BME are a re-seller of QlickView by QlikTec. They claim to have sold 24 of the 30 licenses held by UK Universities. I’m attending their Higher Education User Group today to hear what’s possible, what people are using the product for, what insight and edge it’s giving.
Full programme here
BME Solutions sell more than QlikView; Quantrix (forecasting and modelling), F9 (financial reporting and consolidation), XM Pro (business process management) and associated consulting services to ‘pretty up your models, run training courses, upgrade client hardware, help out with planning).
QlikView allows quickly customisable visualisations of data sets held in existing information systems. The active users I have spoken to are, not surprising given they made time to attend, positive about the product. There are also newcomers here as well as people considering a purchase. We’re being shown as series of active visualisations delivered by client or browser. You can view plenty of these at http://www.he.bmesol.com
BMI sell a product to allow visualisations based on geographical data layers and using QlikTec. The product is GeoQlik and there details at http://demo.geoqlik.com/
Back in 2010 Jisc funded a BI project at UEL. It’s one of the case studies in our BI InfoKit. The work has attracted a good deal of sustainability funding from other sources. It’s based on QlikView and the lead, Gary Tindell presented a demonstration of this work. It’s based on data UEL purchase from HESA and loaded into QlikView allowing competitor analysis and identification of new markets for courses and areas for targeting Widening Participation. The insight gathered was a lack of STEM subject provision in the local Borough of Newham. The action taken to mitigate this was the development of a brand new (Newham Collegiate) Sixth Form College. UEL have since been commissioned to provide similar insight for all 33 London Burroughs but expanded to include student attainment data from HESA. UEL offer this as a chargeable service, so generate income from their investment in data and interpretations.
Number of students in HE by Borough, for example Newham sent 12K students into HE, we see the courses they took, 60% stayed in the local area, UEL was the most popular, it’s possible to adjust the parameters by ethnicity, age range, gender, award type, degree classification, destination of leavers (occupational classification). On and on it goes. There’s a lot of actionable insight here. Of course, the trick is prioritising what to action and how to do it.
As well as HESA data (and bought in HESA data) there are opportunities to identify actionable insights based on institutionally owned data and we saw a model with over 10 million records over 8 years, so big data in terms of volume, but not velocity and variety as it’s archived and structured.
Bath Spa have developed a product called QV Source which is for sale. It is a connector to the HESA data sets using the Heidi API. Heidi is the system that HESA provides to all Universities to collect and distribute their data sets. Out of the box Heidi allows XLS or XML downloads of data for client manipulation. The overhead of creating reports for this is considerable. QV Source connects Heidi data to QlikView as a live lookup obviating the need to manually manipulate Heidi provided data. Rather the data is manipulated via the QlikView client. Presumably standard dashboards can be provided as a service to end users.
Another demonstration here by a University client showed us visualisations of research income performance. We at jisc looked at this through the Liverpool, Glasgow and Huddersfield Business Intelligence projects and these are written up in our BI InfoKit.
It’s clear that QlikView can be used for administrative support to run a university, what was less well understood (by me at least) was its potential to support research and teaching.
I saw a demo of Quantrix at the end of the day so we’re off topic with respect to QlikView and visualising data from disparate sources. Here we have an analytical modelling tool to forecast future student intake. The tool extends dimensionality of spreadsheets from 2 to as many as required in the model.
A few pearls of wisdom;
People are the main problem in BI projects, not technology or data
Focus on the questions you want answers to, or you’ll drown in possibilities
Provide appropriate visualisations to each staff member, the more senior the staff, the simpler the dashboards
QlikView is a reporting tool, not a data source
Visualisations highlight data inaccuracies that users will fix, don’t try to clean all the data before you launch one, you’ll never get started