The Empathy Project is about making the online experience more empathetic and human.

It is an innovative project researching ways to mine social media data to allow us to better understand our customers and anticipate their needs.

The EmpathyAPI pulls data from Twitter within a defined geographic area. How this is applied will depend on the specific demands of the industry sector, but always with the intention of increasing customer insight or predicting future demand based on social media notoriety.

The ClusterSphere is the latest application of this technology.

This gathers tweets into subject-related clusters to present an interactive graphical view of what people are talking about. It is updated in near real-time.

The example below shows the clusters in the Bilbao region of Spain. The block near the top are tweets about the internet and technology, at the bottom is sport, Bilbao's football team Athletic de Bilbao (to the left) and politics to the right.

Zooming in on the block at the bottom, we can break down each cluster to understand exactly what is being talked about. We can zoom in as far as the keywords of the individual tweets.

We can also add in authors and can select individual tweeters. The graphic below shows one organisatino whose tweets cluster over to the right, away from where most of the people are talking. Once developed in further detail (focussing only on those tweets from the relevant audience) this sort of analysis can be useful in driving communication and content marketing strategies.

Once the specific audience is highlighted, this will show the areas which are of most interest to the people you are trying to engage with. For example, people tweeting in the clusters you identify as your core cluster may also be clustering elsewhere. This should influence how you communicate and what content you develop.

This functionality is still under development, but shows great promise as a tool for managing social media marketing and building customer insight.