News

Here are the responses I received to my question about text mining
solutions embedded in e-mail management response systems.

--
Guy Lapalme from U. Montreal wrote

you might be interested at looking at our research project at the Université
de Montréal which deals with automatic e-mail follow-up
http://www.iro.umontreal.ca/~kosseim/Extraction/LUB.en.html

---
Charles Huot, Director at TEMIS (Text Mining Solutions), based in Paris,
France, sent a presentation about their solution called Email Miner.

It contains rules allowing the user to extract knowledge from emails.
Rules include Morpho-syntactic and semantic tagging,
Semantic components building rules, Negation rules, Modalities rules
and Managing the implicit.

Among other interesting features is
quantitative evaluation of the degree of customer (dis)satisfaction.

See www.temis-group.com or contact Charles at charles.huot@temis-group.com
for more information.

---
Alessandro Zanasi of Meta Group wrote

I think that some systems are well suited to solve specific
problems. Customizable (for specific languages particularities)
products will be the core of the more efficient solutions.
Generally speaking, I consider text mining useful to mine
unstructured data as emails. I expect (at least I hope so!) to
present in  Data Mining 2002 conference a paper about email text
mining, used a) to understand the moral inside a large entreprise
employees b) to measure the change in managerial culture in a
period of two years (always analyzing the content of the emails).