Today SPSS Inc. announced new product releases and new naming for its product families. Predictive Analytics Software (PASW) is the new umbrella and the four families are layered below this. The four families are Data Collection, Statistics, Modeling, and Deployment. This week’s enhancements are to the modeling family. In addition, Clementine becomes PASW Modeler 13 and the text mining product becomes PASW Text Analytics 13. More updates to the statistics, data collection and deployment families will come in the next few months. All these families are integrated from SPSS’s perspective so you can go from capturing customer feedback to understanding data and testing hypotheses to automating your predictions with modeling. All three families are tied to the deployment tools and are part of the way in which models are deployed into business processes.
SPSS wants to enable more users – and new users are more like business analysts. SPSS’ general experience has been with statistically deep data mining experts and now they want to empower new users with the modeling tools. As a result lots of automation has been added to the tools. Automated Data Preparation, for instance, helps bring new people into the process but also makes experts more productive. One of the features of this is to recommend sources (based on the amount of work needed to prepare them) as well as report on what it did to fix the data. Auto Cluster is another new feature that allows all three clustering algorithms to be run on a data set and then users can graphically compare details of the different clusters to see which is the most predictive. Business analysts can use this but it also allows experts to drill-down into the cluster details. The integration of the statistics product with the modeling tool has also been extended. Finally, collaboration is also enhanced, for instance with the ability to add comments into the modeling stream to help multiple people collaborate. This also helps experts document their choices for others in the modeling.
SPSS is continuing to use incremental model refresh, in-database mining and parallelism to ensure that performance is good. The ability to update the model with new data without reprocessing every item remains important. The tools now support Citrix and VMWare for virtualization and added support for HP Neoview, MySQL and Sybase.
The unstructured text analysis tools are also getting a refresh. More language support and more language processing techniques and pre-built Text Analysis Packages for categories and dictionaries. These handle issues like customer satisfaction and provide a ready-to-use set of assets that can be pointed at a text source to get some analysis without doing too much legwork. This would allow other experts to do this also so that they could make a specialist vocabulary (for, say, warranty claims notes) available.
The renaming is designed to focus people on what SPSS sees as the value proposition – the integration of the products across the capture-predict-act lifecycle and their focus on Predictive Analytics.