The People's Text Analytics Software

Mar 25, 2010


      Bookmark and Share

BEST PRACTICES SERIES

The People's Text Analytics Software

When a 15-year-old kid doesn't know or care what people think about him, he tends to get a boost in punk-rock credibility. When a company doesn't know or care what people think about it, it tends to go out of business. Brand and reputation management is obviously important to a company, and Amherst, Mass.-based Lexalytics is trying to bring the kind of analytical advantages enjoyed by large corporations to smaller companies with Lexascope, the company's latest web-based API for the masses released Tuesday, March 23.

"We wanted to come up with something that was lower-cost and easier for people to implement and use," says Seth Redmore, Lexascope's vice president of product management. Lexascope is based on Lexalytics' older, more expensive program, the Salience text analytics and sentiment software. In order to bring the functionality to a mass market, the engineers behind Lexascope employed cloud computing to streamline the service and scale it to smaller-volume users.

"It's basically about bringing the Lexalytics functionality to a cloud-based solution," Redmore says. Once inside, Lexascope is a text analytics program whose sophisticated processing engine is belied by its simple GUI.

Lexascope can be programmed to analyze Twitter, RSS feeds, online news, customer surveys, and other documents for trends, sentiment, and linkages between different entities. "What the GUI shows you is entities and themes and companies," Redmore says. "A circle is a company, a square is a person, and triangles are themes."

Any one of those people, companies, and themes is linked to other points it's mentioned with in the articles and is given a sentiment score. The GUI is designed to be simple, and the API is designed to be easily manipulated to interface with a company's internal computing system.

"What I learned during the beta period," Redmore says, "is that there are too many interfaces that people have to deal with, and when you learn one interface, unless you keep using it you have to go back and learn it each time."

One of the major selling points of text analytics programs and services is the number of man-hours it frees up. Instead of having one or more employees coding text, a process Redmore calls "labor-intensive and inconsistent," a computer performs the task more quickly, accurately, and, ideally, cheaply. However, much of the existing analytics software contains expensive and complicated programming that the people behind Lexascope thought the average small-business user could do without.

"You're throwing a lot of software at the problem," Redmore says. "It's sort of like shooting a flea with a bazooka." Armed with a simpler, cheaper weapon, Lexalytics went out to capture markets that previously might have viewed analytics software as out of reach.

"One [market] is larger enterprises that have point text analytics problems they have to deal with. It frees up a person from having to deal with this," Redmore says. "The second is companies that are working to provide vertically specific media and reputation management services-companies that...have a fairly small clientele and want to provide some good analytic services."

Redmore says companies have integrated the Lexascope API into their companies in several ways, from business intelligence to search engine optimization. "They're kind of two sides of the same coin in that they both deal with data," Redmore says.

Lexascope Web Service is free for up to 1,000 documents per day. A $400 per month subscription allows users to process up to 50,000 documents per day. By cutting costs and complexity, Lexascope may become the text analytics tool for the masses.

"It's a simple service," Redmore says. "We're hiding all the complexity."

(www.lexalytics.com)