Yahoo Searching Gets Social

The holy grail for search engines is to return the most relevant results. Because of the nature of the English (and some other) languages, though, two or more people searching on the same word will want different results; “bass” means one thing to a fisherman and another thing to a musician. Yahoo thinks it has found a way around that — by harnessing the power of community to deliver more relevant, personalized search results.

How many times have your friends pointed you to websites that you might never have stumbled across on your own? If your social network is anything like mine, you’ve probably lost track, and gained plenty of interesting sources of information you can use along the way. Don’t get me wrong; I’d never want to give up search engines for navigating the Internet and making some kind of sense out of the chaos. But search engines have a hard time distinguishing certain kinds of information in useful ways; that’s where a human interface can help.

Yahoo! came up with a new service, still in beta as of this writing, that it hopes will tap into social networks to improve its users’ search experiences. If you’re a regular reader of Yahoo’s search blog, you may have seen the post discussing it, dated June 28. It’s called My Web 2.0, but I like to think of it as “social search.” In effect, it lets users formally set up their contacts as “human filters” they can consult when they’re looking for something online.

My Web 2.0 takes personal search history one step further. You may have seen and even used a personal search history feature in the Ask Jeeves, Google, and Yahoo search engines. It allows you to keep track of searches you have performed before, and sites you have visited before; most of them let you turn the history “on” and “off,” search within the sites you have cached, and in general do other useful things with your results. The idea is that, over time, you build up a collection of sites in your areas of interest that, while not exactly bookmarked, serve as a smaller universe of resources which you can consult when you don’t want to be hit with the entire Internet – a universe, moreover, that may be more likely to deliver the information you need, since you have already consulted these sites on related matters.

With My Web 2.0, that universe expands to include your friends. Just as you might consult a friend who is wild about gadgets when you want to buy a cell phone that does everything, now you can consult searches that your friends have performed online if you want to find a good website for gadget reviews.

{mospagebreak title=The Original Problem}

Yahoo’s search blog claims that the genesis of My Web 2.0 goes back about two years, when one of its engineers wanted to buy a plasma TV. Naturally, he did a Web search to learn more about them. The search returned enough sites to make his head spin. So he consulted a friend, who told him about two good sites for plasma TV reviews. Problem solved, right? Well, yes and no; to the folks at Yahoo, that may have solved one problem, but in fact it pointed up the frustrations that many search engine users experience even with relevant results. Searching is just not “personal” enough.

It isn’t exactly the fault of search engines themselves. These modern marvels were designed by humans to serve the majority of humans. This means that they fall short in three areas, according to Yahoo:

  • “Opinion” queries. Which restaurant in South Florida serves the best chili? If you like spicy chili, your preference won’t be the same as mine – and search engines can’t really distinguish between us. While this may seem to be a trivial example, there are many other, more serious matters of “taste” (everything from political views to expertise in certain fields) for which search engines can’t easily make distinctions. Wouldn’t it be nice if a search engine could tell the difference between you and your grandmother querying for a good review site for personal computers?

  • Personal results. Sometimes I think Beatles fans and certain fruit lovers must get very frustrated with search engines. As far as most of the big names in search are concerned, “apple” means Apple Computer. Search on that word in Google, and you don’t even find any mention of the kind that grows on trees until the third page of results. Four pages later, the site for a comic book convention shows up, but nary a sign of Apple Records. Again, it’s not the fault of the search engines; they just deliver what they think is the correct, most relevant results for the majority of users. Of course, this brings up an expression you may have heard before: “the tyranny of the majority.”

  • Serendipity. Who hasn’t hunted for a particular book in a bookstore and found about three others in the same area that also looked interesting and begged to be taken home? While a search engine can give you great, relevant results, it won’t steer you to new items that you might find fascinating. At the risk of sounding clichéd, that’s what friends are for.

My Web 2.0 was designed to address these three search engine weak spots.

{mospagebreak title=How Does it Work?}

My Web 2.0 lets groups of users make their personal Web indexes available to each other. So, for example, if you have a group of co-workers who need to keep informed on particular topics, you can all share the websites you’ve discovered, complete with your attached comments about the sites. Users would be able to perform searches within the indexes, knowing that the results they get would be on target…because others have already determined that these sites are relevant for the kinds of topics they are likely to research.

Yahoo has thought through some of the ways that users will want to fine tune My Web 2.0. It is not necessary to make your entire personal Web index available to your community; you can keep links and pages private (so if you don’t want your co-workers to know that you’re secretly an opera fan, you don’t have to include the searches you performed for Luciano Pavarotti). Conversely, Yahoo plans to allow My Web 2.0 users to make their indexes available to all other My Web 2.0 users if they want (so you can boast to everyone about finding the one obscure website that reveals where science fiction writers get all of their crazy ideas).

MyRank, Yahoo’s personal search technology, powers My Web 2.0. It combines the usual search engine algorithms with the ability to tap into the communities formed by My Web 2.0 users to deliver better search results. Indeed, the entire idea behind My Web 2.0 is the ability to tap into the knowledge of people you trust.

My Web 2.0 works best in conjunction with the Yahoo Toolbar. Users can save pages to My Web directly from the toolbar; otherwise, pages can only be saved from a Yahoo search results page. The toolbar is available for both Microsoft Internet Explorer and Mozilla Firefox browsers.

One interesting change from My Web 1.0 is that Web pages are now saved with “tags” rather than within folders. The idea behind this seems to be that, by giving a page a name, then searching for that name (or tag), you could easily re-find the information. A single page can be given multiple tags, so you can increase your odds of finding it again. Yahoo describes tags as keywords or labels to group common items.
Another useful feature lets you save a copy of a Web page, not just the URL. This can be useful if you really need the content of that page, and you think it likely that the page will be updated. You will even be able to search the exact text of the page.

{mospagebreak title=Why This Matters}

This is a genuinely new approach to search. Yahoo sees it as complementing the regular kind of Web search performed with search engines, not replacing it. Though in some ways, it will allow whole communities of users to, in effect, create their own search engines. In its search blog, Yahoo states that “Over time, we envision communities using My Web to build their own search engines to capture and make accessible the knowledge of their community – search engines populated with the collective experience of a group of medical researchers, a community of PHP experts, a bird watching club, or members of a structural engineering consulting firm.”

Interestingly, Google apparently hasn’t thought of this yet – at least, as of this writing, there aren’t even any rumors that Google is working on something like this. It’s a very rare thing to catch the search engine giant flat-footed. Whether this will be enough to give Yahoo an edge remains to be seen.

The community aspect of My Web 2.0 could potentially give Yahoo’s site far more “stickiness” than it has had before. Users of My Web 2.0 need to be registered with Yahoo, of course; they can’t receive the benefits of My Web 2.0 unless they are signed in. Once they are signed in, they will see that My Web 2.0 fits in perfectly with Yahoo’s whole strategy of keeping people connected with each other (as with Yahoo 360, email, instant messaging, contacts, and many other features).

This is not the first time I’ve observed that Yahoo takes a very different approach to search from Google. For Yahoo, it seems that search is a means to an end – getting people connected with, well, whatever they need to be connected with, be it people or information. Google at least gives the impression that it considers search to be more of an end in itself. We’ll see which approach will win in the long run.

[gp-comments width="770" linklove="off" ]