First Look – Angoss 8.0

I have previously written about Angoss and its products (most recently this post on their 7.5 release) and recently got an update. As a company Angoss is focused on delivering data mining and predictive analytics to improve performance in sales, marketing and risk. They have a set of products for desktop, client/server and cloud-based and a focus on accessibility and usability. They have over 300 corporate accounts globally, across sales/marketing as well as risk. Risk analytics is their historic strength, as it is for most analytics firms, but sales and marketing are an increasing focus going forward. Their products are designed to have an easier on-ramp, a rapid learning curve and are designed so as not to rely on programming skills in users. A typical customer is using Angoss’ tools alongside other data miming and predictive analytic tools.

Directionally Angoss has historically sat in the middle of the spectrum between simple reporting and dashboards (BI) and the kind of advanced predictive analytic that require coding. Their KnowledgeSEEKER product is being extended from this middle ground to close the gap with typical BI users so they can add some predictive analytics to their BI environment. In particular they recently launched their KnowledgeEXCELERATOR product to fill this gap. Meanwhile their KnowledgeSTUDIO product is broadening its support for more advanced analytics to reduce the need to use other tools for advanced predictive analytics. Their current product set includes:

• KnowledgeSEEKER
  The core data mining /predictive analytics product focused on trees – data preparation, data exploration, and profiling, decision trees, strategy trees and model evaluation. Text Analytics is available as an add-on
• KnowledgeSTUDIO
  Aimed at an advanced user, this has all the features of KnowledgeSEEKER as well as  a number of more high-end predictive analytics capabilities like market basket analysis, cluster analysis,  multi-layer neural networks, principal component analysis, scorecard, linear and logistic regression. Text Analytics is available as an add-on.
• KnowledgeEXCELERATOR
  A new product released in September 2012 for visual data discovery and prediction for business analysts based on a guided workflow for discovery, prediction and visualization that integrates decision trees. A sneak peek of the product as well as a trial version of the software can be found at www.whatsinyourdata.com.
• KnowledgeCLOUD
  Cloud-based platform that hosts everything in the cloud and supports both a general platform and industry solutions in claims, mutual funds management and CRM lead and opportunity scoring for instance.

KnowledgeSEEKER has support for a lot of common data preparation and profiling as well as standard segmentation and model evaluation capabilities. The StrategyBUILDER component allows these models to be extended into Strategy Trees. Which essentially combine explicit business rules with the analytic decision trees as well as provides the user with the ability to validate the success of the implementation of those business rules via the validate strategy option. Both KnowledgeSTUDIO and KnowledgeSEEKER extend all this with optional modules for in-database analytics and text analytics (by embedding the Salience engine by Lexalytics). While KnowledgeSTUDIO provides the user with unsupervised learning and advanced modeling techniques. Recent updates include:

• 7.6 enhanced data visualization with various mapping and charting improvements while extending In-database support to add SQL Server and Netezza to the original support for Teradata.
• 8.0 improved data preparation, added text analytics and extended in-database support to include Oracle.

The big themes since my last update are cloud, support for Big Data and text analytics:

• The cloud solution is one recent area with significant new capabilities for Angoss. The KnowledgeHUB platform at the heart of the cloud solution brings data from a wide variety of data sources into a cloud-based analytical data mart. The standard Angoss tools are then used to produce models against this data. Output can be as scored lists, an API for live scoring (allowing any model developed with Angoss, including a Strategy Tree, to be deployed as a web service), reports or through integration with Salesforce.com. The KnowledgeCLOUD options are fully managed cloud-based solutions with analytical development done by Angoss. Angoss also allows customers to buy On-Demand software that allows desktop-client access to cloud-based data on a monthly subscription basis for its KnowledgeSEEKER, KnowledgeSTUDIO and Text Analytics software products.
• Support for Big Data in Angoss comes from support for in-database analytics (to handle larger volumes within enterprise data warehouses) as well as text analytics (to handle a wider variety of data sources including text-based, unstructured data) and 64 bit in-memory support for improved scale. Data can be imported from Hadoop and models can be deployed back to Hadoop. Angoss’ in-database analytics turns common elements of the data processing in model development into SQL that can be executed in-database. This improves performance by moving less data and improves integrity by keeping the data in its original source. The resulting models can also be generated as SQL functions, allowing in-database scoring also.
• Text analytics has been added by embedding Lexalytics’ Salience text mining and sentiment analysis engine seamlessly into the product. This handles sentiment analysis and entity / theme extraction for instance and is very focused on handling text written by amateurs or casual text. Integration of this allows Angoss users to generate datasets from unstructured data sources for data mining, visualize relationships, do segmentation based on document attributes and integrate this data with other data in more advanced analytics.

Up next is the 8.5 release later this year with a number of enhancements specifically around improved visualization, scorecard development (including reject inference calculations) and the ability to bring R datasets into the products directly.

You can get more information on Angoss and its products by visiting www.angoss.com; and Angoss is one of the vendors in our Decision Management Systems Platform Technology Report.