LexaBlog: Our Sentiment about Text Analytics and Social Media

Wow, it has been a while since we blogged. That's bad. Sorry about that, but we've had some new developments coming out of the company, literally.

First, we helped launch a new subscription-based product called Lexascope - www.lexascope.com - and it is powered by Lexalytics' Salience technology. We previewed this in October at the Inbound Marketing Summit and are pleased to have it avaiable for download.

It's still in beta, but we are encouraging everyone to take a look and sign up for the 15 day FREE trial.  Let us know what you think. The first desktop application is geared towards PR professionals who want to discover entities, themes and sentiment from RSS feeds or online content.

In addition, we're working on a new product called Lexalytics Cascade.

Cascade attacks a class of problems not addressed by our Salience product, namely stateful content processing. It provides users the ability to look across a collection of content and perform tasks like content filtering, document similarity and collection level theme rollup. In the initial release, we focused on providing a high performance filtering engine to bucket content via user-defined queries, and to provide a scalable document similarity engine capable of measuring the similarity of terabytes of documents.

If you're thinking, "exactly what is a filtering engine?" The easiest way to imagine it is to think about it as a search engine standing on its head. In a filtering engine, the queries are indexed and the documents flow across the engine and act more like queries, where the documents are "filtered" into buckets represented by the indexed queries.

Filtering engines are designed to operate against live flows of content like newsfeeds or twitter streams. The advantage of a filtering engine over a search engine to bucket content is simple, PERFORMANCE. A filtering engine can filter hundreds of thousands of documents per hour. This capability combined with the engines similarity capabilities means that you can process large flows of content with Cascade, and identify duplicates and/or syndicated documents in your document stream.

Throughout 2010 Lexalytics will further enhance Cascade with a series of new releases focusing on the aggregation and rollup of concepts across whole collections of content. 

We hope you forgive us for being gone for so long. As we approach the holidays we want to wish you and your families all the best for a healthy and happy holiday, and look forward to more (frequent) posts in 2010.

~The team at Lexalytics