Semantic Search Retooled at Yahoo
June 25, 2010
Yahoo!’s Developer Network blog published a brief interview with Peter Mika, a wizard from Yahoo’s Research Division. The topic is semantic search, also the topic of a speech he delivered on the 19th, called “The future face of Search is Semantic for Facebook, Google, and Yahoo!”. But I wonder about the “growing interest” mentioned and how the short article frames semantic search as something that’s just now happening (“Semantic Search will also bring entirely new functionality.”) The future is already here; semantic search and developing metadata to use in search is fairly old hat. The article specifically mentions Sindice, a semantic search engine we noted in 2008. While I’m sure there’s good basic info on semantic search to be had, I’m more inclined to see this release as a pitch for Yahoo!’s SearchMonkey. In fact, how do these different semantic efforts fit together? Where’s the intersection with the Bing?
Jessica West Bratcher, July 25, 2010
Freebie
Yahoo Is Committed to Search—Cutbacks, That Is
June 9, 2010
Short honk: Lots of pre-Microsoft deal chatter about Yahoo search.I thought that the Yahooligan search wizards lacked experience getting bought.Lots of excitement when a big outfit cuts a big deal. Point your browser at “Another Round Of Layoffs At Yahoo, Search Team Gets Hit.” Simple story. Search experts are expendable. What’s that mean for search at Yahoo? Your guess is better than mine.
Stephen E Arnold, June 9, 2010
Freebie
Yahoo Bafflement
May 26, 2010
Yahoo fired up its purple rainstorm in 1995. As Internet ventures go, Yahoo is the example of what happens when a technology company goes consumer. The next phase of Yahoo’s evolution took place when the Google ad system débuted and Yahoo settled out of court regarding an alleged action by Google related to Yahoo’s ad system. At that point, the company, a technical Australopitecus afarensis, watched the emergence of a ripped Homo sapiens.
I read “Carol Bartz to Mike Arrignton” article and found it quite revealing and a source of useful information about corporate development. Mr. Arrington has a knack for getting high profile people to provide useful information. The venue was a “disrupt” conference and I expected fireworks. I was not disappointed.
For me, the key passage was this exchange between Mr. Arrington and Ms. Bartz:
MA: Are you a search company or not?
CB: Half of our revenue is from search. The fact that you can crawl the web is a commodity. We’re about search, but we’re not a search company. We do a lot of things.
“A lot of things” is the key point. In the beginning, Yahoo was a directory. In the intervening years, the company has become a lot of things. The trajectory in business school would be one of those nifty –ism’s: entrepreneurialism, opportunism, etc. In reality, Yahoo has become America Online, and we see the perils that that case example illustrates.
Is Yahoo a search company? No. It never was a search company. It was a directory company. Somehow the baloney about search has morphed into interesting and sometimes horrific forms that confuse and frighten me. The whole privacy, conversation, and collaboration thing has been the “buzz” (no pun intended for several years). The ad thing is not new, but technology permits monopolization in a way that would have made the marketing fellow for Pears’ Soap fall to his knees stupefied.
Yahoo is a company with traffic. That traffic is the point, and the actions the company is taking are dependent upon and related to traffic. In short, whether Yahoo cuts a deal with a flickering candle like Nokia or buys a content mill, Yahoo is doing what leverages its traffic.
So, forget the generalizations about search. Yahoo is following the trajectory of AOL which followed the trajectory of the Source before it. The recent announcements from Google suggest that its management team is packing the station wagon and getting the camping gear together to take a similar trip.
The notion of a commodity is an important one, but I don’t think a Web index is a commodity. The overall diffusion of the technology for what the phrase “the Internet” refers is seeping into other methods. The boundaries are exciting, but once the boundary has been passed, the new world is like a three day camping adventure—fun for a day and then agony until everyone heads back to home. The discomfort, work, and insects are not worth the “adventure”.
Yahoo, therefore, is not going to find it slow going as the company tries to make changes that really matter. From my vantage point, Yahoo is following its predecessors in the consumer online space.
Stephen E Arnold, May 26, 2010
Freebie
Web Search Share and Tea Leaves
May 13, 2010
I read BusinessInsider’s “April Search Numbers Out! Yahoo Gains Share Because Of New Slideshow Tool” and was more confused than before reading the summary of comScore’s league table data. I think that with each passing month, the analyses of what happens to Google, Microsoft, and Yahoo is less and less germane to my work.
Two quick examples.
The first item is from this passage: “On a reported basis, Google lost ~70 bps of search share in April vs. March, while Yahoo! was the main beneficiary, up ~80 bps.” I have no idea what this means, and I am not convinced that the metric is significant. Back to Stats 101 for method, margin of error, etc.
The second item is from this passage: “According to the data, total US core search volume increased 5.3% Y/Y in April, a deceleration from 7.6% growth in March. However, adjusting for the impact of user interface changes, we estimate that search volume was up ~3% Y/Y. The April growth level was also below 1Q’s 10.1% Y/Y growth.” My question, “What?”
When I look at the data, I don’t see much more than tea leaves. Over the last three or four years, my perception says, “Google is far ahead. Microsoft has made some modest gains. Yahoo remains a mystery because it is not in the search business. Other search vendors are marginal players.”
Feel free to disagree but these séances with comScore-like data echo the Ouija board trend. Knee shaking cannot move the dial, however.
Stephen E Arnold, May 13, 2010
Unsponsored post.
Yahoo and Facebook Envy?
May 5, 2010
The Financial Times doesn’t seem impressed with Yahoo’s moves to integrate Twitter and Facebook in “Yahoo Wants to Do What Facebook Did, Only Slower.” Trying to capitalize on the social search phenomenon, the search giant will soon add social capabilities to the address book, showing you when your Yahoo! friends “comment on Yahoo News stories…rate a song or movie, or otherwise interact with Yahoo content.” Are they too late? Or do they just want to grab some of the social search pie? The FT agrees that it makes sense to keep people engaged with the site longer, but points to a four-year quote from a Yahoo exec about social search’s “power by virtue of tapping untapped authority”. Maybe, secretly, they wish they had come up with a social network for college students.
Samuel Hartman, May 5, 2010
Note: Post not sponsored.
Ask and Yahoo Allegedly Gain Share
April 11, 2010
I delight in the league tables for the Web search championship. The data are made particularly compelling because little information about methodology, margin of error, and numerical recipes are provided. Anything goes for the azure chip crowd. Consider “Yahoo Searches Gain 3% of Market Share.” I am not sure what a “search” is nor am I able to explain “market” but never mind.
Yahoo grew, according to Hitwise, one of the PR centric online analytics companies. But even more surprising to me was the factoid that Ask.com grew by 21 percent, from 2.84% of the “market” to 3.44%”. The losers were Google and Microsoft Bing.
Several thoughts:
- With Ask.com on an apparent roll, why don’t I use the service? The results don’t allow me to do my work more quickly or easily. Ask.com strikes me as a service that may be helpful to a small number of folks looking for information but I think these folks are a distinct and interesting segment of the market. Ask.com I have learned by listening at parties where school teachers are in the mix is a hit with the middle school crowd. Some teachers find the results either “safe” or “more understandable” to this age group. I need more data, of course.
- What are Google and Microsoft Bing doing wrong? Once this question is considered, the answer is, “Not much that is within their control?” Both companies are on a marketing and PR blitz. Both companies are working overtime to improve their search offerings. Both companies are high profile brands. Maybe there is a “fatigue” factor? Maybe the Hitwise data need that “margin of error” thing that bedevils some first year statistics students in colleague?
- What are Ask and Yahoo doing right? I don’t have Ask on my radar because once the company became the search engine of NASCAR I dismissed it. Yahoo is retrenching and doing lots of MBA tricks to revivify an aging, unexciting brand. I don’t think Ask or Yahoo has made a substantive change that caught my attention, but I may be guilty of inattention.
- Is Ask getting some link love from IAC Web properties CollegeHumor.com, Reference.com, Vimeo.com, Chemistry.com, TheDailyBeast.com, InsiderPages.com, and other IAC sites? Backlinks are often helpful in making those traffic stats perk up.
To sum up, Web search is no longer the main event. The users who matter are shifting to services like Facebook to find information. Another segment of users who matter are going the mobile route. Despite the fancy mapping methods, once a person has a mobile device, the search experience becomes different from what one can do sitting at a notebook with a fast, stable network connection.
My thought is that these league tables are becoming less relevant to what I am tracking. Your mileage may vary but the era of the text search is ending. The winner of a somewhat unexciting way to get information has as much importance to me as the name of the gladiator scratched on the wall at Pompeii.
Stephen E Arnold, April 11, 2010
A freebie.
Quote to Note: The Foundation of Yahoo, Rotten
March 3, 2010
Navigate to “The Steady, Efficient Decline Of Yahoo.” The article has some useful analysis of Yahoo’s woes. These include silos and unexciting products. The key statement is:
Yahoo’s foundation is rotten.
When foundations are weak, the structure comes apart. Right?
Stephen E Arnold, March 2, 2010
No one paid me to write this. I will report non payment to the manager of the Blair House, which has foundations like Toyota plants.
Yahoo Redefines Search as an Interface
February 28, 2010
Yoiks! The UK media giant Telegraph ran the story “Yahoo and the Future of Search” with a remarkable subtitle: “Yahoo’s search deal with Microsoft could usher in a new purple patch for the former web giant.”
Look at this paragraph in the interview with the former Googler Yoelle Maarek, now senior director of Yahoo Research.”
The [Microsoft] deal, she explains, should be seen as liberating Yahoo! to focus on front-end search innovations, rather than spending time and money on ensuring the back-end technology is working well. If anything, says Maarek, the Microsoft deal has freed the company up to start fighting the search war in the most important area – the bit the consumer can see. Yahoo’s search teams are planning to launch several new initiatives in this area over the coming months, to try and steal share from search Goliath Google, and also, somewhat confusingly, from Bing, the search platform of Microsoft, its new partner.
Now I am an addled goose often besieged by azure chip consultants and swamped with poobahisms. But this passage startled me.
First, as a former Googler, I expected something logical. Search as interface is okay, but it is part of the story. In fact, that story is Microsoft’s pitch for the UX or user experience. The idea is that users can have eye candy, facets, and suggestions so no query is needed. Fine, but to define search as an interface is like defining GM as a Corvette wheel assembly. Important but not the “real” GM.
Second, I quite like the notion of relying on Microsoft’s plumbing so Yahoo can do the UX thing. The hitch in the git along is that the plumbing is not quite up to Google standards. Ms. Maarek may not know how different but I am of the opinion that she will find out pretty darned quick.
Third, Yahoo has to make stuff work. I will not mention Panama. I will not point out the problems with email search. I will not point out the interesting behavior of Yahoo’s help system. Nope. I will just suggest you navigate to Yahoo.com and run a query for a Canon 200 camcorder. Check out the results. Make your own decision about where Yahoo is relative to Amazon and Bing. Even Google’s quirky Products service is more useful in my opinion.
What I find remarkable is that a marketing pitch is presented as laser like insight:
Over the coming months, Maarek says Yahoo! is focusing on three core search areas. Firstly the company is investing in lots of research and “data crunching” to understand how its search engine can better anticipate a person’s “intent” when they enter a search term – for instance, how it can discern whether a person is looking for business news, the record label or the fruit if they enter the search term ‘apple’. Secondly there are new tools being built upon and promoted to make searching via Yahoo! easier and help the company “establish a dialogue” with its users, according to Maarek. ‘Search Pad’ is one of these initiatives. It is a note-taking application which automatically assists a user in saving the addresses of the websites they are visiting on a virtual pad. It helps users collect, edit, organize, save, print and email their notes for immediate or future use. However, unless a user is logged into a Yahoo! account, it will not save or send a user the URLs after the browser window is shut down. Thirdly, Yahoo! is trying to build upon is ‘web of things’ concept – the idea that the web should be seen as an entity built up of ‘objects’ rather than documents. This should be reflected in the way search results are presented. For instance, if a user searched for Lady Gaga, then instead of receiving a list of blue links, the search results will be presented like a mini newspaper – with a variety of types of result including, images, ticket offers, videos (presented as videos and not just links) and news articles.
Does this sound like marketing speak to you? It does to me.
My view:
- Yahoo is yesterday in my opinion and rolling up other companies’ technology is a tricky financial wicket in my experience
- The technical infrastructure has not been rationalized so cost control for Yahoo will remain a problem. You can only sell off so much and fire so many people before deterioration accelerates even if revenue ticks up a notch.
- AOL is moving in a new direction, not talking about a new direction. I think the odds against AOL are significant but, hey, AOL is giving the content farm game a whirl
- Time is running out for Yahoo. Top line revenue growth is needed now.
What happens if the Microsoft Yahoo tie up does not yield big bucks and significantly greater market share? Microsoft moves on and Yahoo twists in the wind.
Stephen E Arnold, February 28, 2010
Xerox Legal Eagles Swarm at Google and Yahoo
February 24, 2010
Quite a surprise. I have not given Xerox much thought. True, about 11 years ago we had a job to hook one of the DocuTech scanners to the main DocuTech copy machine. Not too tough, but work is work. Since that time, I don’t pay much attention to Xerox. I know about Xerox Parc’s history of innovation, of course. I do recall learning that the company has rolled out an information system for law firms, but I don’t think of Xerox as a document management or eDiscovery company. Xerox to me is a maker of photocopy machines which makes clear why the headline “Xerox Files Patent Suit against Google, Yahoo” caught my attention. I thought, “What?”
The main idea is that Xerox has US6778979, “System for Automatically Generating Queries”. I think I met Greg Grefenstette at one time. The invention, according to the patent document’s abstract states:
A system generates a query using an entity extractor, a categorizer, a query generator, and a short run aspect vector. The entity extractor identifies a set of entities in selected document content for searching information related thereto using an information retrieval system. The categorizer defines an organized classification of document content with each class in the organization of content having associated therewith a classification label that corresponds to a category of information in the information retrieval system. The categorizer assigns the selected document content a classification label from the organized classification of content. A query generator formulates a query that restricts a search at the information retrieval system to the category of information in the information retrieval system identified by the assigned classification label. The short length aspect vector generator generates terms for further refining the query using context information surrounding the set of entities in the selected document content.
Xerox also asserts that the Google infringed on US6236994, “Method and Apparatus for the Integration of Information and Knowledge.” This invention, according to that patent document’s abstract states:
The present invention is a method and apparatus for first integrating the operation of various independent software applications directed to the management of information within an enterprise. The system architecture is, however, an expandable architecture, with built-in knowledge integration features that facilitate the monitoring of information flow into, out of, and between the integrated information management applications so as to assimilate knowledge information and facilitate the control of such information. Also included are additional tools which, using the knowledge information enable the more efficient use of the knowledge within an enterprise, including the ability to develop a context for and visualization of such knowledge.
The TechWeb article reported:
Xerox is seeking treble damages because it claims the defending companies are aware of its patents and that their infringement is willful.
I know zero about the legal world. I do know big bucks when I read about this type of claim. What’s interesting is that Xerox seems happy to talk about the legal matter. According to the write up:
“We have been in dialog with Google and Yahoo for some time about licensing these patents, without reaching a resolution,” a Xerox spokesperson said in an e-mailed statement. “We believe we have no option but to file suit to properly protect our intellectual property.”
The economy may be struggling, but the lawyers involved in this may have a Veyron in the drive way by next spring.
Stephen E Arnold, February 24, 2010
No one paid me to write this. Unlike attorneys, I guess, I work without compensation. I have to report non payment to the USPTO. I hope that group’s online system someday includes more patent documents easily accessible via a search system that does not violate another party’s patent.
Microsoft and Yahoo, The Challenges
February 21, 2010
eWeek, once one of the big dogs in the Ziff Communications kennel, ran the story “Microsoft, Yahoo Face Integration Challenges, Analysts Say” on February 20, 2010. No kidding? I set this short write up aside because I was not sure how to comment on the analysis by the analysts. I decided to point out the challenges expressed in the article even those these were scattered and not grouped to make explicit that Microsoft and Yahoo have some challenges ahead. Here goes:
- Nine months to achieve integration, full shift by 2012
- Microsoft’s ad system is ready to tackle the Google in hand-to-hand combat
- Combined market share about 30 percent. Google’s market share is 65 percent of US market, maybe more so that’s like a handicap in golf, right?
- Yahoo’s hot search features will add lift to Bing. What about Bing’s UX?
My thought? We will know at the end of 2012 if not sooner. If this flops, what is Plan B? Lots of assumptions, lots of challenges. No Plan B. Even Alexander the Great had a Plan B until he fell ill and died.
Stephen E Arnold, February 22, 2010
Nope, no one paid me to write about Alexander the Great. Ah, a disease. I must report getting no dough for this short item to the NIH?
-
- Subscribe to Beyond Search
Feature archive
News archive
-
