Episode 24 “Bucket search system”

Responsibility for fostering a digital culture
Acquisition of groundbreaking technology Limited target users

Ichitaro was originally intended for use in operational documents, but as its user base expanded, I came to think that we should no longer provide dictionaries solely by ourselves. This was because I felt that our company was responsible for fostering a digital culture of the Japanese language.

This is how the “ATOK Supervisory Committee” was formed in 1992, with writer Jun-ichiro Kida as chairman. Advanced Technology Of Kana-Kanji Transfer, abbreviated ATOK, is still used in Ichitaro’s dictionaries.
Originally, I created the dictionaries by referring to newspapers and other information sources and assuming the word order, etc., that users would use frequently, and we have also hired personnel with a background in Japanese linguistics, but it was still a great help to have normative standards for vocabulary selection based on the knowledge of experts.

At the same time, we were actively involved in the international consortium for “Unicode,” a universal character encoding standard, to establish the status of the Japanese language overseas, and we have achieved results.

At the same time, we focused on discovering the seeds of new businesses. Therefore, in 1996, we opened a research institute in Pittsburgh, the U.S., named “JPRC”. Then one day, we received a call from its director. He said he had an interesting technology to show us.

At that time, I was seeking a search technology to accumulate and apply a large amount of text that users had created using Ichitaro. At that time, more than 10 years had already passed since the first version of Ichitaro was released, a large amount of document data had been accumulated by companies and local governments.

It was like a sea of letters. I wondered if we could develop a technology to extract and apply the appropriate text at any given time from this sea. That was my idea.

I used to call this technology I was seeking a “bucket search system”. This was meant to imply finding what one wanted in the bucket of text that was accumulating steadily. And the “concept-based” technology that the director of JPRC discovered was this very bucket search technology.
I flew to Pittsburgh with Hatsuko and other researchers and was briefed on this technology firsthand at the JPRC. And I was immediately convinced that this was exactly what I had been looking for. The feature of this technology is that it can not only search for keywords, but also search for sentences as they are still in concept.
David Evans, the developer of this technology, started his own company while working as a professor at Carnegie Mellon University, but was struggling to grow his business. I therefore made the immediate decision to acquire his company. The purchase price was over 2 billion yen, the maximum we could afford, but we had found a breakthrough technology that was worth it.

As you may have already noticed, this “concept-based” technology was, I believe, ahead of Google in the U.S. The first product incorporating this technology was launched in Japan in July 1997. Meanwhile, Google was founded in Silicon Valley in September 1998.

However, my vision at the time was to deploy this “concept-based” technology in Japan, and to target Ichitaro users in particular. It makes no sense to talk about “what if” when it comes to business, however, I wonder what the future would have been like if I had deployed this technology overseas from the very beginning. I might have changed the world.
In 1997, the company went public on the OTC market. This was not for raising funds, but because Hatsuko and I felt it was necessary to ensure the liquidity of the shares we held. That is why we never reacted nervously to fluctuations in stock prices after our listing.
However, the business environment surrounding Ichitaro, our flagship product, gradually became harsher, and we posted a loss for the fiscal year ended March 1998, immediately after our listing. This result was met with a series of harsh criticisms by investors and the media. On the other hand, the “segment strategy” of focusing on sales to schools and local governments was bearing fruit. Nevertheless, Microsoft’s cross-selling of basic OS software and Microsoft Word had a significant impact on Ichitaro’s sales.

With David Evans (middle)
and my wife Hatsuko (right) in July 1996