Text Mining & Analytics – Correlating Social Intelligence with Traditional Data

March 25, 2011
295 Views

We talk a lot about analyzing social media conversations to surface consumers insights, considerations and preferences.  But what we don’t comment much on is that the same technology we use for social data can also be used to analyze an organization’s internal, private data. This is important especially if your organization is working towards augmenting existing consumer data with social insights.

We talk a lot about analyzing social media conversations to surface consumers insights, considerations and preferences.  But what we don’t comment much on is that the same technology we use for social data can also be used to analyze an organization’s internal, private data. This is important especially if your organization is working towards augmenting existing consumer data with social insights.

Why would analyzing private data be important?

Let me give an example and it’s one that I have been thinking about for some time. I’ve always entertained this idea of applying our semantic technology to something like the WHO data sets. Last year, the WHO had a contest that allowed you to analyze a group of data sets and the winner was selected based on the type of results they surfaced during the analysis. I think that would have been fascinating but unfortunately I happened upon the contest after the fact. But it got me thinking. What if you gave each nurse or doctor access to an application that allowed them to enter details about a patient’s symptoms and profile details? This information could then be loaded into a data management system, where a semantic engine (like ours) could be used to identify emerging trends, like flu or malaria. Demographic information could identify which audience segment was most likely to be at risk and the percentage of each group currently infected. Geo-specific information could highlight areas of high and low infection rates and perhaps map the spread of the contagion.

How much more prepared could a region be for an outbreak if agencies had access to this information? The greatest impact is that patient details could be analyzed as often as required, which means responses can be made based on relevant and timely information.

Nice Example of Text Mining but I Sell Running Shoes

What’s nice about our approach to analytics and text mining is that our technology can analyze social assets for consumer perception and intention regarding brands, products and categories. This social intelligence can then be mapped back to your organization’s existing consumer records. This social media analytics feed can be integrated into your existing data management systems to provide a real-time, always on, unified view into your consumer’s behavior and can provide ongoing feedback of your organization’s campaign or outreach efforts.  We strongly believe that social is another data point –  critical, real-time, true voice of customer – and the data should be available across the organization, integrated with your other traditional data to show correlations, and work with your existing data systems.

Social Data Mapping Into an Organization’s CRM

So, how might this mapping work? Our semantic technology is able to surface social intelligence about an audience or an author from social media conversations. Below is a selection of a post we call a snippet, which is  a portion of a social media conversation, which includes only content relevant to key terms we have defined.

Click image to enlarge

From a single snippet, we can identify an amazing amount of information:

  • Social handle
  • Approximate time of day
  • Location
  • Gender
  • Subjective

(All of this information is publicly available. We strongly believe and adhere to all privacy laws. )

This social profile of author chriskoffman can then be matched to an existing consumer within your organization’s CRM, transactional or outreach data system.  This integrated view of your social consumer can help inform 1 to 1 or 1 to many outreach efforts, campaigns and customer care initiatives.

I’ll be honest that this type of merging of social data with more traditional records can cause both organizations and individuals some concerns. Your organization will need to carefully navigate how quickly  to move towards targeted outreach or custom campaigns.  In some ways, your organization may need to rely more fully on social media to carefully listen to your customers to determine the level of engagement they prefer.