Innovative predictive analytics techniques
IBM is using the latest predictive analytics techniques—among them geospatial, text and entity analytics—to help organizations along on the journey to big insights. For a quick overview of geospatial analytics, including what it is and how it is being used, watch this one-minute video.
As described in the video, geospatial analytics uses geospatial data: longitude and latitude. It’s a good fit, then, for law enforcement, which uses geospatial analytics to predict where and when crimes will occur. However, IBM geospatial analytics is also proving valuable in the healthcare, retail, law enforcement and nonprofit sectors, among other sectors. Geospatial analytics is emerging as an important source of insight in both traditional and big data analytics—hardly surprising considering that IBM was among the first to integrate geospatial capabilities with predictive analytics.
Then there’s text analytics, which looks at unstructured data, information not organized in a predefined manner. Unstructured data is often text-heavy, and it can contain dates and numbers as well. Text analytics takes unstructured data and categorizes it. IBM software can find the value in unstructured data, identifying the sentiment and meaning behind it and allowing it to be used to make predictions. For example, suppose a mobile phone service provider wants to know how—and whether—it can retain a customer who suffered five dropped calls in the same area. Mining text data, then categorizing and analyzing it and using it for predictions, can improve customer service.
Entity analytics looks at the many different ways of referring to an entity, be it a person, a company, a product, a thing or an event. For example, my name is Armand Ruiz. I can be referred to as Armand, Ruiz, A, Armand Ruiz, or even by my rank. With so many combinations of names, ranks and locations applicable to me, I could be identified many different ways in a single database. But entity analytics looks at all my data as being associated with one person, not many.
Yet IBM offers even more innovative analytics techniques than these. For example, IBM SPSS Modeler can help you build and deploy predictive models directly into your business processes—and its premium version can make use of advanced capabilities, including entity analytics and text analytics. Moreover, the active and growing IBM open-source community offers resources to help extend predictive analytics software, among them blog posts, videos, tutorials and a library of 6,000 predictive extensions. Notably, such extensions, or add-ons, help users expand the capabilities of IBM predictive analytics using popular programming languages such as R, Python and Java.
Learn more by contacting an IBM expert or visiting IBM online to get predictive analytics working for you.