When Google itself comes out with predictions for where search is headed, things get even more interesting. This is obviously because Google is such a huge and critical part of the search landscape. Google’s Matt Cutts discussed some of his own predictions for search in a recent upload to Google’s Webmaster Central YouTube channel.

One thing Matt stressed is that Google is always looking for new types of data to search. He gave examples of searching email with Gmail, books with Google Book Search, and patents with Google Patent search. He predicts Google will continue this trend and find more data sources to provide search functionality for.

Another prediction he gave was that Google will continue to improve search over harder problems. Specifically, he noted things like determining what is really going on with the words in documents and in queries – semantic search if you will.

"A lot of people think that if you type in ‘A B C,’ all Google does is crawl the web and return pages that match ‘A,’ ‘B,’ and ‘C’. And that’s not it," says Cutts. "We do a lot of sophisticated stuff. Think about synonyms, morphology…all sorts of ways where we can kind of find out, ‘oh, this is really related to them conceptually.’ Whether you want to call it semantic stuff or statistical processing, we do a lot of stuff to try and return relevant documents."

As part of this prediction, Cutts says Google will continue trying to find new ways to extract "good data" from the web. He mentions Google Squared (which is still in an experimental stage)  as an example of doing so. Google Squared, in Google’s words, takes a category and creates a starter ‘square’ of information, automatically fetching and organizing facts from across the web.

Cutts also predicts that people will get more comfortable with storing their data in the cloud. He expects more people will migrate their data from their hard drives to different cloud services, and that this will make it easier and better for search, and contribute to the delivery of more relevant results.

He also mentions real-time and mobile as playing significant roles in the future or search. No surprise there.

"It’s going to be a lot of fun. Search is nowhere near done, and every time we make search better, people ask us harder and harder questions, " he says. "So the nice thing is knowing that we’ll pretty much always have more to do to make search better."

Cutts recently discussed the possibility that page speed could play a role in search engine rankings. He made no mention of this in this set of predictions, but that is another thing to consider as we get ready to move into 2010.

What predictions do you have for search? Any that Matt didn’t go into? Discuss here.

Related Articles:

> Google Testing a Revamp of the Search Results Page

> Google Puts Its Own Dictionary in the Spotlight

> Critical Local Search Factors To Pay Attention To

There is clearly great emphasis placed on Wikipedia results, and I have to admit that as I began to read Microsoft’s explanation, I was starting to think, "Why not just cut out the middle man and search Wikipedia?"

The company says things like:

"The easiest way to find the Reference vertical within Bing is to look for the Wikipedia result in Bing’s algorithmic listings. Each one of those results has an “Enhanced View” link (see below) that will take you to the that entity’s Wikipedia page, within the Bing Reference vertical."

But getting further into the explanation, you start to see some more useful things going on. For example, if you type a question as your query, you may get Wikipedia results (which is fine)in some cases, but Bing will actually go so far as to highlig actual answers within result snippets.

They give the examples of "When did earthquakes hit San Francisco" and "who did Texaco" require. Here is what Bing delivers:

Notice in the Texaco example that it highlights answers about who Texaco acquired, and not who acquired Texaco. Furthermore, it determines these answers with the use of different words besides "acquired". You get answers highlighted that say things like "purchased," "bought," "acquisition of," and "took over the operations."

Powerset is behind these semantic results, and Microsoft hopes that these kind of results will sway users to start using Bing for their reference searches. Read the rest of Microsoft’s explanation to see the ways Bing handles Wikipedia content.

"Delivering the best answers though innovations in semantic search technology is the direction in which the search industry is headed, and Ask is best-placed to lead the industry in this regard given our database of hundreds of millions of questions, and our core search technologies," said Scott Kim, Ask.com’s EVP of Technology. 

"With countless answers on the web, consumers face challenges in getting the answer they’re looking for due to varied formats, redundancies, and junk answers. The Ask Q&A channel aims to cut through the clutter – and with today’s advancement, we’re well on our way," he added.

Last fall, Ask launched its own core search technology, and it has achieved the 300 million Q&A pairs through crawling and indexing questions and answers from "numerous and diversified sources" across the web. From there, Ask says it applied its semantic search technology advancements in clustering, rephrasing, and answer relevance to filter out insignificant and less meaningful answer formats.

"The result is a Q&A database that is fine-tuned to give consumers the best answer, the first time, every time through streamlined, localized, concise results to their questions," explains Ask. "For example, the question ‘How do I train my dog?’ yields in-depth answers with step-by-step instructions from a series of diverse and authoritative sources across the web – giving Ask.com searchers a selection of options and resources they can use to take action."

It should be noted that Ask’s Q&A results are also blended into its regular search results. The database can be searched specifically by using the "lots of answers" link on Ask’s homepage.

Ask.com is used three times more than any other search engine for search queries that are in the form of a question, according to research from comScore. Ask says its own research indicates that user frequency, retention, and loyalty are significantly higher within the Ask.com Q&A channel specifically.

Ask has now announced some advances in its semantic search technology. If you’re unfamiliar with the concept of Semantic Search, Wikipedia explains it:

Semantic Search attempts to augment and improve the search process by leveraging XML and RDF data from semantic networks to disambiguate semantic search queries and web text in order to increase relevancy of results.

Tomasz Imielinski, EVP of Global Search and Answers explains Ask’s changes on the company blog:

In October last year we introduced our proprietary DADS(SM) (Direct Answers from Databases), DAFS(SM) (Direct Answers from Search), and AnswerFarm(SM) technologies, which are breaking new ground in the areas of semantic, web text, and answer farm search technologies. Specifically, the increasing availability of structured data in the form of databases and XML feeds has fueled advances in our proprietary DADS technology.  With DADS, we no longer rely on text-matching simple keywords, but rather we parse users’ queries and then we form database queries, which return answers from the structured data in real time.  Front and center. Our aspiration is to instantly deliver the correct answer no matter how you phrased your query.

Many of the categories consumers care about most are rich with structured data – meaning almost anything you’d want to know about that category exists in a database or XML feed somewhere. Extracting it for practical use is another matter.

Imielinski uses the example of TV Listings to illustrate what he’s talking about. Queries like " Football on TV this weekend " and " Movies on TV now " illustrate how Ask’s search engine delivers these semantic results:

As Peter Young at Marketing Pilgrim puts it, "Whilst integration of this technology is currently limited, it certainly shows a level of innovation sometimes lacking from some of Ask’s competitors. On a personal level I am not sure this is going to lift Ask above competitors such as Google, Yahoo and MSN in the short term, however it is certainly indicative of an ongoing commitment to search, which is often in doubt."

If you ask me, they should have never dropped the "Jeeves," a brand that many people came to recognize. Imielinski promises more improvements on the semantic search front in the future.

(Coverage of SES Chicago continues at WebProNews Videos.  Keep an eye on WebProNews for more notes and videos from the event this week.)

 Nagaraju Bandaru

Nagaraju Bandaru, the cofounder and CTO of BooRah, kicked things off by pointing out that semantic search is already increasing the relevance of search results; both search engines and users are embracing it.  His company uses metadata to make improvements.

Kathleen Dahlgren, the founder and CTO of Cognition Technologies, then stepped in to mention specific areas targeted for improvement including tokenization, morphology, parsing, and formal semantics.  "We want to go into the natural language processing," she said.  "We want to take search technology down to word sense."

This means search engines would be more easily able to distinguish between, say, a strike on the head, a workers’ strike, striking a match, and striking out.

Tim Musgrove, the founder and CEO of TextDigger, believes that using natural tags to denote similar concepts will lead to real improvements.  Yet he allowed, "Semantic search will change our lives, but it will not happen overnight."

Scott Prevost, Microsoft’s principal development manager of Live Search/Powerset, agreed, "Search has a long way to go."  He feels keyword techniques involve only a shallow representation of documents’ meaning and users’ intent.

Larry Cornett, Yahoo’s vice president of consumer products, had at least a few good things to say about the direction SearchMonkey is taking, though.  Experimental enhancements have it applying structured data and rich content and giving users and publishers control.

Here are a few of the sessions I’m planning on covering:

More Customers, Fewer Costs – Why Marketing to the ‘Long Tail’ Makes Sense
Who doesn’t want more customers and fewer costs? Isn’t that a business’ dream? This session will be interesting as there have been recent stories on how the long tail is not as effective as we once thought. I still believe in the long tail though.

Semantic Search: How Will It Change Our Lives?
What? Change my life? That’s a pretty big promise. However, that’s what search does. It works to give the best results, which continually improve based on users needs. Just think back over the past few years; there has been blog search, universal search and local search all up and coming. Sounds like semantic search is next. Wouldn’t you agree?

Storyteller Marketing: How the Art of Storytelling Matches Up With the Business of Marketing
Ahh, story time. Everyone enjoys a good story, but where does that fall into play with SEO and online marketing? I can tell you a story about most anything, but throw in some package testing and spigot lingo and it’s a bit more difficult.

7 Proven Ways To Get Your Website On Page 1 Organically Then Convert
If it was only that easy right? I’m a bit skeptical here. If there are 7 proven ways, then as soon as they tell us, everyone will use them and, well, you can only have 10 on page 1. I hope this isn’t to high level and just about content, links and the stuff we already know.

Identify, Analyze, Act: SEM by the Numbers
Analytics is huge to a site’s success. How your visitors get to, and interact with, your site is extremely important and the stats hold all the secrets. But finding all the bits of gold that is buried in analytics can be tough to find. Hopefully they’ll show some secrets and insight into the best places to be looking. A little sad though that analytics based on Star Trek doesn’t appear on the schedule.

Google Dance
After a few refreshments Thomas will also be getting his groove on at the Google dance; even if he doesn’t know how to groove well.

Getting Vertical Search Right
Yes, it’s the world wide web, but you don’t have to compete with everyone. Being the best in your vertical is all a site needs to do. But how does one target some, but not everyone?

SEO Rehab & Intervention
Let me grab by celebrity attitude and I’ll be ready to go. I know none of my site’s have issues, but I’ll be sure to find issues in your sites.

Advanced Paid Search Techniques
PayPerClick is like the black sheep in SEO. It’s not organic or social, but it is a proven method for generating traffic and conversions. These paid links are OK! Just be careful not to blow your entire marketing budget in a few days or get a bunch of unqualified traffic. There has to be a happy medium.

SearchBash
More groovin and movin and networkin’ with ya’ll!

With all those sessions, the California palm trees and happy cows may not get enough attention. However, with all the knowledge poring in and great networking, there will be more than enough to keep busy.

If you’re going to SES San Jose next month, what sessions are you planning on attending? If you’re not sure, check out the full conference schedule.

Comments

While I’ve always wondered why you’d want to do that unless a search engine’s results were lacking, I figured it was about time to see how that idea was going after almost three months.

Pluses:

• The network doesn’t require you to register, just type in the query.
• You can contact the authors of the messages (if they’ve chosen that option), report abuse, remove your own messages and rate messages on a scale of 1 to 5 stars.

Minuses:

• Limited number of queries with responses.
• Inability to respond to other messages on the message board.
• No dates on messages—no idea whether the person you’re responding to even remembers this listing.
• Extremely unhelpful responses.
• Few, if any, understand what the utility is there for.

For one query which Hakia listed among its popular rooms, [care health] (no kidding), there were fifteen responses. Not bad, right?

Well, it’s not so good, either, when 9 of 15 listings are job listings, two of which are completely unrelated to the query (yeah, I was looking for work-at-home opportunities under [care health], how DID you know?). The other six include people looking for doctors (no indication of where they’re located), random statements, and plagiarized news items.

Am I surprised? No. I still don’t understand why you would want to connect with other people using the same query, especially one as generic as [care health] (which, incidentally, is the same room as [health care]). A query that generic will give you fifteen different “right” answers from ten different people.

Judging by the responses I’ve seen, very few people have quite grasped the concept of why this social network is here, mostly preferring to use it for self-promotion, spam and the online equivalent of graffiti. It could be marginally useful, but not in its present state.

What do you think Hakia could do to improve this utility?

Comments

Tag:

“Xerox Corp. says its new search engine based on semantics will analyze the meaning behind questions and documents to help researchers find information more quickly,” reports the AP’s Stephen Singer.  “FactSpotter promises to help by returning a specific portion of a search document that is relevant to the query.”

Singer then relays an example of what semantic search is all about.  “For example, common searches using keywords ‘Lincoln’ and ‘vice president’ likely won’t reveal President Abraham Lincoln’s first vice president.  A semantic search should yield the answer: Hannibal Hamlin.”

But here’s where Xerox may have a problem, as, following a Google search for “Lincoln vice president,” I can see Hamlin’s name in the second, sixth, ninth, and tenth results – without even bothering to click or anything.  Whatever Xerox can do, Google may be able to do better.

Then there are the companies that only do semantic search.  Powerset’s engine hasn’t yet been released, but a lot of people have been impressed by some early previews.

Still, Xerox’s product isn’t yet “out there,” either, so it’s too early to judge how things will go.  “FactSpotter was introduced in Grenoble on Wednesday and will launch next year,” states Singer, “initially to help lawyers and corporate litigation departments plow through thousands of pages of legal documents.  Xerox expects the technology to eventually be used in health care, manufacturing and financial services.”