Original text on nytimes.com
SAN FRANCISCO, Feb. 8 — Early in the decade, a struggling Xerox Corporation was trying to sell off a stake in its Palo Alto Research Center, which it could no longer afford to support. But with the technology bubble bursting, the price that investors were willing to pay for a piece of PARC, as the center is known, kept going down.
So in 2002, Xerox switched to Plan B: it spun off the center into an independent subsidiary and sought to prove that it could sustain itself by licensing technology and forming partnerships with outside companies.
On Friday, PARC is announcing a deal that underscores that strategy. It is licensing a broad portfolio of patents and technology to a well-financed start-up with an ambitious and potentially lucrative goal: to build a search engine that could some day rival Google.
The start-up, Powerset, is licensing PARC’s “natural language” technology — the art of making computers understand and process languages like English or French. Powerset hopes the technology will be the basis of a new search engine that allows users to type queries in plain English, rather than using keywords.
In the fall, Powerset raised $12.5 million in its first round of financing from venture-capital firms and individual investors. The challenges facing it are immense, and the odds of success are low. But the PARC technology, which is a result of 30 years of research, is certain to lend it an aura of credibility.
PARC’s natural-language technology is among the “most comprehensive in existence,” said Fernando Pereira, an expert in natural language and the chairman of the department of computer and information science at the University of Pennsylvania. But by itself, it will not guarantee Powerset’s success, Mr. Pereira said.
“The question of whether this technology is adequate to any application, whether search or anything else, is an empirical question that has to be tested,” Mr. Pereira added.
As part of the deal, a leading natural-language researcher at PARC, Ronald M. Kaplan, will join Powerset’s staff of about 40 as chief technology and scientific officer. PARC will also receive an equity stake in Powerset and earn royalties from the company. Additionally, Powerset will sponsor a handful of researchers at PARC.
The specific financial terms of the agreement are not being disclosed. But Mark Bernstein, president and center director of PARC, said: “It’s one of the biggest deals that we have done, and we hope that it grows into the biggest in terms of the length of the relationship and the amount of value we can create together. It represents a commitment of some of the intellectual crown jewels that PARC has created.”
As part of the business model forged when it was spun off, PARC has struck various business relationships with outside firms and organizations.
About half of its research is still sponsored by Xerox, Mr. Bernstein said. But the lab is also conducting paid research for Fujitsu, Dai Nippon Printing and others. Some of its researchers work on federally financed projects, and the lab is working with ipValue, a intellectual-property licensing firm, to commercialize some of its research.
PARC has also formed a partnership with the Scripps Research Institute in San Diego to develop a system that uses laser printing technology to detect cancer cells.
And in the deal that most closely mirrors the alliance with Powerset, PARC has helped incubate SolFocus, a start-up that is developing solar power technology.
PARC now has about 220 employees, including 160 researchers, down from its peak of 318 employees and 100 contractors and temporary workers in 1995. Mr. Bernstein said PARC, which has an annual budget of $55 million to $60 million, is profitable.
“We are very pleased with where it is today,” Anne M. Mulcahy, Xerox’s chief executive, said about PARC. Ms. Mulcahy said she did not rule out the possibility of selling an equity stake in PARC in the future, but added, “We are very comfortable with continued ownership.”
In Silicon Valley and beyond, PARC has often been called the lab of missed opportunities. It has been credited with many breakthroughs, including the graphical user interface and the Ethernet networking technology, that have revolutionized the computer industry, but that were commercialized by others.
“There’s no way anyone can top what they did in the past in terms of dramatic research developments,” said the futurist Paul Saffo, a fellow at the Institute for the Future. But Mr. Saffo praised PARC for finding a business model that has allowed it to survive at a time when many research groups at American corporations are being cut.
“This is an organization that has done well at keeping researchers, and spinning out a steady stream of little products,” Mr. Saffo said. “PARC has been a very quiet success.”
The success of its bet on Powerset is another matter. Over the past year, PARC researchers have worked with Powerset engineers to build a prototype, but the company does not expect to release its search engine to the public until the end of this year.
Meanwhile, other start-ups and several of the search giants are also working to develop natural-language search technology. The appeal is clear. A successful natural-language search engine could, in theory, answer real questions — for example, what companies did I.B.M. acquire in the last five years? — that existing search engines are not equipped to handle. And it could turn the process of finding information on the Web into a conversation between the search engine and the user.
“For a lot of things, keyword search works well,” said Barney Pell, chief executive of Powerset. “But I think we are going to look back in 10 years and say, remember when we used to search using keywords.”
Researchers have predicted breakthrough applications for natural languages for years, but the technology has proved usable in only limited contexts, turning many experts into skeptics about its potential, at least in the short term.
“My general feeling about natural-language processing in search is that I’m a bit of a skeptic in the sense that even the best systems, and I include there the systems from PARC, make many mistakes,” said Mr. Pereira of the University of Pennsylvania.
In a November interview, Marissa Mayer, Google’s vice president for search and user experience, said: “Natural language is really hard. I don’t think it will happen in the next five years.”