CATEGORY: Data Mining & Visualization

REVIEWER: Dr. Dursun Delen, assistant professor, Oklahoma State University, Department of Management Science and Information Systems.

BACKGROUND: Oklahoma State University is a well-known, comprehensive, land-grant university and research institution that focuses on people and opportunity.

PLATFORMS: Microsoft Windows NT, 2000 and XP environment.

PROBLEM SOLVED: Non-traditional students, working full-time in industry, bring a wide variety of classification and prediction problems to be solved with STATISTICA Data Miner's tools and techniques. Some of the recent problems include: the prediction of diabetic illnesses based on demographic, social and recreational parameters; target marketing models for better promotional mailing; prediction of financial indicators such as the S&P 500; and foreign exchange rates.

PRODUCT FUNCTIONALITY: STATISTICA Data Miner has provided everything we need for our data mining projects. The comprehensiveness, efficiency and accuracy of data mining algorithms in the product are impressive. Unfortunately, STATISTICA Data Miner does not have text mining capabilities. In addition, the process model interface for data mining projects is not as intuitive and as comprehensive as we would like it to be.

STRENGTHS: We have been impressed by a variety of strengths in STATISTICA Data Miner including the previously mentioned algorithms, the rich set of visualization tools and processing speed. As far as the algorithms are concerned, based on my years of experience in industry and in academia, I can confidently say that STATISTICA Data Miner has one of the most comprehensive data mining algorithms on the market, covering such application areas as prediction (e.g., regression), classification (e.g., classification and regression trees, artificial neural networks, discriminant analysis), clustering (e.g., k-Means, Konenan SOM) and association rule mining (e.g., apiori algorithm). In addition to one of the most efficiently implemented modeling algorithms, a rich set of data preprocessing and data visualization algorithms is also integrated into STATISTICA Data Miner. The product takes advantage of multiprocessor environments by facilitating parallel processing on the server site and, by doing so, dramatically increases the execution speed of the data mining projects. Moreover, the graphical tools and their outputs are phenomenal.

WEAKNESSES: Though it is clear that StatSoft has expended much effort to make the GUI as user-friendly as possible, some of the features are not very intuitive to the end user. For instance, the user is provided with only predicted values for a developed model. To obtain other diagnostic output, additional steps must be taken.

SELECTION CRITERIA: After looking into the leading data mining tools, we chose to use STATISTICA Data Miner because it had all of the data mining algorithms we were interested in with the exception of text mining; a graphical, interactive, user-friendly interface, which includes a process map where a complete project can be drawn using graphical icons and connection arcs and be run all together at once; and, most importantly, the fact that STATISTICA Data Miner can be launched as a Web application so that everybody can use it from within a Web browser without installing any client-side components. These features made STATISTICA Data Miner our premier choice to be used for teaching data mining to graduate IT students and to be used for our academic research endeavors. Lastly, despite its advantages over other tools, we found STATISTICA Data Miner less expensive than other comprehensive toolkits in the market.

DELIVERABLES: The tool provides all kinds of graphical and tabular outputs for a wide variety of data mining algorithms. Based on the selection made by the user in the project specification interface, each data mining algorithm can produce simple, enhanced and comprehensive output reports. STATISTICA Data Miner also provides users with a deployment feature by which the end user can easily turn the developed data mining models into production system.

VENDOR SUPPORT: We have received exceptional support from the technical personnel of StatSoft, Inc. They came to our site to install the software on our Web server and have answered all of our questions about the tool and its features. They have been very patient, courteous and helpful to us throughout the entire purchase process. Overall, it has been a great experience knowing and working with the data mining experts at StatSoft.

DOCUMENTATION: The documentation that comes with the product is very comprehensive, covering all of the aspects of the tool and the problem domain. The tool comes with step-by-step tutorial that helps new end users learn the product in a relatively short period of time.

Register or login for access to this item and much more

All Information Management content is archived after seven days.

Community members receive:
  • All recent and archived articles
  • Conference offers and updates
  • A full menu of enewsletter options
  • Web seminars, white papers, ebooks

Don't have an account? Register for Free Unlimited Access