However, I’m not unsympathetic to the White House’s IT staff. According to a NYT article, they receive over 15,000 messages a day (no word on how much of that is spam.) They need a classification system. Fortunately there is such a tool. Here’s what I wrote to the White House Web team.
I read about the new email feedback system for the White House from a couple of sources:
Having tried [the new web form] myself, I’m not satisifed with the soulution either. However, I certainly sympathize with the problem of classifying the 15,000 emails you receive each day, as the Times article cites.
May I suggest an alternative strategy for classifying the inbound mail?
Bayesian rules, as Paul Graham describes in his article at http://www.paulgraham.com/spam.html, can not only be used to filter spam, but classify non-spam mail as well.
PopFile (http://popfile.sourceforge.net/) is an implementation of a Bayesian filter that will run on both Windows and Unix systems. It provides a web interface that an intern or non-technical specialist can use to develop filters which will stamp inbound mail from both POP and IMAP with additional headers and/or additions to the subject line which any conventional email client can use to deliver messages to inboxes.
More importantly, it does not require the person sending a message to pick a subject from a list. Thus, as new issues develop and citizens write to the White House, the filter can be trained to recognize those new subjects.
The new web mail form tries to deal with the flood of messages, but inadvertently creates a barrier to citizen participation. A filtering system that routes messages post-hoc removes the hassle for the citizen, while bundling and sorting messages at the receiving end.
Like I said, I don’t doubt the White House is drowning in email they don’t know how to parse, but there are tools for this, and you don’t have to annoy the citizen writing to the President in order to use them.
Possibly Related posts (machine generated):