Guest Author Stephen Arnold

One of my business associates in Canada sent me a link to an interesting search engine named Cluuz.com. The system–unlike the shy Powerset, a media darling developing a semantic search engine–is available for anyone to use. Navigate to Cluuz.com. Make sure you add the extra “u”, or you will be looking at a plain text page from the graphically restrained Clue Computing operation in cow country.

Cluuz.com takes results and applies semantic processes to them. Some of the company’s display options are a bit too sophisticated for my 64-year-young eyes, but I found the system quite useful. Let’s run through a basic search and take a cursory look at some of the features that I found interesting. Then I want to comment on the semantic search boom or boomlet (depending on how jaded you are), and conclude with several observations. In the last few days, the shrinking violets in the Big Name search vendors’ public relations department have reduced their flow of 30-something insights. Perhaps my comments about semantic search will “goose” them into squawking. I certainly hope so. Life’s no fun in rural Kentucky without well-groomed Ivy League wizards asserting their intellectual superiority in email speak.

A Query for Cluuz.com

Navigate to the Cluuz.com splash screen. Make certain that you have checked the option under the search box for “Charts”. We’ll look at the other options in a moment. Now enter the test query as shown in italics: Google +”programmable search engine”. Here’s my result for this query on May 7, 2008:

The system processes results from MSN (search.live.com) and Yahoo, processes them, and displays this map. Note that the system identifies important people and companies. The system correctly identifies the Google Forms service as related to the “programmable search engine”.

The system offers other ways to view the results set. For example, you can look at hits from the search engines to which the query is passed as a traditional laundry list. Other choices include a cluster display and a Flash display which is, in my opinion, cluttered with sliders, controls, and options.

You can also enter a more complex query using the Cluuz.com advanced search page. In my tests, the system did a good job of dealing with specific Boolean queries. You can also set preferences, which may not be necessary for a metasearch-based approach to generating hits.

Cluuz.com’ s Owner

Cluuz.com is a service of Sprylogics International Corp. This Canadian company develops advanced search, analysis, and information display tools and services. The company asserts,

These solutions enable users to search large amounts of unstructured data on the Web and in internal corporate databases, and convert it into actionable intelligence. The core technologies driving Sprylogics’ solutions are embedded in the Analyst and Evidens analytical workflow and Cluuz Search Engine platform which enables both consumers and corporate users to methodically search the Internet and internal corporate resources and find the information they are looking for. Cluuz search results are visually displayed through patent pending semantic graphs and result in improved decision making capabilities.

The company, like Kroll (a unit of Marsh & McLennan and Silobreaker) has a management team with some experience in law enforcement and intelligence. The firm’s president, Avi Shachar, has experience with the Royal Canadian Mounted Police and Israeli intelligence.

This real world experience is evident in the result set shown in the screen shot above. The default results for the query shown in the graphic in this essay focuses on people. Result lists are useful in certain types of library work. But an investigator wants to know who the key people are and how one or more of the individuals relate to one another, companies, and particular subjects.

Observations

My experience with Cluuz.com was quite positive. I see and test a large number of search and content processing systems, and Sprylogics is worth monitoring. Several other comments are warranted by my test drive:

1. For many years, getting a results list that identified people required expensive software and a lot of fiddling around. The established vendors such as i2 Ltd. in England made products that were not generally available for business intelligence. Sprylogics, it seems to me, wants to bring this type of useful service to a broader audience.
2. If I were able to push additional information into the Sprylogics’ Cluuz.com engine, I would be able to get a useful result without much, if any, manual opening and closing of links in a result list. The bird’s-eye view of the information pertinent to a query about a little-known topic like Google’s PSE is quite useful.
3. The interface permits customization. This is good because some of the fine-grained controls available in the Cluuz.com service as a Flash interface can be easily stripped down for “line of fire” work.

My two complaints are relatively minor but important to me. First, I could not locate pricing information about the company’s products or services. While not unusual, the lack of a floor cost means a phone call and email plus playing the “he’s on the road” game. Second, it’s difficult to determine how easy or hard it would be to integrate this system into a behind-the-firewall intelligence operation. I have a hunch the company has the details, but it has chosen to keep some data close to the vest. Again, this makes the management team happy but it bedevils me.

I have added Sprylogics to my watch list. Take a look, and I will update this posting as I gather more information about the company and gain more experience with the system.

Powerset Launches Wikipedia Search

Semantic search engine Powerset, which we’ve written about here before, has just launched its initial release. The current release is limited to indexing Wikipedia content, but it provides a great showcase for their technology and user experience.

For example, my search for “Alexander the Great” provided the following results page:

Here’s the interesting part: If you look at the corresponding article in Wikipedia, you see a long text article, with biographical details in a table on the right. The Powerset results page adds to this by pulling out a great deal more data (listed as “Factz” in the results page) – such as territories conquered, cities founded and armies defeated. The Powerset engine has extracted this information from articles.

Which brings us to the next question of interest: how will Google respond?