Hitmetrix - User behavior analytics & recording

Salience: making smartphones smarter

There is too much information coming from that little smart phone you take everywhere with you. That digital teddy bear is turning into a monster. And it always wants your attention.

Eager to tame the beast, Lexalytics released Salience for Android, a native text analytics package that will sort through incoming message streams to pick out all the important stuff and put it on your screen. “It’s our first foray into the mobile market,” said Seth Redmore, Lexalytics’ Chief Marketing Officer.

Salience for Android is an abbreviated version of Salience, a cloud-based SaaS application used to analyze text and sentiment in real time. “We didn’t put everything in the phone that we could,” said Redmore. “The small screen means you have a limited amount of real estate to display stuff.”

Salience for the cloud will find out new things, track things and filter things out. But for the handheld, the limitations became apparent. “Our gut tells us it will be most interesting on a local platform with filtering capability.” Redmore said. By knocking the app down to smartphone size, Lexalytics then relies on its filtering capabilities to prioritize and present information to the user. “The phone is an extension of you,” Redmore said. It knows your contacts. It knows your schedule. “It is a convergence of information sources you set up,” he said.

Once set, Salience for Android should present the information you want to look at when you want to. But the filtering and language analysis can also serve up “imperatives”—messages from someone who is asking you to do something, Redmore explained. Extraneous messages don’t surface as clutter.

The ideal user is someone who has too much information to track and too many tasks to juggle, Redmore said. This might be your average middle manager or executive. Or it might be an IT support person who has 20 things to do without knowing for sure which one to do first.

To analyze language, Salience tears apart sentences. It distinguishes punctuation marks, then “chunks” the pieces and then relates the chunks to each other, then figures out the relation of the sentences to each other, Redmore said. That capability is delivered through a combination of machine learning, natural language rules and reference to dictionaries.

Lexalytics chose the Android platform as a starting point for its mobile efforts, Redmore said. Android is an easier platform to work with, he noted, while developing for the Apple iOS is more difficult. Salience for Android only takes up 16 megabytes of space on the Android smart phone, Redmore noted. This compares well to the 2 GB footprint Salience has in its cloud iteration. More importantly, Salience for Android  is made to work “locally”—on the smart phone itself. This cuts out the need to search for the app to search bandwidth to communicate with its larger cloud-based cousin.

Lexalytics is also offering a software development kit to craft apps for Salience for Android. If the app produces less than $5,000 a month for the developer, then Lexalytics will give away the SDK for free.

Related Posts