Microsoft Acquires Semantic Search Engine, Powerset

You know, these days whenever I see these topics I think, okay, there goes another $100 million dollars that could have been best used elsewhere. Even if this is the technology of the century, could Microsoft do anything with it? I mean, they've pretty much shown that they are desperate right now and will do anything at all to gain another percent of the search market or to not lose any more of what they have.

So now the battle moves to fighting for #2, #3, #4 and #5. #1 Google is clearly out of reach even for #2 through #5 combined.

Right now, I'm rooting for anyone who can even stay in the race. ;)

I wonder how it handles keyword stuffing and autogen content. Also trust and spam detection in general. That is where google comes into its own in comparison with other engines.

>>Google is clearly out of reach even for #2 through #5 combined.

There's always a place for a #2, and Avis did just fine being #2. They just figured out how to capitalize on it and turn it into an advantage.

>>Also trust and spam detection in general. That is where google comes into its own in comparison with other engines.

Except for back-link spam. The tricks for clicks crowd will always stay a step ahead on that one, by being very trend-savvy and agile.

FWIW, this way cheaper and less disruptive to MS than trying to buy Yahoo. If this semantic thing is any good, I give it 6 months before it goes live and Google has a panic attack.

The problem facing MS at this point is Google is a trusted brand, it's not just about who's best and which are the better relevant results, it's about something familiar and trusted. It will easily take 12 months for the buzz to shift a majority of people's browsing habits from Google to Live, assuming the semantic search makes it online in the first place.

If only they'd put that $100 million into OS development...

It's only one part of the mix. Microsoft have to find a way to monetize its offering - that's where the Yahoo deal would have come in.

This is a cost effective way of adding search technology, and I hope it's going to work for Microsoft. I'd rather have three strong competitors in the search market than have one way out and two some way behind.

I can almost always find what i want on google and it would have to taking something mega good to make me change.

Something mega good to make me change.

How much more do we need? I mean, there is only so much you can do with Search, isn't there? And Google have pretty much capitalized the market in that area. If anything mega good is coming to Search, I'm almost certain it will have Google's brand on it.

I was following all of Google's Acquisitions at one time. It just became a bit too tedious for me so I backed off. If there is "technology" involved, Google is involved. And, if any promising new technology hits the marketplace without Google's name on it, they are quick to buy it for some off the wall 7-8 digit number (not including decimals).

I'm interested to see whats in store for 2nd tier search as the 1st tier is locked up solid for the time being. It looks like it will be that way for the rest of this year and next at the current rate. Each month that passes by, Google grabs another 1.0% of market share. I believe 2008 June numbers are probably going to show Google at 70%+.

So Microsoft, Where's the Beef? Hey, there's an idea. Bring back the "Where's the Beef" lady and put Powerset in her hands. That would be the ultimate test. :)

How much more do we need? I mean, there is only so much you can do with Search, isn't there?

Next step as I see it is to get rid of that troublesome need to click on search results to see the content. First to manage that wins!

I don’t understand one thing – why people cant concentrate on something they are good at.

MS is Vista the last we are seeing from you?

A simple example why Google is top today is they concentrate on what they are best at – Search!

Microsoft's interest in semantic analysis isn't anything new:

[webmasterworld.com...]

It's the ingegration/implementation that's been the issue.

A simple example why Google is top today is they concentrate on what they are best at – Search!

Google bought Applied Semantics (CIRCA Technology).

God I'm getting sick of hearing the word semantic.

Aren't the search engines semantic enough already? And who said I wanted them to be overly semantic? When I search for something I want to find IT, not something maybe like it. If I put in a term that is too ambiguous that's my own fault, and I follow it up with a more defined version of the original result.

I think semantic is just a buzzword that means "Artificial Intelligence". I don't want my search engine to think for me. I just want it to deliver results for terms I put in it. Is that so much to ask?

If only they'd put that $100 million into OS development...

Apparently Vista cost $10 bln to develop. If you look at all Microsoft costs of OS development over the last 20 years you may find that $100 mln is peanuts.

>>When I search for something I want to find IT

What about if you're looking to buy a present for 'er indoors?

Interestingly I think the search battle is kind of moving away from quality and into branding. Google is good at search, everyone knows that. There may be better techniques out there, but they will need to be 10 times better to get folks to abandon the trusted brand they have come to love over the years. Google is a household name. It is no longer enough to have better search. It is now a battle of the better brand. And Microsoft is losing on every new product they release... Vista... Zune... Windows mobile... etc. Xbox had some success... but you don't hear about that any more...

Instead of merely searching for keywords on a page, as Google does...

Isn't this a huge assumption?

Powerset search results feature Adsense; have MS bought an MFA site for 100M?

Vista... Zune... Windows mobile... etc. Xbox had some success... but you don't hear about that any more...

They should have given that 100 million to Xbox owners that had to wait 3 months to get their console back after getting the red ring of death. For that many consoles to just up and die is ridiculous and truly another instance of Microsoft rushing to get their product to market before testing it enough.

Oh, and don't forget HD-DVD. That was a big failure as well.

Mod Note - Please, let's keep the discussion about the Powerset acquisition.

Back on topic:

This has to be a benefit to Microsoft's Live Search team.

If the price is correct, then it's not a bad deal, IMHO.

Once integrated, it should make a big difference to the existing search. Let's just hope it's easily scalable.

We really have to keep in mind that for decades, MS desktop software has been available and successfully implemented in multiple languages worldwide; so they're not exactly novices in Natural Language Processing. The big however, though, is that desktop software, as cutting edge as it must be, isn't functioning within an adversarial environment like organic search is.

I tried Powerset a while ago, on a single search, and it did better than Google. But that was on a Wiki page, now they must scale it to the web.

implemented in multiple languages worldwide; so they're not exactly novices in Natural Language Processing.

Those translations are typically done by hand by localisation engineers, not translation software or semantic processing or anything like that.

Semantic search is the beginning of machines understanding the intent of the searcher, not just hitting the database to see which documents rank best for the keywords entered.

I guess I must be missing something. I ran some tests using their search...then ran the same query using google for a wikipedia site search. The results were basically similar.

So that makes me wonder, what's so special about Powerset that it's supposed to be the "google killer?"

How much more do we need?

A simple example why Google is top today is they concentrate on what they are best at – Search!

A CSR lady that works for me told me yesterday that she has hard times finding anything on Google, that she now starts looking on page 3 and further to find stuff. Ditto for myslef, and I am an experienced surfer and can do advanced things. Google has destroyed non-commercial page 1 results, all you get is Wikipedia and some newssite these days, which is useless for the most part. The rest of the results are commercial.

And, coming up on Google "improved" results - their own affiliate program results on Page 1. Yu-hoo! I mean ... Yahoo!?... Altavista? It almost feels like we are back in the 90th, except Google has the $$$ to control their own destiny.

What Google is good at is finding spam. If MS or anyone else can even come anywhere close to that, they will get the market share, 'cause Google's SERPs ain't that good anymore. Don't know if any semantic soft is going to help MS though.

Re < I can almost always find what i want on google and it would have to taking something mega good to make me change>

Amen. How good (excellent, wonderful, etc. can good, excellent, wonderful get? Google is so damn good, it's hard to imagine something better, but.............possible, but it would have to be a whole lot better by a factor of 10 to impress most and switch most from a trusted brand search engine to something else.

I guess I must be missing something. I ran some tests using their search...then ran the same query using google for a wikipedia site search. The results were basically similar.

So that makes me wonder, what's so special about Powerset that it's supposed to be the "google killer?"

That is in and of itself is a major accomplishment if it can be scaled to the entire web.

I think the problem is finding quality results in a search query. There are tons of results in all the search engines , thousands of pages .

My search experience has come to the conclusion that majority of page one results are placing a nice adsense ad in the middle of their page to maximise revenue meanwhile if they maximise user experience i think they would make a whole bunch more money.

I also like the idea that the internet allows smaller business to get a slice of international business , this help entrepreneurship and enable people to grow their business.

I recently saw a talk by Jeff Bezos and he claimed the internet was a bit like the discovery of electricity. When first discovered it took thirty years to invent the plug. I think the internet hasnt even been around for thirty years yet so we still have loads more to come.

How this relates to search is that we still have far to go , however Google will be around for years , ms as well . I think there still is room for start up companies to take search to a new level.

Thats the whole thing about technology , its always changing and there will be new companies coming and going .

Good luck

>That is in and of itself is a major accomplishment if it can be
scaled to the entire web.

Yes, Google isn't a search engine company, although they have a search engine. They're the world's largest ISP, specializing in ultralarge datasets. (That's one of the things it takes to be able to scale a search engine to the web.) A potential competitor will have to compete in that space, just to get into the search engine competition. Not many companies could do that at all. Most companies that could, have better sense than to waste their time doing it just to compete with Google.