Technology firm hopes to answer questions (Kozoru)
- The Johnson County Sun
Technology firm hopes to answer questions By: Adam Lee, Sun Staff Writer July 29, 2004 John Flowers is the founder of the Overland Park-based company called Kozoru. He hopes to develop a platform to make searching the Internet easier and more intuitive.
John Flowers is haunted by his first encounter with a computer.
The machine was a TRS-80 Model 1. Flowers was 9 years old.
"I remember the first thing that I did was I sat down to this blinking, green cursor and typed in a question, thinking that I could interact with the computer that way," Flowers said. "I think that was everyone's first experience."
Twenty-four years later, Flowers believes he has figured out a way to create that degree of interaction between man and machine.
Banking on that belief, Flowers launched a search technology company called Kozoru this month in Overland Park.
Kozoru's goal is to transcend the likes of Google, Yahoo, and Microsoft by creating an entirely new platform that would allow language-based - not keyword-based - searches.
In other words, you'd able to sit down to a blinking, green cursor, type in a question and get an answer - the right answer.
Flowers and his team believe today's search technology user is being sold on the idea of "more is better" instead of "correct is better."
Technology such as Google, Yahoo or even Ask Jeeves overwhelms the user with information, Flowers said, much of which is inaccurate and/or irrelevant. He equates this situation to asking for a needle and getting back multiple haystacks.
Google uses a page-ranking system that basically determines the relevance of a Web page by counting how many other pages link to it.
In 2003, Google was used for more than 70 percent of all searches in the United States.
And while it currently leads the industry, Google's technology boss Craig Silverstein told the BBC in March that the company is trying to develop a more intuitive way for users to locate the information they're searching for.
"A rank list is not the most natural way to find information. We think about that a lot but we're not convinced the alternatives are better," Silverstein told the BBC.
"The right way to search is to do what a person would do, that is the ideal we are going for. The problem is that we have no idea how to get there."
That's a good source of positive reinforcement for Kozoru, according to communications manager Justin Gardner.
"He's basically saying that the way Kozoru is approaching this problem is the right way to do it," Gardner said.
It's a problem that Flowers, 33, has been trying to solve ever since that first session on the TRS-80.
The fundamental paradox, as Flowers puts it, is that computers are really good with math but really bad with language.
Flowers struggled with this dilemma through a stint working for Microsoft and while helping develop the Inquisite Web survey system and starting his first entrepreneurial effort, nCircle Network Security.
"Then I gave up, frankly. January 2003, I said the heck with it. Technology was no longer interesting to me, and the really hard problem that I wanted to solve is unsolvable," Flowers said.
Flowers, who holds degrees in English and philosophy, spent the next few months writing books and screenplays. His wife, Gretchen, gave birth to their son, Nicolai, and Flowers soon forgot all about the tech industry.
Last summer, the Flowers family decided to leave Silicon Valley on a road trip without any sort of itinerary. Their initial destination of Chicago later became Boston, then London, Paris, Rome and so on until 10 months later, when they found themselves in southeast Asia.
"The idea was Zen travel, to aimlessly wander the globe to a place we hadn't been and relax," Flowers said. "And get as far away from technology as we could."
When in Thailand, Flowers decided he would like to explore Buddhism and asked if he could study at a temple. His wife and son went back to the United States, while Flowers spent the next two and half months "wearing robes and wandering around southeast Asia."
Flowers said the experience taught him some very valuable lessons, one of them being that he isn't very good at meditating.
"I sit down and I try to focus and it just doesn't work. At least not focusing on the things they ask you to focus on," Flowers said. "Instead what I focused on was all of these third-order predicate calculus problems, where language could be converted into some sort of mathematical representation, and the computer could solve questions."
In February of this year, Flowers came up with an answer and came back to the states to put it to work.
"In math ... you have an equation and you have a solution. There's a language counterpart to that, it's called a question and an answer," Gardner said.
After translating more than 980,000 words in the English language into codes of ones and zeroes, Kozoru's first objective will be to establish a knowledge base.
To do this they will first turn to the most objective source for language information, the dictionary. After establishing that system, they will incorporate the most objective source for historical information, the encyclopedia.
"That creates the very base for language and how language works, because it has all the rules included in it and it also has all the history," Gardner said.
He said this knowledge base will include a variety of primary sources such as medical and legal dictionaries and encyclopedias.
This first iteration of Kozoru won't be for the Internet, Gardner said, but will still have immense implications and applications.
"It will be a very valuable tool for academics," Gardner said.
It could also be used commercially to offer businesses quick, direct access to their databases and documents, no matter how expansive or unrelated those items may be.
"You can pour that data into the Kozoru engine and it will be able to understand what you're looking for by how it sees the information within those documents given the baseline we have created with language," Gardner said. "It will be able to infer, understand and dynamically create questions and answers."
On top of that knowledge base, Kozoru plans on adding established sources of news, the most objective source of information available in real time, to make a more adaptive and immediate source of information.
When Kozoru makes the final step of becoming a free Internet search tool, the company plans to weigh opinion on top of its three objective sources of authority to comprehensively evaluate all sources of information on the Internet.
"It will be able to truly judge, in real time, the knowledge and information being created every single day," Gardner said.
Flowers said he hopes to have the initial Kozoru prototype developed in the next nine to 12 months. It's a tight turn-around for developing such a revolutionary technology, but Flowers has developed a highly skilled team from his past software connections.
It wasn't exactly easy to sell them on an idea that hundreds of people had been working on for years and he had come up with while sitting on a dirt floor in Thailand.
"They absolutely thought I was insane," Flowers said.
Adding to the skepticism was Flowers' decision on where to locate his venture. Instead of going the traditional route of setting up shop in the Silicon Valley, Flowers decided to stake a claim on the digital prairie.
Flowers had spent some time in the Kansas City area doing some consulting work; his son was born in Overland Park; and, he said, his wife fell in love with the area. He said the quality of life and cost of living is unequaled on the coasts.
"I feel a real sense of community and a sense of geography with this place," Flowers said.
He eventually was able to sell everyone he recruited on not only the Midwest location but his plan for Kozoru. That includes David Warthen, co-founder and architect of Ask Jeeves, and Ridgely Evers, who developed Intuit's Quickbooks software, who both are members of Kozoru's board of directors.
Flowers' investors include the state-managed Kansas Technology Enterprise Corp.
Flowers said companies like Netscape, Apple and Microsoft all became successful, in part, by not doing things in the traditional way. Kozoru is hoping to take that idea to the next level, too.
"We're not trying to think outside the box," Flowers said. "We're trying to live outside the box."
©The Johnson County Sun 2004