[...] After even more glitches than usual with their content management system, Computerworld finally posted the second part of my series on enterprise text technology architectures. I already posted the main points of the column here several weeks ago, but of course the column includes further material. In particular, I draw an analogy between text technologies and business intelligence, inspired in part by various direct ties between the two disciplines. Dave Kellogg makes a similar point, focused on general market development. [...]