The Next NOT Big

Chicago Scraps…,” “San Francisco Abandons…,” their blueprints for urban WiFi, thanks to Earthlink’s issues. So, the next big in wifi, has become the not so big…

How to Properly Query a Semantic Search Engine

I have been using when I wax completely wigged out over some info I cannot dig up via Google. Hakia, is a semantic search engine in a Beta phase. I was just reading the blog “over there,” which has an intriguing post on the proper way to test new semantic search engines. The biggest lesson I took was this:

“A proper “test” case must include all possible variations of a query/question as listed in the table below. The column called “sampling” indicates the minimum number of cases to be tested for each variation.”

Table-1: Query Types Sampling
keyword, phrase, sentence, what, where, when, how, why, which, who, is/was/does 11+

from the Hakia blog


Some Info I Missed on Powerset

Mark Johnson, a Powerset product manager, reminded me that the search engine also offers up search results with some nifty highlighting so you really can see how relevant your search is. Visually pinpoint the data for which you’re sifting on a page of blah, blah, blah–in a jiffy. Very usable.

How Keywords May be Replaced by Old Fashioned WORDS

In the Google, Yahoo! universe keywords have become a commodity–the monetary muscle that drives search development, much the same way as Big Oil has driven energy–up til now. A couple posts ago I referred to the next big push in search–natural language. This sounds way academic for most commoners, but really what it means is this: your next gen search engine may actually be more finetuned to word meanings than your current search vehicle.

I’m a big time Google-head. I do nothing but online research day in and day out. My search capabilities have matured in the last couple years. I’ve gone from one and two-word search queries to whole sentences. For example, today I wanted to know about a rumor I’d heard about a high-end grocery store going in, so I typed in: plans for fresh market chapel hill nc. I was immediately returned a page of results that included two local newspaper articles with the keyphrases: fresh market and chapel hill. I consider that a successful search. But have I become keyword-centered?


When I checked into the Powerset blog this morning I found one of the posts most illustrative of the language flexibility that the search engine will have. The example of a Powerset search query for who proved fermat’s last theorem? returns results that not only correspond to the terms fermat and last theorem, but also understand the question is about a “who.” Not very impressive maybe until I plugged the exact same phrase into a Google search box. My results only corresponded to the keywords fermat’s last theorem, with no apparent recognition of the fact that I had asked a question about a “who.” Results were not nearly as concise as those delivered by Powerset. This, then, is a small indicator of the linguistic muscle being built into next gen semantic web.

Petaflop Supercomputer: What Big Questions Will it Solve?

Thanks to a big OOPS, the world is now privy to pretty specific info on the IBM-National Science Foundation deal–the one where the NSF grants IBM the right to build the world’s biggest supercomputer. Well, IBM currently holds the record with its Blue Gene/L. The next-gen super-duper will apparently take over with the larger (read “more important/costly”) condundrums of the moment. Already, according to a NYT article, there’s a line at this oracle’s gate. Magic 8-Ball….

Perhaps….Not Likely…..Ask Again Later……

BTW: define: petaflop-“one thousand trillion mathematical operations a second.” NYT, IBM Near Supercomputer Contract