Svizzer Update: Swiss Search
April 1, 2009
I looked at the Svizzer desktop search tool in 2006. Yesterday (March 30, 2009), one or my two or three readers in Europe asked me, “What’s become of Svizzer?” I had to admit that I had not thought about the Swiss company for two years. Svizzer was in 2006 a Microsoft desktop search system created, according to my yellowing notes, by G10 Software AG. The company crated “an information cock pit for everything”, I wrote. Categorical affirmatives trouble me, and the notion of a cock pit with “everything” strikes me as a quick way to confuse the pilot and probably contribute to increasing the chances of an error, particularly in a combat situation. Under fire, less is more usually works better than floods of data.
At one time, the company asserted, “Microsoft n’a pas de solutions adéquates en ce moment et il n’y a pas de solution convenable sur le marché en vue dan un futur proche.” This statement was true in 2006, and in my opinion, it is true as I write this on March 31, 2009. My notes identify Alexander Rossner, Peter Biewald, and Dieter Eschebach as key executives.,
The Svizzer interface from Version 3.5 in 2007.
The 2006 version of the software included a Copernic like metasearch feature. I noted that it could remove ads from the results, but since I have not tested the system in two years, I don’t know if today’s ad technology will be caught by the system. The last download link I had was for Softpedia here. Give it a try.
The idea, I noted, was that Svizzer provides “a single point of search access to site search, enterprise search, desktop search, Web log search, news search, and Web search”. The system, according to my notes, performed a Vivisimo like aggregation and deduplication function.
The license fees for the system begin at about $100 per user. A customer can buy an annual service plan that includes support, upgrades, etc. I noted that the software included an advertising function, but I was not sure how well that would work in some more traditional organizations. I thought that an administrator could use the ad function to put specific information in front of users, a version of hit boosting.
The company is privately held, according to my notes. There were about 15 employees. I estimated the firm’s revenues in the $2.0 million per year range.
The company was European centric, and I noted, had plans to expand within continental Europe. The company’s Web site at www.g10.ch is off line. The last active date I have in my files is June 2008.
I am not certain the company is still viable. The last Web log entry here is dated March 2007. If a reader has more information, please, use the comments section on this Web log to update the information.
Stephen Arnold, March 31, 2009
Cutting Information Technology Costs
April 1, 2009
I read CIO Magazine’s “Five Things You Need to Know: Budget Cutting” here and realized that there are journalists and their are more sophisticated financial types. Before you read the article, click here and learn how Qantas slashed its information technology costs.
Now you can romp through the tips and come away vulnerable to the machinations of a cost analyst. These tips provide a sense of false security which may give way to some even bigger organizational realignments in certain situations. One realignment may be suggesting the person who followed these tips find his or her future elsewhere. In short, there’s more to cost cutting that the tips suggest.
I don’t feel comfortable parroting the five tips. I want to highlight one and offer a few comments. Let’s look at item three:
Break down exactly how you spend your budget. Start by identifying your expenses. A CIO might know she has employees with BlackBerrys, but how much is each person spending on their phone bills? With more specifics, you can set more accurate goals for saving.
Sounds pretty safe, right? The problem is that analysis of a segment of a budget means that the numbers reflect the person’s understanding of costs. Most information technology managers are clueless about the secondary costs triggered by routine information technology actions.
Let’s look at one example. A marketing team must produce a proposal. The marketers named Trent and Wendy need to output the document in two forms: PDF and to an ftp server. The client wants to get ftp access on a Monday and receive the FedEx package with the six copies of the proposal no later than Tuesday, before 5 pm. The Trent and Wendy duo understand the PDF part, but neither knows how to move files to the organization’s ftp server. Trent calls the information technology department and asks for help. The IT person says, “We’ll be there before 2 pm.” The IT person does not arrive. At 5 pm, Wendy calls. The phone rings in space. Trent and Wendy call a consultant, explain the problem, agree to a fee for the ftp part of the job, and complete their work. The cost for the ad hoc consultant is lost in the organization’s budget. The IT manager and the CIO are clueless about the cost their unit created.
Sound familiar?
The problem with the recommendations is that the actions will not return a comprehensive picture. A true cost analysis will surface dependent costs and indirect costs, not just the obvious direct costs. Once tallied, these costs can be tracked back to the root cause of the over or under run.
In my experience, that’s too much work. As I learn about companies going out of business, the casual approach to cost analysis bites back and bites hard. Read the five tips. Track down a cost analyst and enlist his or her help. The present financial climate does not look kindly on those involved in projects such as search. It’s easy to blow through $500,000 in a matter of three or four months and have a non functioning system. Undisciplined thinking about enterprise systems and their costs can crush a career and an organization. The five tips omit that point.
Stephen Arnold, April 1, 2009
Passwords List
April 1, 2009
Short honk: you can get three lists of common passwords here. These lists often come in handy when filtering government information prior to putting documents online. ArnoldIT.com has used this method for years. If you are indexing an organization’s documents, you might want to filter test your corpus. Might be helpful.
Stephen Arnold, April 1, 2009
Search Certification
April 1, 2009
A happy quack to the reader who told me about the new AIIM search certification program. Now that will be an interesting development. AIIM is a group anchored in the original micrographics business. The organization has morphed over the years, and it now straddles a number of different disciplines. The transition has been slow and in some cases directed by various interest groups from the content management sector and consulting world. CMS experts have produced some major problems for indexing subsystems, and the CMS vendors themselves seem to generate more problems for licensees than their systems resolve. (Click here for one example.)
This is not an April’s Fool joke.
The notion of search certification is interesting for five reasons:
First, there is no widely accepted definition of search in general or enterprise search in particular. I have documented the shift in terminology used by vendors of information retrieval and content processing systems. You can see the lengths here to which some organizations go to avoid using the word “search”, which has been devalued and overburdened in the last three or four years. The issue of definitions becomes quite important, but I suppose in the quest for revenue, providing certification in a discipline without boundaries fulfills some folks’s ambitions for revenue and influence.
Second, the basic idea of search–that is, find information–has shifted from the old command line Boolean to a more trophy-generation approach. Today’s systems are smart, presumably because the users are either too busy to formulate a Boolean query or view the task as irrelevant in a Twitter-choked real time search world. The notion of “showing” information to users means that a fundamental change has taken place which moves search to the margins of this business intelligence or faceted approach to information.
Third, the Google “I’m feeling doubly lucky” invention US2006/0230350 I described last week at a conference in Houston, Texas, removes the need to point and click for information. The Google engineers responsible for “I’m feeling doubly lucky” remove the user from doing much more than using a mobile device. The system monitors and predicts. The information is just there. A certification program for this approach to search will be most interesting because at this time the knowledge to pull off “I’m feeling doubly lucky” resides at Google. If anyone certifies, I suppose it would be Google.
Fourth, search is getting ready to celebrate its 40th birthday if one uses Dr. Salton’s seminal papers as the “official” starting point for search. SQL queries, Codd style, preceded Dr. Salton’s work with text, however. But after 40 years certification seems to be coming a bit late in the game. I can understand certification for a specific vendor’s search system–for example, SharePoint–but I think the notion of tackling a broader swath of this fluid, boundaryless space is logically uncomfortable for me. Others may feel more comfortable with this approach whose time apparently has come.
Finally, search is becoming a commodity, finding itself embedded and reshaped into other enterprise applications. Just as the “I’m feeling doubly lucky” approach shifts the burden of search from the user to the Google infrastructure, these embedded functions create a different problem in navigating and manipulating dataspace.
I applaud the association and its content management advisors for tackling search certification. My thought is that this may be an overly simplistic solution to a problem that has shifted away from the practical into the realm of the improbable.
There is a crisis in search. Certification won’t help too much in my opinion. Other skills are needed and these cannot be imparted in a boot camp or a single seminar. Martin White and I spent almost a year distilling our decades of information retrieval experience into our Successful Enterprise Search Management.
The longest journey begins with a single step. Looks like one step is about to be taken–four decades late. Just my opinion, of course. The question now becomes, “Why has no search certification process been successful in this time interval?” and “Why isn’t there a search professional association?” Any thoughts?
Stephen Arnold, March 31, 2009
-
- Subscribe to Beyond Search
Feature archive
News archive
-
