10.5 MINING FOR INFO ON USENET VIA E-MAIL


Grizzled Usenet veterans (you can always tell them by the coffee-stained 
leather jackets they wear) proudly recall the days when they could read 
every single article posted on the network each day and still find time 
to do some work. 

But now, with the number of newsgroups approaching 10,000, that, of 
course, is impossible.  That causes a potential problem, though.  What if 
there's a discussion going on somewhere you might be interested in?  
Sure, Usenet is divided into hierarchies and newsgroups with the goal of 
helping people find discussions on specific topics, but given the number 
of people who now post each day, even that might mean you'll miss 
something.  And if you go on vacation and you come back to 2,000 new 
articles in your favorite group, the temptation is awfully high to just 
mark them all as read rather than trying to dig through them for 
useful/interesting messages. 

Meet Stanford University's Netnews Filtering Server.  Somewhere at 
Stanford sits a computer that creates a daily index of all Usenet 
messages that pass through it.  Through simple e-mail commands, you can 
get this machine to filter out articles for you and then send you a daily 
summary of what it finds.  If the summaries of each article look 
intriguing enough, you can then have the entire articles mailed to you. 

The basic commands are really simple.  You tell the computer what to look 
for and how frequently you want to receive its reports. Send an e-mail 
message to 

     netnews@db.stanford.edu. 

Leave the subject line blank, and as the message, write 

     subscribe phrase or word 
     period 1 

For example, 

     subscribe boston bruins 
     period 1 

would set the machine to searching for references to the Boston Bruins 
and then report back to you every day (if you substituted "period 2," it 
would report back to you every two days; you can go as high as 5).  
There's an optional third command, "expire,'' which you would use to tell 
the computer how many days to keep looking for you.  For example, 

      expire 30 

would end the search after 30 days. 

Now let's say you do get an article you want to read more about.  Each 
article will have a message number.  To get it, write back to 
netnews@db.stanford.edu and as your message, write 

       get news.group.# 

for example, 

       get alt.sex.hamsters.duct-tape.4601 

You can also search the Stanford database for existing articles.  Again, 
write to netnews@db.stanford.edu.  As your message, write 

       search word or phrase 

You'll get back a list of possibly relevant articles.