Jennifer Zaino
SemanticWeb.com Contributor

The paper, published by Seth Grimes, founder of DC-based consultancy Alta Plana, identified the top business applications of text analytics as the following: brand-product-reputation management; competitive intelligence; voice of the customer/customer experience management; and other research.

Based on a survey of 116 respondents, most of whom already use text analytics, Grimes finds that the text they want to focus on includes material that's outside the enterprise firewall -- information in blogs and other social media; news articles; and online forums, as well as more internal data such as survey results and emails.

But discovering and analyzing the text that's out on the web is part of the way companies can know what users are saying about their products and brand, what competitors are doing and what they themselves may be exposing online, and react quickly to what they learn, Grimes told attendees at a recent webinar entitled, "From Metadata to Meaning: Intelligence in the Semantic Era." (The webinar was hosted by Nstein, which offers web content and digital asset management, and text mining engine solutions, and which also was one of multiple sponsors of Grimes' editorially independent paper, Text Analytics 2009: User Perspectives on Solutions and Providers.)

The direction that companies want to take text mining, encompassing a wealth of unstructured text data, presents some problems in the traditional search realm.

"Search is designed to help you find information you already know about, but it's not very sophisticated," Grimes noted.

But that won't stop this trend, he said. "The semantic web is coming, and it will offer real business value in the next few years," Grimes said during the webinar. "Not this year, maybe in the next year (though probably not), but definitely after that."

So be ready for it by publishing tagged information (using microformats, RDFa, RDF) to gain the opportunities the semantic web will offer as users start adopting it, he advised.
Back to the issue of text, which is in a form that is difficult to decipher automatically. Text mining via semantic analysis enables smarter search to better respond to users' goals. The semantic web opens the door to augmenting text by generating metadata and tagging content to support distribution and retrieval, enable the extraction of information to databases for analysis, support exploratory analysis and visualization, and automate content categorization and classification.

Text mining today can extend also to semantically tagged audios and videos, and that information can be analyzed via business intelligence tools, too.

"By automating the reading process, text analytics allows analysts and researchers to tap material that had not previously been systematically mined," Grimes writes in the paper. "It allows them to work far faster than before and to analyze far greater volumes of information than ever before. Importantly, text analytics can make a huge difference in text analysis and processing costs and enable the creation of new information products and services."

For online content publishers, the business value users want to gain from the information on the web that they are so instrumental in producing translates into an opportunity for them to leverage semantic web technologies to their customers' benefit -- and their own. According to the survey, users primarily want to extract or analyze named entities (people, companies, locations, etc.); topics and themes; sentiment, opinions and emotions; concepts; events and relationships; and metadata such as document authors or publication dates.