2007/06/04

Automatic (censoring) filters

Por razões que compreenderão se lerem tudo, o resto desta mensagem será escrita em inglês. Outras questões que se prendem com o caso que aqui descrevo, podem ser lidas de forma bem mais completa no comentário deixado ficar pelo Pedro.

Dear worlwide scientific event organizers (and others using email marketing to promote worth related events),

Some of us will not be abble to tell you, that your efforts to promote scientif gatherings and other related serious activities, are not arriving to some (plenty?) of us.

The reason?

Due to massive attacks of spam in organizational settings, the so called «intelligent agents», known to the commun worker as «organizational filters» are blocking your messages promoting work related events.

Those «intelligent agents» do not really understand the difference between «Las Vegas» as in «Gambling», from «Las Vegas» as in «Call for papers»! See example below, from a message i've received (i've ommited the sender and the receivers of the message):

Sent: Sunday, June 03, 2007 7:44 AM

Subject: FW: SITE 2008--Technology & Teacher Ed. Call (Las Vegas, Nevada)

BANNED CONTENT ALERT

A banned word or phrase was discovered. The file has been removed. See your system administrator for further information.

Context: '[Message Body]'
Content Rule(s): 'Gambling (English) > Severity - Low'

I understand that this is a fiercely battle between «have to block unsolicited mail» and «have to find 'but' exceptions». Both parties, Spammers and Researchers, are doing their best. When researchers find a way to get better «intelligent filters», spammers read their published results and get even better. Ultimately they will both keep on getting better until... maybe it's time to use more of other ways of getting the events promoted, like blog posts, categorized event feeds... distributed micro content. In other words, human filters. The kind of filters that give that extra touch to software: social software.

For those of you that are sending out «email surveys», just let me add that this organizational filters also concern your data collecting in research. I've experienced that, back in 2003 when i was collecting data for my master thesis: my survey was a word document, with embedded macros (for easy click response), with pictures in it, and i had the express agreement of the individuals i was sending the messages (expatriates working for a multinational). The results? The filters of that organization didn't «like» macros nor did they allowed messages with images. You get the picture?


PS - In the organization i work for, it is easy to solve this kind of blockings, since the IT department trusts our judgments, and open to listen to our grounded arguments. But i do not know if this is the norm with all the people, in all the organizations. I suspect not!

No comments:

Post a Comment