Dan Farber reminded me again last week of the fascination I had for PowerSet when they first debuted their algorithm for natural language search & semantic parsing analysis last year.
While there are several firms that play in Powerset's space (e.g., Radar Networks, Hakia, Mahalo) they're innovative application of the Palo Alto Research Center's (PARC) Natural Language Parsing (NLP) technology really struck a chord. Who couldn't be fascinated by algorithms capable of "...enabling machines to understand and respond to what (actually) people mean...?!
Unlike most search engines, Powerset promises the ability to understand the context of search-query phrasing through linguistic comparison and analysis that iterates and grows the more it is used. It breaks down whatever verbiage we use by parsing the subjects, objects, verbs and synonyms to extract contextual concepts and their intended meaning. A far cry from even the best and most complex keyword search.
The commercial opportunities are pretty amazing. From IP-licensure to end-product-consumption, Powerset's application of Chesbrough's open innovation has built an opportunity-rich model ensuring they'll be around for some time.
The potential application of Powerset's capability to social-networking is equally fascinating when you consider the value offered through an improved contextual understanding for relationship connections.
Why couldn't we, afterall, apply similar algorithmic logic to understanding the impetus behind social network connections that are highly sought and highly active?
An organization's interest in acquiring conscious understanding for the contextual nature of relationships that sustain themselves -- because they "make life easier" -- seems a natural extension of the forms of social network analysis we'ved discussed before.
And this brings me to Twine...
Twine is an online service that basically combines elements of social-networking, collaboration, communities-of-practice and knowledge management all under one roof.
Using semantic capabilities similar, in some respects, to Powerset's underlying algorithmic capabilities - Twine offers a unique collaboration platform for connecting individuals around specific content themes. Once you become part of a "twine" you have access to fellow subscribers interested in your topic - but, more importantly, you have access to all of the content they've self-selected as relevant and useful...regardless of its form.
I've selected 3 twines and decided to create my own on...of course, enterprise social software.
So - imagine the context for social networking application within business software...
Everyday, the business applications we use offer automation for specific transactional steps - usually defined by software vendors - devoid of context and specific relevance for a company's culture and the individual experience of its employees.
And, while there are often many individual "users" of a software solution - e.g., a sales force automation solution within a CRM application - rarely do these users engage with one another and collaborate "inside the solution" unless they are approving transactional step amongst one another.
Application of capabilities similar to Powerset's and Twine's as embedded components to the management of a business process, would change that.
Instead of only waiting for my pipeline entry to be approved as an official pipeline opportunity by my VP - to extend our SFA example in CRM - I could have access to all of the content, best practices and experiential knowledge he (or she) has that could actually help me to close the deal, as well.
Applying semantic-web functionality to the "process management" capability of today's business applications would be a true coup for any software vendor.
It will be interesting to see who leads the charge...
Link to original post