Microsoft Looks At Other Search Engines

ZuperDee writes "It looks like Microsoft is now looking for another search engine to buy. They are looking at Ask Jeeves and Looksmart, but they recently dumped Looksmart, after deciding that its results don't stack up well. So would anyone be surprised if they bought Ask Jeeves? It can't hurt that according to Netcraft, they already run Microsoft IIS."

Comments

BS based on rumors

[melted]

Come on folks, RTFA. The article is just a bunch of rumors carefully worded to sound believable.

Re:Why not ask Jeeves

[sosume]

According to this link, AskJeeves and Look Smart are the same company...
so is this a marketing hype to keep Google stocks cheap for a hostile takeover??

Microsoft COULD Make it Work!!!!

[ZuperDee]

I think I should also point out that Ask Jeeves also own Teoma, which is absolutely nothing to be sneezed at.

Not only that, but Microsoft has a world-class research arm with Microsoft Research. With Microsoft Research's world-class research, and Microsoft's deep pockets, you can bet that any improvements Teoma would need to compete with Google WILL be made.

Re:But does anyone use them?

[BadCable]

But then google is also starting to suck. People are learning how to cheat their page ranking system, more and more "trap" sites exist which do nothing but link to a central site while also linking to each other. (For example, try finding jewlery sites - at least 30% of the responses will be fake sites that point to ONE site.) Sure Google isn't "dead" and it's not "dying" but it's certainly becoming more vulnerable to abuse and less accurate than it once was. With the millions that MS has to invest in an engine it's likley that they'll be able to provide just slighly better results than Google and use the "default page", "page not found" etc traps to promote their seach engine.

Ask Jeeves uses better CYC Tehcnology

[thenarftwit]

They use the CYC AI database thecnology so that you can ask a question in plain english..(much better than all the over "dumb" search engines combined). I find that the Ask search engine is better at finding stuff that the dumber search engines are not capable of remotely gettin close too. As far as microsoft buying them, I am not too exitied about that company owning more things in the whole universe, it's bad enough that 98% of all computers runs their crap OS's, and that they spend all their time constantlly changing (but not improve) their languages and OS's (to take over the world), now they have grabbed of the better search engines.

Re:Why buy, when you can build?

[Anonymous Coward]

I don't understand why they need to buy an engine. It may be shortsighted of me, but building one would probably cost less and could be done failry quickly.

They tried this. I have been contacted no less than three times over the past year by headhunters from Microsoft looking for somebody to architect a new search engine for them. Given the timespan it would seem that they aren't having much luck finding qualified people. I told them to bugger off myself because I wouldn't want to work for Microsoft and I don't think they would want me there either - it's blatantly obvious from my website that I am hooked on Linux and abhore Microsoft, and I can only conclude that they have entirely incompetent headhunters (they were third party headhunters) or they are having a very hard time finding people (which must be true to some degree given that it has taken them this long).

Best Jeeves Answers:

[missing000]

Are at Satirewire.

Re:What exactly does Microsoft want?

[stratjakt]

This is all speculation. Noone has any proof that MS is looking to buy anything.

Kartoo is nifty!

[Thud457]

Kartoo is helpful when you're not having luck with the obvious google searches. You can start with a more broader search, and then use the results it brings up to refine you search. Errr, just go play with it, and you'll see.

Re:But does anyone use them?

[aridhol]

Maybe they should follow redirects and index only the final page. This avoids indexing the intermediate pages that are there only for redirects, and also allows you to move your site somewhat more easily by adding a redirect to your current page.

As for the text/background idea, what about running the page through something similar to SpamAssassin. Not all tests are applicable for websites, but a new ruleset shouldn't be too hard for them to write.

Also, do they keep all pages for their cache, or just ones with a high enough PageRank? How difficult would it be for them to compare two pages for similarity, and lowering the rank for identical pages? Or for pages that have the exact same set of links (to reduce the effectiveness of link farms).

Due to the nature of the PageRank algorithm, lowering the score on the referring pages will have a result on the target page, while ensuring that it's difficult to get someone dropped completely (since other, high-ranking sites may also be pointing there).

Re:But does anyone use them?

[hankwang]

>I wonder why google doesnt parse the page for redirects, and drop the rank if the page does? [...] if the text and background colors are the same

They do. Search for google penalty redirect, google penalty hidden.

Redirects (with a meta refresh tag) are penalized if they happen within 10 seconds. However, Google probably won't catch redirects implemented with javascript. To parse javascripts that potentially end up in infinite loops is not an easy task.

Re:But does anyone use them?

[selkirk]

I wonder why google doesnt parse the page for redirects, and drop the rank if the page does? ... But their weakness is they all do a redirect.

Google does. However, search engine spammers use a technique called cloaking. They know which IP numbers google uses to scan the web. So, they dynamically show a page to googlebot that does not have redirects and that ranks highly for keywords. You are not coming from the "cloaked IP," so they show you a different page -- the commercial redirected page. Report this to google when you see it. Its the only way they can find these cloaked redirects.