Smartbrief uses Classification

  1 m, 11 s

Here's the fourth video taken from our LUG conference.

Smartbrief is one of the first adopters of our Concept Topics wikipedia-based classification technology. In this presentation, Zack Mully discusses how concept topics help them sift through the news. The goal at Smartbrief is to "make people smart and save them time", but doing so is a lengthy process for the Smartbrief team. By using Salience, the team automates a large portion of its work, and creates sample sets that allow the engine to sift through data and drop them under various umbrellas.

Things are currently running smoothly, but a lot of trial and error took place behind the scenes before Smartbrief reached their present state. Zack was nice enough to walk us through their process, going from what attracted them to Lexalytics in the first place (the straightforward OEM and the excellent Java integration), to what worked well with the Salience text analytics engine, as well as covering various pain points, and how they worked around them.

Zack Mully oversees SmartBrief's technology department. He has the unique distinction of being the most tenured employee at SmartBrief – he started at the company in February 2000 as its fifth employee. With his strong systems operations focus, SmartBrief has never missed a day of publishing because of technology outages. As the titular head of the technology department, he's turned on by the big data movement, nosql, graph databases and undocumented system procedures.

Categories: Categorization, Events, Lexalytics User Group, Partners, Presentations, Topic Extraction