Expert System, by realizing the limitations of how computers operate, has taken a unique view of databases and bulk information. The company’s main goal is to improve multilingual cognitive computing, and in an effort to reach that goal, has opened two new research and development labs in Washington, D.C. and Silicon Valley.
The company’s announcement contains the words of its chief technology officer, Marco Varone, who admits that the scientific community cannot exactly explain how humans comprehend language but that we know enough to allow computers to mimic our behavior. Expert System will work inside those limitations to expand the capability of language understanding in computing systems while also remaining realistic about what businesses can achieve.
“No one knows exactly how human comprehension really works, but we have made encouraging progress in programming computers to understand language,” Varone said.
“Our newest Cogito Lab will be dedicated to further expanding the core functionalities of our technology and take new steps towards bringing to market software that really reads like humans and learns from humans,” he continued. “Additionally, we will have a related effort in developing tools that make this power available to almost anyone.”
The two new labs come as a welcome addition to Expert System’s current labs in Modena, Roverto; Naples, Italy; Grenoble, France; and Madrid, Spain. The operations in D.C. and Silicon Valley will continue with the same mission as the others: to move beyond the type of search of which many people are familiar, such as those at Google.
The common search algorithms do well when they look for keywords. They perform operations quickly and do exactly as they are programmed. However, they also have no concept of understanding what a search for “dog” or “computer” means in context. They only know, for instance, that those keywords may be located in a number of different Web pages.
In order to dig through mountains of contextual data a derive meaning from those stores, Expert System wants to create computing systems that understand text how humans understand text. Imagine searching for the phrase “expert poodle handler great job” in a Google search. It may be able to find local and regional websites associated with dog trainers and even provide definitions of what a dog trainer does.
In contrast, when searching for those common terms in a store of data, companies may want to know how each word relates to the paragraphs around it or what a customer’s comprehension of those terms means. Does a customer want to hire a trainer? Praise one for past performance? Find a message board related to dog training? Brief phrases such as the one above can deliver a lot of meaning with the context of larger text stores or audio files (perhaps translated to text). In any case, Expert System wants to dig deeper.
Humans can intuitively provide insight about the feeling or context of any terms they consider. Computers may be on their way to that reality, but not without the help of companies such as Expert System. The two additional research labs in the U.S. are expected to pull their weight and push forward the ability of computers to truly understand the meaning of data. Varone and his company appear hopeful that their investment will provide a complete return of investment.
A rapidly increasing need for APIs and similar measures is expected to drive big gains in this market through 2021.
Doug Waller of Flowroute will be taking part in a panel at the All About The API event, collocated with ITEXPO, to discuss common mistakes that occur …
Google has teamed up with H&M's digital fashion house Ivyrevel to use its Awareness API in a unique way. The duo wants to digitally design customized …
Open source has become an integral piece of every developer's arsenal. The power of the community, the wisdom of many, and the ability to hook into va…
Google's in an interesting place with Hangouts and chat with both the enterprise and the consumer. One thing is clear - what it's doing with Hangouts …