remember altavista, and when you first started using google, you felt guilty? for abandoning altavista?
1st gen search engines: search engines “crawl” links to pages, they make a copy in something called a index, they find pages you are looking through, originally via term frequency. this was too spammable, because control was in the hands of the webmaster.
2nd gen search engines: use factors off the page that wemaster can’t easily influence
- clickthrough
- links “democratic nature of the web”
- pagerank
- anchor text, actual words in the text
- then people began overtly manipulating links, thinking about votes, campaigning for votes (miserable failure), even buying votes
3rd gen?
- vertical search
- focus on a particular topic, such as news
- personalized and social search
- reshape based on what you are doing
- what your friends are doing
- what groups are doing
- http://searchengineland.com/070827-121805.php
- Google personalized search
- ranking is reorganized based on whats deemed to be your personal preferences
- Changes are subtle, but will likely change over time
- personalize influencers
- google homepage
- google bookmarks
- search history
- web history
- social search
- eurekster experimented with friends clicks reshaping results in 2004
- Yahoo My Web promised to let us tag and use a network to reshape results
- Neither really has suceeded
- the promise & reality of mixing the social graph with search engines
- eurekster says “swikis” are much better
- yahoo dropped many feautres quietly
- but what about facebook.
- Social graph (ugh) social network data potentially useful
- watch what others are searching on
- monitor clicks in a more closed environment – harder to spam, identity is persistant
- reshape results based on what your friends seem to like
- but WHO are your friends?
- www.dumpfolder.net/?p=193
- friend pollution
- do you really want to go through and pick out friends you trust enough to influence search results
- what about unfriending, privacy, and what you want to share?
- tagging? search basically works, and sharing queries is undesirable/unnecessary
- Does facebook instead work on a aggregate level? use networks, for example.
- And what’s the underlying platform? They’ll probably partner with someone else (Facebook unlikely to build a search engine from scratch, it’s not easy)
What shall they do?
- Go vertical? People search?
- plenty in space, spock et al
- searchengineland.com/lands/peoplesearch
Search vs discovery
- search is an on demand thing, have particular need to fulfill activity
- Discovery is related but less specific in what you want
- stumbleupon, digg
- iGoogle related magic tabs
QUESTIONS
Q: Maholo?
A: I like it, i think it could help. Google says, we won’t touch it we’ll fix the algorythm but hey, fix it now! But maholo is starting to morph into a wikipedia. SInce it’s made by an editor, it’s more of an about than a social search engine.
Q: don’t people already want to separate top friends, professional friends, etc?
A: if you are a heavy user, then yes, but most people that’s too much overhead. have to be a poweruser.
Q: don’t you think people search is fundamentally different than web search?
A: absolutely, you are looking for a page with bio, contact info, etc. But the problem is where are they? myspace, facebook, etc? Easier to go to google or a people search.
Favorite quote
speaker “google will say we’ve got iran, and we’ve got brazil and we’re coming for you”
audience “the axis of orkut”