Kosmix and Its Positioning
June 26, 2009
I have been tracking Kosmix for two reasons. Its approach is in tune with certain demographic’s interest in having everything in one display, which is hard on the addled goose’s aging eyes. Second, the founder may be a fellow traveler with a Googler for a very interesting approach to information. Manifold opportunities perhaps? The two fellows live and work not too far from one another. And keep in mind that one person is a super wizard at the Google. Kosmix, therefore, is on my radar.
The company is on the radar of the San Jose Mercury News too. “Kosmix Tries New Twist on Search to Avoid Google’s Dominance” by Scott Duke Harris explained Kosmix’ market positioning. Mr. Harris wrote:
But Kosmix founders Venky Harinarayan and Anand Rajaraman, say they know better than to take on Google or anyone else delivering those familiar lists of blue links. As the Web constantly adds information and images — and as smart-phones and social networks like Twitter and Facebook add new dimensions to search — Kosmix, based in Mountain View, is getting attention for its efforts to differentiate itself as a better way to navigate the growing online clutter. To look up, say, wonton soup on the search giants leads the user to a list of blue links, some with photos. Kosmix delivers a multimedia showcase. The page is topped by a Wikipedia summary, and a quick scroll leads to a sizable window for how-to videos, blog commentary and conversations, nutritional data and more. A column on the right includes items like Chinese cookbooks on eBay and 24-pack instant wonton shiitake soup from Amazon. There are also a few “sponsored links” — the advertising that is key to the Kosmix business model.
Kosmix has implemented in its service a number of the features and functions touched upon in various Google patent applications and technical papers. My hunch is that there is more than friendly competition between the Google and Kosmix. How will this dynamic duo party down in the months ahead? Good question.
Stephen Arnold, June 27, 2009
Two Countries Squabble — China and Google
June 26, 2009
In Google Version 2.0 (2007) I spelled out an argument that has been ignored. Now my analysis of Google’s options in China are playing out in real life. I asserted that Google had become a new type of company. In effect, it is operating as a supra national enterprise, which is tough to understand (particularly by regulators) and even tougher to define in terms of existing laws, guidelines and regulations. Few stop to consider that Google is not “anywhere”. To make a conceptual idea more clear, Google could plop floating barges outside a three mile limit and leave it up to its lawyers to explain where a particular computer function took place. You can follow the rest of my analysis if you snag a copy of Google Version 2.0.
Now a series of news stories highlights the squabble between China and a single commercial enterprise. The BBC’s story “Google Access Disrupted in China” makes it clear that Google is being given the digital equivalent of a shot across its bow. In addition to a service disruption, others are reporting that Google is distributing objectionable information.
Now Google has to figure out how to respond to the type of challenge that strikes me as like those issued at the outset of the 100 Years War. Google is not being handled like a mere company. Google is getting the equivalent of nation state treatment by China.
What is at stake? For starters, control of information. Another issue is the shaping of that information. In theory information is neutral. The reality is that information has mass, can be a catalyst, and behaves in weird uranium like ways. Will Google find a diplomatic solution to this problem, or will the company find itself engaged in a long cyberwar. Quite interesting to the addled goose I must say. The Google Microsoft battle pales in comparison with this first real test of a supra national digital entity against a modernizing nation state. How will Google respond? What steps will China take to make Google even more aware of the consequences of its actions within political boundaries? Will China attack Google in interesting ways? How will cultural factions line up? Exciting and new in my opinion.
Stephen Arnold, June 26, 2009
Microsoft Search Infrastructure
June 25, 2009
Short honk: Microsoft has hired some Yahoo wizards and now information popped into my RSS reader about Autopilot. You can read Codename Windows’ write up about this program and ask yourself, “How long will it take to complete this project?” and “What is its cost?” and “How much time will be needed to catch up, then pass Google in this engineering sector?” I don’t have answers but Google did plumbing then search. Microsoft seems to be approaching the problem the other way around if this information is on target. Just my opinion.
Stephen Arnold, June 25, 2009
Seeing Crime the Professional Way
June 25, 2009
Most visualization is gratuitous. Where graphic representations are useful is in law enforcement. Crimes are committed by people. People have to be someplace. Putting the data, the people, and the events together makes it possible to “see” patterns. Visualization of crime related information is often a short cut to resource deployment, anticipatory planning, and budget management. You can see some open source, no cost examples in the article “20 Visualizations to Understand Crime”.
Flowing Data did a good job on this write up. Recommended for those who wonder about the value of monitoring. Adding real time data to these visualizations is a very useful innovation that some organizations are now exploring.
Stephen Arnold, June 25, 2009
Rhode Island and Rooster Pricing
June 24, 2009
When a lad in Illinois, I recall visiting one of my relative’s farm. I learned how to kill a chicken. Learned what tough bird meant. Rooster pricing one of the farming Arnolds told me was that a good bird would fetch a pretty penny. The problem was that once a farm had a rooster, two roosters would be a problem. So roosters had a chance of not being worth too much.
I thought about my early rooster pricing lesson when I read “Why A New (And Unusual) Pricing Strategy By A Rhode Island Paper Will Fail.” This quite interesting article explains that a newspaper reader in Rhode Island (lots of roosters at one time) would charge a premium to get the newspaper in electronic form. Paper only was cheapest. Paper and online slightly more expensive. The online newspaper costs about $350 per year.
PaidContent.org’s write up said the idea will fail.
My view is that it might work when someone really needs only that digital version of the Newport (R.I.) Daily News. I don’t agree. The problem is rooster pricing. I think a few people who want only the digital edition will pay. In my opinion, the number of buyers will be as rare as hen’s teeth. A couple of sales won’t pay the bills. In my opinion, bad idea that rooster pricing. This article inspired me to collect my nine mysteries of online essays in one PDF and make the set available without charge. I will announce the download link in this Web log. No strings attached. No registration silliness. Some of that information will offer alternative pricing ideas. Sorry, no rooster option included.
Stephen Arnold, July 24, 2009
Text Mining and Predicting Doom
June 23, 2009
The New Scientist does not cover the information retrieval sector. Occasionally the publication runs an article like “Email Patterns Can Predict Impending Doom” which gets into a content processing issue. I quite liked the confluence of three buzz words in today’s ever thrilling milieu: “predict”, “email”, and “doom”. What’s the New Scientist’s angle? The answer is that as tension within an organization increases, communication patterns in email can be discerned via text mining. The article hints that analysis of email is tough with privacy a concern. The article offers a suggestive reference to an email project at Yahoo, but provided few details. With monitoring of real time data flows available to anyone with an Internet connection, message patterns seem to be quite useful to those lucky enough to have the tools need to ferret out the nuggets. Nothing about fuzzification of data, however. Nothing about which vendors are leaders in the space except for the Yahoo and Enron comments. I think there is more to be said on this topic.
Stephen Arnold, June 23, 2009
A Google Vulnerability Exposed
June 22, 2009
Erick Schonfeld’s “When It Comes to Search Trends, Google Is Lagging Behind Bing” identifies a potential Google weakness. I think TechCrunch is on to something, but I think the visible vulnerability explained by Mr. Schonfeld is a symptom of a deeper problem.
The weakness is an ability to handle what’s new and what’s happening. Mr. Schonfeld, wrote:
As Microsoft tries to take away market share from Google with its new search engine, Bing
, it is battling Google feature by feature. One feature where Microsoft seems to be edging out Google is with displaying recent search trends. This may not be a major feature, but it shows a weakness in Google’s armor.
Mr. Schonfeld presented sample queries that illustrate this issue. The bottom-line is that for the most recent information, I may want to use more than Google. Bing.com is one option and there are the numerous real time search systems available.
My take on this is different. Keep in mind that I think Mr. Schonfeld has identified a symptom, the deeper disease is “time deficiency.” As zippy as the Google system is when responding to queries, the Google is not as fast on the intake and indexing of real time data flows such as those from social networks.
My research has identified several reasons:
- Google’s attention is on its leapfrog technologies such as Google Fusion and Google Wave. Both of these are manifestations of a larger Google play. While the wizards focused on these innovations, the real time content explosion took place, leaving Google without a here-and-now response
- Google is big and it is suffering from the same administrative friction that plagued IBM when Microsoft pulled off the disc operating system coup and that hobbled Microsoft when Google zoomed into Web search. Now the Google finds itself aware of Facebook, Twitter, and similar services yet without a here-and-now response. Slow out of the blocks may mean losing the race.
- Google’s plumbing is not connected to the real time streams from social and RSS services. Sure, there is some information, but it is simply not as fresh as what I can find on Scoopler and some other services.
What we have is a happy circumstance. If Microsoft can exploit that weakness, I think it has a chance to capture traffic in the real time sector. But having identified a weakness does not mean that hemlock can be poured into Googzilla’s ear.
There are some other weaknesses at the Google as well. I will be talking about one at the NFAIS conference on Friday, June 26, 2009. Get too many weaknesses, and these nicks start to hurt. Addled geese have to be very careful but big companies are often too big and tough to be worried about a few nicks. If there are a thousand of them, well, the big outfit might notice.
Stephen Arnold, June 22, 2009
Twitter Tools
June 22, 2009
Now that outfits like the New York Times and CNN have concluded that Twitter is useful when reporting certain events, the Social Media Guide’s round up of Twitter tools may find some use in the newsroom. The round up “The Ultimate List of Twitter Tools” is long, grouped, and quite good. Highly recommended for dinosaurs and new forms of sentient information life. A reminder: there are other sources of real time info as well. Keep those options open, the addled goose honks.
Stephen Arnold, June 22, 2009
Library Teaches Search – More Instruction Needed
June 22, 2009
My recollection is that libraries taught search as far back at 1980. I recall that either database vendors would run demonstrations or that librarians skilled in the use of online would provide guidance to those who asked. I recall running a class in ABI/INFORM at Chicago Public Library and there was an overflow crowd of both staff and research minded patrons. I was delighted, therefore, to see an article in the Sacramento Bee that described the Sutter Library’s classes in finding health and medical information online. The class is a reminder to me that:
- Librarians and information professionals often know how to search and have an interest in sharing that knowledge
- Patrons are smart enough to know that despite the marketing hype and the pundits’ assertions that search is a “done deal” additional instruction attracts people and finds its way into The Sacramento Bee
We have a long way to go before information professionals will be relics of a long gone time. The people who tell me that they “know how to search” and “can locate almost anything online” are kidding themselves. I think I am a reasonably good researcher. But if you spend time monitoring how I find information, you will learn quickly that I turn to experts who make my search skills look primitive. Even my nifty Overflight system pales with the type of information that my research team generates by:
- Knowing what content is located where
- Understanding the editorial method behind or absent from certain online systems
- Leveraging hard-to-manipulate resources such as information from government repositories, specialized services, and individual experts.
I would like to see more libraries move aggressively into online instruction, market those programs, and raise the level of expertise. Most of the people who claim to be experts at search are clueless about how bad their skills are. Among the worst offenders are self appointed search experts who have trouble figuring out when something is likely to be baloney and when something is just plain wrong. Enterprise search, content management, and text mining are three disciplines where better research will be most beneficial in my opinion. Then we need critical thinking skills. Schools have dropped the ball. Maybe libraries can help in this area as well? Search procurement teams will be well served if the team has one or more librarians in the huddle.
Stephen Arnold, June 22, 2009
Associated Press and Facebook
June 22, 2009
Ryan Tate’s “AP Tells Reporters to Muzzle Facebook Friends” may be an indication that the Bozeman Syndrome is spreading. Bozeman, as you may recall wanted job applicants to provide user names and passwords for social networking site. I have a short item about this that will run in the future. Bozeman’s officials changed their minds, but the idea, in my opinion, may have meme power. (For information on the Bozeman request KTVQ, a Montana news outlet, offered up “Bozeman City Job Requirement Raises Privacy Concerns”.)
Mr. Tate’s story is not directly connected with Montana. But I perceive thin shoots of reaching toward the idea that an employee may not have tight control over social network memberships or participation. Mr. Tate wrote:
Someone sent us the Associated Press‘ guidelines for staff social networking and, in keeping with company tradition, they’re on the paranoid side. You should probably read them, since basically everyone in the world must now follow them. The AP’s Facebook and Twitter policies are less draconian than, say, Bloomberg’s, but that’s not saying much. They do sound, on the whole, reasonable, until you stop and ponder a few of the specifics.
My hunch is that this idea will find a quick uptake because some senior managers see social networking participation as falling within their span of interest. One of the goslings manages my social network participation, and I wonder if such control is possible. I delegate it, and I can think of several work arounds. When I squeeze a tube of super glue too firmly, the substance gets out of the tube. The metaphor may apply to social network controls.
The information can be “out there” and real time search tools allows me to find it. What’s missing is the “real” name of the person or software pumping out the content. So, what’s more important: finding the person who sent out the info or the info itself? Maybe both. Will stopping one ensure stopping the other?
Stephen Arnold, June 22, 2009
-
- Subscribe to Beyond Search
Feature archive
News archive
-
