Improving the Data Analytics Process

Improving the Data Analytics Process
Dates: 18-21 July
Location: London
Organisers: Ian Horrocks (Oxford), Chris Williams (Edinburgh)
The task of going from data to actionable knowledge involves many stages and processes, with feedback cycles. It is widely cited that the data preparation/wrangling process takes up around 80% of a typical data mining project, dealing with steps like understanding what data is available (or collecting it), data integration, handling missing, messy and anomalous data, extraction of features etc. These steps are often carried out in a hand-crafted/ad-hoc fashion, and have had much less attention paid to them than other parts of the process (e.g. predictive modelling). This workshop will look at the process end-to-end and identify key aspects that need improvement or better integration in order to maximise the benefit that can be obtained from the data. We will choose two problem areas with unrestricted data, and consider the whole process of going from raw data to knowledge. Of crucial importance is the availability of tools to deal with the wrangling and analytics problems. We envisage tutorials on the various stages of the data analytics process, along with practical work on using these on the exemplar problems.

Attendance is by invite only.