Skip navigation

Data mining technique developed at Northumbria helps call centre improve efficiency

Valley Care is a telecare service with more than 5,000 vulnerable and elderly customers and receives around 130,000 calls every year. Its aim is to assist people to live independently in their own homes.

Analysing data and extracting trends from the calls it receives is a huge challenge because of the amount of high-dimensional data. But using a data mining technique developed at Northumbria University, researchers were able to help the call centre to improve the quality and efficiency of its service whilst operating in the same budget.

Called targeted projection pursuit (TPP), the technique enabled Professor Maia Angelova and her colleagues to classify and visualise large data sets without losing information. The technique had already been applied to a variety of problems including gene expression data for classification of leukemic cancers and identifying business trends from Companies House data, but this is the first time this technique had been applied to call centre data.

The research enabled Valley Care to provide more efficient workload planning for call centre operators; provide a more efficient allocation of warden visits; prioritise calls to ambulance services and relatives; and eliminate false alarms.

It also identified a number of users at a higher risk of falls; seasonal peaks in the use of the service; and the frequency and type of calls.

“The old system used by Valley Care could neither quantitatively nor automatically prioritise calls and so our research has transformed the operational efficiency of the call centre” said Angelova. “For example, being able to better allocate warden visits and eliminate false alarms has made a big difference.”

Case studies

Latest News and Features

Northumbria University launches series of events to help businesses retain top talent
Prof Katie Jenkins
More news
More events

Upcoming events

Rewriting The Rules - Unlocking People Potential
Roma Agrawal Lecture - 30 March 23

Back to top