I didn’t make it to IBM’s Information-On-Demand (IOD) event in Las Vegas this year, but that’s not to say I didn’t keep up with the proceedings. Thanks to an active social media sphere and some interesting press releases, blogs and tweets there was plenty to keep me engaged from the other side of the pond. One announcement that did catch my eye is Project Neo, a cloud-based data visualisation and discovery tool that has been incubating in IBM Labs for a little while now. This was particularly pertinent since I’ve recently been looking at the role of visualisation in helping improve the interpretation of data.
Project Neo is IBM’s answer to data visualisation and discovery for business users. It promises to help those who don’t possess specialist skills or training in analytics, to visually interact with their data and surface interesting trends and patterns by using a more simplistic dashboard interface that helps and guides users in the analysis process. Whereas previous tool incarnations are often predisposed to using data models, scripting or require knowledge of a query language, Project Neo takes a different tack. It aims to bypass this approach by enabling users to ask questions in plain English against a raw dataset (including CSV or Excel files) and return results in the form of interactive visualisations.
For example, a marketing manager can ask questions and discover through visualisations and guided analytics what is causing sales to fall in a particular quarter. Behind the scenes Neo is semantically analysing data and applying advanced analytics to it to determine what concepts exist in the data and return results in the form of analysis, visualisations and plain English explanations of what was discovered. Likewise it also highlights insights within the data and suggests alternate questions that you may want to ask.
Although I haven’t seem a demo of the product yet I understand from reports that under the covers its uses DB2 BLU for in-memory columnar data storage and querying as well as utilising SPSS Analytic Catalyst for a scoring and predictive engine, and has Rapidly Adaptive Visualization Engine (RAVE) as a basis for its visualisation layer. RAVE is also the underlying technology for Many Eyes, IBM’s visualisation online community for experts and practitioners.
Of course IBM isn’t the only vendor banking on discovery and visualisation to help lower the skills threshold for data analysis. SAP’s Business Objects Explorer, for example, is a keyword and natural language based search query interface designed to provide an easy and simple way for business users to explore and analyse information by using an Internet search paradigm. That said, I think that Neo is much more likely to be a competitive move against vendors such as QlikTech and Tableau who have taken a lead in this market. Part of their proposition (quite rightly) focuses on the user analysis experience and the visual display and communication of relationships in the data through interactive visualisations. It’s been a successful market for these vendors and IBM clearly wants a bigger slice of the pie.
IBM has many of the technology ingredients needed to make this a success; the use of in-memory, natural language processing and visualisations provides a powerful combination for helping users better understand their data, but the addition of advanced analytics (such a predictive analytics) also brings the promise of a deeper and more forward-looking level of insight too.
Neo looks like a very interesting project for IBM. If anything it represents yet another facet to Big Blue’s work towards addressing the analytics skills gap—as I’ve written about before here. Having said that, I believe this initiative, like others, needs to continue to address some of the other stumbling blocks to bringing analytics to a wider business audience. Very broadly these concern ensuring users don’t get themselves into analytic difficulties and end up with misleading or erroneous results, and secondly, helping users align and use the insights found within their data. These are complex issues to deal with but nonetheless will be important factors in helping business users become more data and analytics literate.
Organisations that want to get their hands on Neo are in for a little wait. The project is in the Labs with its beta program scheduled to start in early 2014.