Creeping Inevitability: Search Engines Index Tweets

The Fail Whale

Microsoft’s search engine Bing has struck a deal with Face­book and the hot micro-messaging ser­vice Twit­ter, a brash attempt to add real-time web updates to its search results in order to make Google look like a lum­ber­ing dinosaur.

While we’re still digest­ing the news of Bing adding Twit­ter to its search engine, Google has some news of their own: they’re about to do the same exact thing.

Search­ing Twit­ter traf­fic ini­tially seems kind of odd.  Then, after a bit of thought, it starts to sound a bit bet­ter.  Finally, after a lot of reflec­tion, it starts to ven­ture into the realm of the bizarre.  Here a quick sum­mary of the stages of acceptance:

1 — Odd­ity:  Why would you want to add Twit­ter traf­fic to a search engine?  Twit­ter is all about quick thoughts between you and your ten thou­sand clos­est friends.  Search engines are sup­posed to be good at answer­ing ques­tions, and most ques­tions require more than 140 char­ac­ters to answer com­pletely.  (I’ll blithely ignore the ques­tion of whether most peo­ple doing searches actu­ally care about com­plete answers.)  On first blush, it doesn’t seem like a great match.

2 — Sense:  So why would Google and Microsoft be inter­ested in index­ing tweets, then?  Assum­ing it’s more than just a PR chasing-buzzwords stunt (not nec­es­sar­ily a good assump­tion, but…) how would index­ing tweets add value to the core of their search busi­nesses?  Well, if you treat tweets more as meta­data than search data, it starts to make more sense.  A good per­cent­age of tweets con­tain links, and once you fil­ter out spam you’re left with a lot of links that have been determined–by actual humans!–to be inter­est­ing.  Machine intel­li­gence is great, but humans are still bet­ter at fig­ur­ing out which pages are worth­while and which aren’t…this is the idea behind Mahalo and the like.  If Google and Microsoft can mine that data to improve the qual­ity of their hits, index­ing tweets sud­denly makes more sense.

3 — Non­sense:  But if that’s the goal, why make the tweets them­selves search­able?  That’s going back to treat­ing the tweets as actual data again, which seems ques­tion­able at best.  If peo­ple start to see their tweets show up in search engines that will change the way Twit­ter is used.  Right now it’s treated as an ephemeral medium; incor­rectly in the­ory, since Twit­ter is already search­able, but given how well Twitter’s search engine works, it might actu­ally be true.  With sto­ries pop­ping up all over the place remind­ing peo­ple to be cau­tious about what they put on Face­book or MySpace, does Twit­ter really want to be included in the list of ser­vices to fear?  I’m sure there are good reasons–probably money–for Twit­ter to get involved in this deal, but it’s not with­out risk.

via Bing Part­ners With Twit­ter and Face­book for Real-Time Search and  BREAKING: Google Announces Search Deal With Twit­ter.

Creeping Inevitability: The Internet Improves Literacy

Image by sebilden

Image by sebilden

The fact that stu­dents today almost always write for an audi­ence (some­thing vir­tu­ally no one in my gen­er­a­tion did) gives them a dif­fer­ent sense of what con­sti­tutes good writ­ing. In inter­views, they defined good prose as some­thing that had an effect on the world. For them, writ­ing is about per­suad­ing and orga­niz­ing and debat­ing, even if it’s over some­thing as quo­tid­ian as what movie to go see. The Stan­ford stu­dents were almost always less enthu­si­as­tic about their in-class writ­ing because it had no audi­ence but the pro­fes­sor: It didn’t serve any pur­pose other than to get them a grade.

There’s an inter­est­ing piece of research out there called the Stan­ford Study of Writ­ing that points out that the sky, in fact, is prob­a­bly not falling when it comes to writ­ten lan­guage.  Despite the occa­sional puff-piece arti­cle describ­ing a 4th grade teacher’s frus­tra­tion with LOLessays, there is more writ­ing going on now than at any time in the past, and the qual­ity is actu­ally improv­ing with all that prac­tice.  Maybe that’s just because the study focused on col­lege fresh­men instead of 4th graders, I don’t know.

I won­der, though, if the gains that have been made by forc­ing social­iza­tion to be text-based are tem­po­rary.  There is a con­stant push to make computer-mediated com­mu­ni­ca­tion more audio and video based that ranges from pro­grams like iChat and Skype to YouTube’s user-generated broad­cast model.  Though­out the devel­op­ment of the net there has always been this idea that we’re just wait­ing for the tech­nol­ogy to get us to the point where we can use it as a glo­ri­fied video­phone.  Text has a lot of advan­tages, though…maybe video isn’t where we actu­ally want to end up.

via Clive Thomp­son on the New Lit­er­acy.

Creeping Inevitability: Using Twitter Hashtags, Just Like Usenet!

Hash­tags are essen­tially a sim­ple way to cat­a­log and con­nect tweets about a spe­cific topic. They make it eas­ier for users to find addi­tional tweets on a par­tic­u­lar sub­ject, while fil­ter­ing out the inci­den­tal tweets that may just coin­ci­den­tally con­tain the same key­word. Hash­tags are also often used by con­fer­ence and event orga­niz­ers as a method of keep­ing all tweets about the event in a sin­gle stream, and they’ve even been used to coor­di­nate updates dur­ing emergencies.

via HOW TO: Use Twit­ter Hash­tags for Busi­ness.

Not a whole lot to say on the con­tent of the arti­cle itself as its pretty basic and self explana­tory.  What I find inter­est­ing, though, is the hash­tag idea itself.  It breaks the idea of a social net­work because you’re no longer just broad­cast­ing to your friends, you’re attach­ing a global tag that can be found any­where.  I won­der what would hap­pen if Usenet were intro­duced as a new prod­uct today?  That’s essen­tially what we’re look­ing at with tags, except for the char­ac­ter limits.

The fun part about Usenet was that it cre­ated com­mu­nity by start­ing with the tags (group names, as top­ics, are essen­tially just tags).  Twit­ter, though, starts with a loose com­mu­nity and has now emer­gently added tags to it.  What do the net­work maps look like for both ser­vices?  It seems to me that Usenet was actu­ally a more ‘social’ ser­vice in the sense that new com­mu­ni­ties were formed…users used to be more atomic, whereas now they’re more locked in to their indi­vid­ual, pre-determined cir­cles.  Not that Usenet was a utopia of course, even the groups that weren’t over­run by spam, porn, and bina­ries even­tu­ally devel­oped into impen­e­tra­ble tribal cliques…but that seems to be largely how Twit­ter is start­ing out.

Creeping Inevitability: AdSense for Feeds

googleads

Well, the creep­ing may have been done in the past since the news is pretty old at this point, but it scores high enough on the inevitabil­ity scale to make up for it.  Google has a ser­vice that inserts their AdSense ads into RSS feeds.  The pro­gram hasn’t exactly taken off, but then again, nei­ther has RSS as a whole, depend­ing on what you think RSS is really for.

RSS is sup­posed to be about syn­di­ca­tion (hence the name).  His­tor­i­cally, syn­di­ca­tion mod­els haven’t tried to send con­tent direct from the author to the con­sumer, they just spread con­tent to a num­ber of dif­fer­ent out­lets that then sell the aggre­ga­tion of that con­tent (see also: news­pa­pers).  Peo­ple have an idea that RSS is about direct-to-customer dis­tri­b­u­tion, and that has been rein­forced by all the news­reader pro­grams that are out there and sev­eral years of hype.  But it is even more pow­er­ful when it’s actu­ally used for syn­di­ca­tion:  get­ting my con­tent on another web­site is more use­ful to me than get­ting my con­tent sent directly to the user.  If my con­tent is on another site then that is pub­lic­ity that can draw in com­pletely new audi­ences.  Send­ing my con­tent to the user is con­ve­nient for the user and poten­tially makes them ‘stick­ier’, but it also deprives my site of hits and can only reach one per­son at a time.  To put it in MBA terms, RSS is far more pow­er­ful when it’s used in a business-to-business model than direct-sale.

I’m a lit­tle shaky on the idea of 3rd party ads being inserted in RSS because it isn’t most effec­tive as a direct to the user tech­nol­ogy and it makes my feed less attrac­tive to other con­tent providers.  I know that, as a devel­oper, if I wanted to use a feed on my site and saw it had ads in it the first thing I would do is fil­ter them out (assum­ing I didn’t just drop the feed com­pletely).  I gain noth­ing by let­ting some­body else sell ad space on my site, and as a prac­ti­cal mat­ter, none of my sites are intended to be com­mer­cial so ads would just look out of place.  If the the­ory is that AdSense allows con­tent pro­duc­ers to mon­e­tize RSS feeds it seems like the mar­ket just isn’t there to sup­port it.