Buys “Answer Extraction” Engine Brainboost for $4 Million in Cash and Some Shares of Restricted Stock

Word from that they’ve acquired an “answer extraction” database that utilizes NLP (natural language processing), for $4 million in cash and 439,000 shares of restricted stock.

The developer of Brainboost, Mr. Assaf Rozenblatt, has joined the team as Director of Natural Language Research. My experiences with Brainboost over the years have been mediocre at best. At this point answers to questions (What is…? How many…? Who are?) are derived from open web content that doesn’t always give a quality factual answer. Does a teacher want a student citing, ? Of course, this is new technology, and something we’ll likely to be seeing more in the future. So,’s investment is likely a good business move. In many ways, it’s similar to the ZoomInfo tech we blogged about earlier today.

Brainboost ( works by scouring digital content and extracting candidate answers to natural language queries. It then ranks those candidate answers heuristically and displays the highest-confidence results in simple English form. will apply Brainboost’s answer extraction techniques not only to the Web at large, as implemented currently, but to’s own growing content library of attributable reference sources.

The last sentence can only be a good thing for Brainboost with the implementation of material that does go well beyond Wikipedia data. Plus, having Answer’s resources both human and financial can only help.

Again, it’s the librarian in me talking here but it’s one thing to get an answer (something) but it’s also crucial where (the source) of that answer. Sometimes the best answer (currency, scope, authority) doesn’t come from open web. Of course, Brainboost can help the searcher get to an autoritative source but the question is will the person take the next step.

Ads on results pages are provided by Google.

Since you were likely wondering what type of answers you can find. Here are a few sample questions that came to mind in the past five minutes (literally), remember results may vary:
+ Who owned the Chicago Cubs in the 1960’s…No mention of Phillip K. Wrigley (yes, the gum guy).
+ Who are the members of Coldplay?…No mention of Chris Martin and the rest of the guys.
+ What does WGN (the TV and radio station) stand for? Correct? World’s Greatest Newspaper
+ Where was Carl Bernstein born (Woodward and Bernstein)? Date yes, location no.
+ When is Labor Day in 2006? No dates given.
+ What is Boomer Esiason’s real name? No answer
+ What does the luggage or baggage tag MSY stand for? No answer. Btw, the correct answer is New Orleans.
+ Who is the founder of the Wikipedia? No answer. Sorry Jimmy.
+ What is the name of the stadium where the Washington Wizards play? No go.
+ What are the names all six children on the brady bunch? No go
+ Who is the Prime Minister of Canada? Nope. No mention of Paul Martin.
+ What is the nickname of Texas Christian University? Nope. Answer, Horned Frogs.

However, my friend Tara from ResearchBuzz had better luck. Back in March, she had some positive comments about the serivce.

Postscript: Ask Jeeves in July launched Web Answers that mines or extracts the open web for “answers” in real-time. This is not to be confused with Ask’s Smart Answer program which takes content from structured sources. More about web answers in this post (2nd half). Like BrainBoost, caveat emptor with this type of service. Here is a question (Where is Diego Garcia) answered by a AJ Web Answer.

Related reading

How to speed up SEO analysis API advantages for SEO experts (with bonus)
Common technical SEO issues and fixes, for aggregators and finance brands
faceted navigation in ecommerce
marketing automation for SEOs, five time-saving strategies