The Power to Know

IN THIS ISSUE:

Text Mining Advances Educational Research and Undergraduate Learning

Using SAS® Text Miner to Analyze Open Source Software Development
Students Use SAS® Text Miner to Analyze French Education Market

Applying Text Mining in Bioinformatics
Students and Professors Put SAS® Text Miner to Work
Text Mining Reveals Hidden Value and Competitive Advantages
Report from SAS Asia Pacific Academic User Conference
New Statistics Certificate Program at Kennesaw State University
New African Business Intelligence Hub Established
Forthcoming Book from SAS® Press to Include Chapter on Text Mining
Data Mining for Educators To Be Offered in Three Locations in 2007


SUBSCRIPTION
This newsletter is a non-obligatory gratis service from SAS.
To subscribe to this newsletter, please click here.
To unsubscribe, please click here.


RECOMMEND A FRIEND
If you know someone who would also like to receive this newsletter, please forward the subscribe link to them.

CONTACT US
We welcome all your opinions and your suggestions for this newsletter. Feel free to contact us by e-mail.


SAS Academic Program University Newsletter Winter 2007 Edition


The Butler Group reports that 85 percent of all data stored is held in an unstructured format. Large volumes of such unstructured text exist in everything from patient records to e-mails to help desk reports. This data represents a potential gold mine of insight awaiting those with the skills to extract information, explore new ideas or detect trends. Text mining applications are growing, and this newsletter will focus on how academics around the word are using SAS Text Miner to evaluate open source software success, identify trends in higher education, provide students with new skills and even compare applies to oranges. For readers who want more information on SAS' offerings, SAS provides an overview of the product, a fact sheet and what's new in SAS Text Miner 3.1 here. Finally, read how Dr. Pat Cerrito's research into text-based medical records hit close to home. As always, we welcome your questions and comments at aigeditor@sas.com.


Back to Top


Text Mining Advances Educational Research and Undergraduate Learning

When Dr. Tasha Inniss of Spelman College in Atlanta, Georgia tells her undergraduate students to dig in, she means it. Her students dig into data and learn the basics of data and text mining for a series of research projects that will be presented at Spelman's Annual Research Day in April 2007. The lessons learned will become the foundation of a data mining course. Additionally, Inniss plans to offer a six-week summer research program in introductory data and text mining for select Spelman sophomore mathematics majors. In 2006, Inniss received a SAS course development software mini-grant, which provided SAS Enterprise Miner and SAS Text Miner for six months to develop these learning activities for her students.

Inniss has also used text mining techniques in her research. With a National Science Foundation (NSF) Small Grant for Exploratory Research, she and educational researcher Dr. Karen King of New York University are discovering key concepts or opinions in student interview data. Their paper has been accepted to the American Education Research Association Annual Meeting to be held April 2007. They hope to expand their research by mining tutoring sessions to ascertain whether a particular mathematical concept is being learned over time and to explore the use of concept links to address an NSF focus area called "visual analytics."

Back to Top

Using SAS® Text Miner to Analyze Open Source Software Development

Given the increasing corporate use of open source software development, there is a need for evaluation models of these projects. Dr. Uzma Raja has taken on this task. As a 2006 SAS Student Ambassador from Texas A&M University Raja's work won Best Contributed Paper in the Data Mining track at SAS Users Group International in San Francisco. In her paper , she used logistic regression, decision trees and neural networks to investigate success in open source software development. In August Raja became an Assistant Professor of Management Information Systems at the Culverhouse College of Commerce and Business Administration at the University of Alabama. In October, she presented a paper titled "Exploring Open Source Software Development and Maintenance Using Data Mining and Text Mining" at M2006 as an invited speaker.

Back to Top

Students Use SAS® Text Miner to Analyze French Education Market

EISTI, a French engineering school, offers students hands-on experience in business intelligence architecture as well as a dual-degree with Grenoble Ecole de Management. Students study the entire SAS business intelligence platform, including ETL, data warehousing, OLAP, business reporting, statistics and data mining, and then they apply their learning to business cases. A recent example analyzed French higher education, which is undergoing change to facilitate integration with the broader European system.

Students used text mining techniques to analyze existing postgraduate programs' curricula, which were posted on a higher-education Web site. Applying the %tmfilter SAS macro, they reviewed all the links on a specified page and all the links from the linked pages, drilling down across three levels. The result was a large, free database of programs in management, marketing, finance, computer science and economics. The programs clustered into two groups, not by the type of skills taught but by the maturity of the curriculum. New programs describe the content of the courses: topics, percentage of lecturers from industry and expertise. More mature programs focus on positions attained by former students after graduation: the type of job, the recruiting companies and salary.

Back to Top

Applying Text Mining in Bioinformatics

Dr. Tasha Inniss from Spelman College in Atlanta, Georgia, led a six-person team in using SAS Text Miner to determine a common vocabulary for describing age-related macular degeneration. The team included fellow professors from Spelman, the University of Iowa and the University of Chicago, as well as colleagues from industry. Their paper was accepted to the First International Workshop on Text Mining in Bioinformatics, published in the conference proceedings and presented in November 2006.
Read More

Back to Top

Students and Professors Put SAS® Text Miner to Work

At the Leeds School of Business at the University of Colorado at Boulder, SAS Text Miner is in use for a number of teaching and research projects. All undergraduates in operations and information systems learn SAS Text Miner, along with SAS Enterprise Miner, and then apply their knowledge to in-class projects done for industry partners. For example, last year the class delivered analysis and recommendations to Wild Oats Marketplace, Lifepics, and Spectralink.

SAS Text Miner is also used for research projects as diverse as understanding the relationship between newspaper coverage and executive compensation in S&P 1500 businesses to mapping the structure of business school research across nine academic disciplines (e.g. accounting, marketing, finance, information systems). Professors have also used SAS Text Miner to develop a novel approach to conducting literature reviews. And the Leeds team is collaborating to improve approaches to naming text clusters by comparing the SAS approach to two approaches from academic literature as well as two approaches developed at Leeds.

Back to Top

Text Mining Reveals Hidden Value and Competitive Advantages

I remember my fourth-grade math teacher telling me that it's impossible to add apples and oranges. She was wrong.

Now there's a way to throw apples and oranges into an equation without making a mess. The newest text mining technology enables decision makers to crunch words and numbers at the same time, yielding practical knowledge that can be leveraged for business growth. An article from the Dec. 6, 2006 edition of DM Review by SAS' Mary Crissey explains.
Read More

Back to Top

Report from SAS Asia Pacific Academic User Conference

This year's conference was held Oct. 5 at the Saroj Sadan, Alkesh Dinesh Mody Institute of Financial and Management Studies at the Indian University of Mumbai on its Kalina Campus. The event began with the traditional lighting of the lamp by Professor A.D. Sawant, Pro-Vice Chancellor of the host, the University of Mumbai, and a welcome address from Mr. Sudipta K. Sen, the CEO and Managing Director of SAS India. Keynote speakers for the opening session were Dr. Deepak B. Phatak, Head of the Kanwal Rekhi School of Information Technology at the Indian Institute of Technology; Professor Subash Sarnikar, Chairman of the ICFAI Group; and Dr. M. Rammohan Rao, Dean of the Indian School of Business, who shared their insights on innovation in education and the importance of preparing students with practical skills to be competitive in the business world. Mr. Prasar Sharma, Chief Manager of the Technology Management Group for ICICI Bank, spoke about the corporate demand for SAS skills. Sun Microsystems was the event sponsor. Additional information about the conference and abstracts are available.
Read More

Back to Top

New Statistics Certificate Program at Kennesaw State University

FORTUNE 500 companies will take notice when they see Kennesaw State University listed on a resume, thanks to a new certificate: Statistical Methods Using SAS. Located outside of Atlanta, Georgia, the university requires students to complete 12 hours of courses that offer SAS training to qualify for the certificate. "We chose to train our students using the SAS program because it is the most widely recognized leading statistical software available," said Victor Kane, Chair of the Department of Mathematics and Statistics. Professors in the department of mathematics and statistics view the new certificate as an opportunity for their students to differentiate themselves from all other graduates and gain a cutting-edge advantage in the marketplace through hands-on training.
Read More

Back to Top

New African Business Intelligence Hub Established

A new partnership will address the growing demand of African companies for business intelligence (BI) specialists. This effort builds on two existing partnerships between SAS and the Grenoble Ecole de Management (GEM) and GEM's own partnership with ESCA, a business school in Morocco with a strong record of attracting foreign students.

This BI hub has three focus areas: training, with the November 2007 launch of a postgraduate program in BI, offered by Grenoble; research, through applied projects carried out by researchers and PhD students with SAS support; and knowledge dissemination, with a BI Training School for professors and executive managers to be launched July 2007.

The partnership was signed on Nov. 30, 2006 and preceded by a press event with 15 Moroccan and French journalists (see CNN article) and a business intelligence conference attended by 200 participants, including Thami Ghorfi, President of ESCA; Jean-François Fiorina, Dean of ESC Grenoble (one of the schools of Grenoble Ecole de Management); and Ariane Sioufi, Academic Director of SAS France. Local SAS customers Meditel, a telecommunications company, and the Ministère de l'Economie et des Finances also joined the conference to share their experience.
Read More

Back to Top

Forthcoming Book from SAS® Press to Include Chapter on Text Mining

In his new book titled CRM Segmentation and Clustering Using SAS® Enterprise MinerTM, Randy Collica uses SAS Enterprise Miner and other commonly available techniques for customer relationship management (CRM). The book includes a chapter on segmenting textual data. Examples clearly demonstrate the concepts of segmentation and clustering in the context of CRM, and several chapters also contain additional exercises. The text is also well-suited for use in a business data mining analytics course in a college MBA program or course of study. The anticipated publication date is March 2007.
Read More

Back to Top

Data Mining for Educators To Be Offered in Three Locations in 2007

SAS' Higher Education Consulting Group will hold its sixth annual Summer Institutes in Data Mining in 2007. The program consists of SAS Enterprise Miner training and discussions about using data mining in the university environment. The training is free of charge; however, attendees are responsible for providing their own transportation, meals (other than lunch), and lodging. This summer three sessions will be held, including the first outside of the US. They will be at SAS world headquarters in Cary, North Carolina (Aug. 5-10), California State University at Long Beach (July 15-20), and the Universidad Anáhuac in México (May 21-26). A new course (Applied Analytics Using SAS® Enterprise MinerTM 5) will be offered in all three locations, and both US locations will include an advanced track. Attendance is by invitation only, and all dates are tentative. If you are interested in learning more, please send mail to Susan.Walsh@sas.com.

Back to Top

Don't miss important updates from SAS! Please add sas.com as a domain in your safe sender list.

SAS and all other SAS Institute Inc. product or service names are registered trademarks or trademarks of SAS Institute Inc. in the USA and other countries. ® indicates USA registration. Other brand and product names are trademarks of their respective companies.
Copyright © 2007 SAS Institute Inc. All rights reserved.
SAS Institute GmbH, P.O. Box 10 53 40, Neuenheimer Landstr. 28-30, D-69043 Heidelberg, Germany