Ness Computing Announces $5M Series A Financing to Develop Personal Search Engine
From the post:
SILICON VALLEY, Calif., July 19, 2011 /PRNewswire/ — Ness Computing is announcing that it raised a $5M Series A round of financing in November 2010. The round was led by Vinod Khosla and Ramy Adeeb of Khosla Ventures, with participation from Alsop Louie Partners, TomorrowVentures, Bullpen Capital, a co-founder of Palantir Technologies and several angel investors. This financing is enabling the company’s team of engineers and scientists, with expertise in information retrieval and machine learning, to pursue their vision to change the nature of search by building technology that delivers results and recommendations that are unique to each person using it.
The technology, which the company calls a Likeness Engine, represents a new approach to this complex engineering challenge by fusing a search engine and a recommendation engine, and will power the company’s first product, a mobile service called Ness. The Likeness Engine is different from traditional search engines that are useful for finding fact-based objective information that is the same for everyone, such as weather reports, dictionary terms, and stock prices. Ness Computing’s vision is to answer questions of a more subjective nature by understanding each person’s likes and dislikes, to deliver results that match his or her personal tastes. This can be seen in the difference between a person asking, “Which concerts are playing in New York City?” and “Which concerts would I most enjoy in New York City?” Ultimately, Ness aims to help people make decisions about dining, nightlife, entertainment, shopping, music, travel and more, culled expressly for them from the world’s almost limitless options.
Impressive array of previously successful talent.
I am not sure I buy the “objective” versus “subjective” information divide but clearly Ness is interested in learning the user’s view of the world in order to “better” answer their questions.
Depending on how successful the searches by Ness become, a user could become insulated in a cocoon of previous expressions of likes and dislikes.
That isn’t an original insight, I saw it somewhere in an article about personalized search results from search engines. Nor is it a problem that arose due to personalization of search engines.
The average user (read not a librarian), tends to search for terms in a field or subject area that they already know. So they are unlikely to encounter information that uses different terminology. In a very real way, user’s searches are already highly personalized.
Personalization isn’t a bad thing but it is a limiting thing. That is it puts a border on the information that you will get back from a search and you won’t have much of an opportunity to go beyond that. It simply never comes up. And information overload being what it is, having limited, safe results can be quite useful. Particularly if you like sleeping at the Holiday Inn, eating at McDonald’s and watching American Idol.
Hopefully Ness will address the semantic diversity issue in order to provide users, at least the ones who are interested, with a richer search experience. Topic maps would be useful in such an attempt.