Modeling Family

Previously PASW Association

To find hidden links in your data

IBM® SPSS® Association can help you identify patterns and links between products, actions and behaviour; enabling you to use information about association and order to predict future events. Once you identify these links, patterns and sequences, you can use them to improve decision making in a wide variety of applications.

IBM® SPSS® Association enables you to:

• Mine for associations among items in a large database and identify behaviour associated with a particular outcome
• Find useful associations that occur less frequently
• Discover order-sensitive “sequential associations”

Features

Apriori
CARMA
Sequential association algorithm

Formerly PASW Classification

For better prediction

IBM® SPSS® Classification, extends the capabilities of IBM® SPSS® Modeler by providing best-of-breed algorithms to create better classification and regression models, combined with automation to help you find the best model more easily.  Modelling techniques include machine learning, decision trees, subgroup identification, statistical methods and multiple model generation.

Using these algorithms, you can build models to help you:

• Improve customer loyalty - Predict which customers are likely to churn, allowing enough time to implement retention strategies
• Prevent fraud, waste and abuse - Detect suspicious patterns in transactional data or claims data and flag them for further investigation
• Lower credit risk - Profile risk groups based on customer information, such as account activity, and offer the right credit line to the right applicants based on risk group
• Better predict customer buying habits - Predict whether or not a customer will purchase a particular product or service and target campaigns at customers that will most likely respond to them
• Make better medical decisions - Improve the accuracy of diagnoses and predictions and prescribe more effective treatments

Features

Nearest Neighbor
Binary classifier and numeric predictor
Self-learning response model
Time–series forecast models
C5.0 decision tree
Neural Networks
Support Vector Machines
Bayesian Networks
Cox Regression
Binomial and multinomial logistic regression
Discriminate analysis
General linear models (GLM)

Previously PASW Segmentation

To target the most valuable relationships

Expand your data mining capabilities with IBM® SPSS® Segmentation, which helps you cluster data into similar groups and detect unusual or anomalous cases. Using this information you can improve market segmentation, gain customer or constituent insight and detect fraud, waste and abuse.

IBM® SPSS® Segmentation enables you to:

• Detect anomalous and suspicious occurrences and/or constituents and prioritise them for the auditing process
• Identify distinct groups among a customer base allowing precise targeting of sales efforts
• Discover similarities in customer behaviour, create new segment variables for further analysis

Features

Auto Cluster
Kohonen Network
Two-Step Clustering
Anomaly Detection

Previously PASW Modeler Solution Publisher

Make better decisions through cost effective deployment

Data mining solutions can be integrated into every point of your business strategy to allow optimal organisational change with measurable results.

IBM® SPSS® Modeler Solution Publisher enables you to:

• Aid decision-making with efficient deployment of data mining solutions
• Update data mining solutions quickly and cost-effectively

Previously PASW Text Analytics

Unlock the power of text data

Up to 80 percent of your organisations data is contained in an unstructured format, including customer emails, call center notes, open-ended survey responses, Web forms and other textual sources. This means that data mining projects focusing only on structured data may use less than 20 percent of the information available. IBM® SPSS® Modeler Premium enables you to combine this valuable unstructured data with traditional structured data to provide a 360° view of your customers or constituents, adding more predictive power to your models.

Examples of IBM® SPSS® Modeler Premium applications include:

• Product development and refinement - Identify trends in customer complaints to improve existing products and services and develop tailored offerings
• Improve campaign effectiveness - Analyse data to understand your customers’ needs, enabling the implementation of targeted marketing campaigns
• Cross-selling - Improve sales by using information about customer preferences to better target products to existing customers
• Churn prevention - Discover why some customers leave and take steps to prevent defection by analysing customer communication records for recurring problems or complaints that precede churn
• Market intelligence - Survey competitor and industry Web sites, RSS feeds and other publicly available textual information to maintain a current view of your competitors
• Security - Discover potentially suspect behaviour by analysing website content, chat rooms, e-mails, blogs and other sources of information to identify patterns and associations in the data
• Fraud detection - Identify potentially fraudulent patterns in banking transactions or insurance claims and use investigative resources more effectively

Features

Linguistic extraction
Web feed node
Text analytics modeling node
Text link analysis
Resource Editor