Scalable Semantic Analysis

by David Johnson on June 2, 2010

I’ve spent a fair amount of time researching structured and unstructured text analysis and analytics tools that are used to uncover meaningful insights from complex data sets.  I’ve seen some new tools emerge that are starting to enable data-mining from corporate archives, CRM systems, legal archives, U.S. Patent and Trademark Office and the web.

This is an increasingly crowded space in a market that is in its infancy.  There are solutions based on keyword search, Boolean search, natural language theory, artificial intelligence, latent semantic indexing, hybrids and many others.  Those that have spent any time in this space understand that it’s just not as simple as getting data, and running a text analysis algorithm against the dataset.  In fact, there are several bottlenecks in the text analysis process that are very challenging: data capture, cleansing, storing, categorization, indexing, retrieving, analyzing, interpreting and reporting.

Companies that are performing these services are doing so on a very small scale or they have deployed very specific industry solutions based on high-performing data warehouses or data warehouse appliances (algorithms embedded in servers and/or storage frames like Netezza).  To date, I’ve not seen any solutions that are capable of scaling to the breadth of challenges that companies or brands face.   More often than not, it seems like many of the solutions are “technical” in nature and they force the client to become the text analysis expert, analytics expert and/or behavioral science experts.  To my mind, that means that 90%+ of the market is underserved, but there is an option.

I can’t divulge the IP, methodology or distribution strategy of this solution.  However, you should know that there is something available that will provide valuable insights that leads to measurable results.  Without getting into too much detail, the platform I am speaking of scales to the needs of nearly all companies, one that scales both functionally and technically.  By functionally, I mean that it leverages several text analysis and analytics algorithms that can be applied to a multitude of challenges across all industries.  From a technical perspective, it enables access to many, many, many petabytes of data, thus enabling valuable insights to be derived from insane volumes of data…in real-time.  The  solution is also language independent, so it is multi-lingual right out of the box.

If you’re interested in learning more about how text analysis, semantic analysis, sentiment analysis, multi-channel mix analysis, social monitoring and other analytics solutions can be used to capture meaningful insights to propel your business, please feel free to contact me at 678-761-5395.




Leave a Comment


Previous post:

Next post: