Applications:

SAS Text Mining Software Turns Prose into Profit

Words fill the world of work. Every day, we're flooded with e-mails, Web pages, memos, agendas and countless other documents filled with simple words.

That's why SAS, a leading provider of business intelligence, announced the general availability of SAS Text Miner, adding advanced text mining to the company's market-leading data mining solution. Text Miner allows users to discover and extract knowledge from a wide variety of text documents, including sales call reports, medical records, Web pages, patents, warranty claims and e-mailed customer communications. The software can access text data in Microsoft Word documents, PDFs, HTML and many other kinds of files stored in a wide variety of databases and file systems.

For instance, text mining can help a sales manager examine free-form text notes collected from customer e-mails and by telesales representatives, categorize these notes into meaningful clusters and then make informed decisions on up-sell or cross-sell opportunities given each customer's position in the sales cycle.

Medical researchers can use text mining to review patient records, analyze medical conditions and treatments, and then identify trends that can help healthcare providers understand what combination of approaches works best with certain patients.

"Text Miner solves the problem of too much unstructured, text-based data piling up in organizations. Everybody knows this data is valuable, but until now no one could exploit it," said Jim Davis, executive vice president and chief marketing officer for SAS. "With SAS Text Miner, organizations finally can apply analytics to discover unseen patterns and unearth intelligence otherwise buried in these collections of free-form text."

Text Miner goes one step further than other available text mining products by providing full text preprocessing within the powerful, easy-to-use SAS Enterprise Miner process flow environment. Users can obtain further insight and intelligence by integrating unstructured textual data with existing structured data such as age, income and purchasing patterns.

"SAS is an emerging leader in enabling the analysis of unstructured data, including text, from within an enterprise data mining workbench," said Andrew Braunberg, senior analyst at Current Analysis. "With text mining working in concert with data mining, SAS' analytical solutions are even more compelling to companies looking to solve real-world business problems."

Through a strategic text mining alliance with Inxight Software, Text Miner ships with rich multilingual text analysis capabilities in English, French and German. Support for 13 additional ASCII-based languages is available through Inxight.

Text mining plays key role in healthcare

Hospitals affiliated with the University of Louisville report healthcare information to organizations like the Joint Commission on Accreditation of Hospital Outcomes (JCAHO) and the National Committee for Quality Assurance (NCQA), two of the many industry groups that create performance measures and accreditation standards for reporting healthcare information to the public.

Outside organizations rank hospitals and healthcare plans throughout the United States based on information reported by healthcare providers. According to Patricia Cerrito, a professor of mathematics and biostatistician for the University of Louisville, it's important to consider the source of the ranking -- the reported data -- when hospitals seek to improve their rankings.

In an effort to standardize reporting practices, Cerrito is using SAS Text Miner to explore medical records from more than 10,000 open-heart surgeries performed across the country in the last five years. The SAS solution pulls relevant variables from patient charts and makes it easy for Cerrito to look at patterns in the way healthcare professionals treat patients and report patient information.

For example, Cerrito has discovered that the under-reporting of patient risk factors has a large impact on hospital rankings. Hospitals that under-report patient risk factors will have lower predictions for patient mortality. So even if their success rates are equal to other hospitals, their ranking will be lower, because they report a greater difference between predicted mortality and actual mortality.

Using Text Miner, Cerrito compares information from patient charts to the information that is abstracted for reporting purposes. She can easily see what has been missed and determine how improved coding of diagnoses and risk factors can increase the hospital's ranking.

"By examining the data for all open-heart surgery patients, we've discovered that the patients treated at our hospitals have been more ill than we've been reporting," Cerrito explained. "We use SAS Text Miner to identify other health conditions that commonly affect open-heart patients -- such as diabetes, urinary tract infections and allergies -- and then we can go back to our physicians to make sure they fully document all the ailments impacting our patients. That way, our physicians and the rest of our healthcare staff can get the credit they deserve for the great job they're doing."

Already, Cerrito's research has helped solve problems and improve national accreditation rankings at local hospitals. But ultimately, her results should lead to better patient care. "Better reporting of quality measures makes everyone look more closely at what they're doing," Cerrito says. "And more effective reporting will help everyone see the value in the existing measurements."

About SAS

SAS is a market leader in providing a new generation of business intelligence software and services that create true enterprise intelligence. SAS solutions are used at more than 38,000 sites -- including 99 of the top 100 businesses on the Fortune 500 -- to develop more profitable relationships with customers and suppliers; to enable better, more accurate and informed decisions; and to drive organizations forward. SAS is the only vendor that completely integrates leading data warehousing, analytics and traditional BI applications to create intelligence from massive amounts of data. For 25 years, SAS has been giving customers around the world The Power to Know. Visit SAS at www.sas.com.