Subscribe to DSC Newsletter

The next Internet revolution: news feed aggregation and optimization (it will create new types of analytic jobs)

Why did Twitter receive so much VC funding (more than $100 million), yet they barely make any money.

I believe that Twitter could be the new Google and the new source of information, in 5 to 10 years. Twitter could become the largest repository of news and content. The feature introduced this month - list management - is an extremely powerful mechanism to automatically generate targeted content in real time, and publish it on your blog, website, newsletter, network, etc. A very powerful SEO tool, and traffic generation mechanism.

We have created two Twitter lists: one for jobs, and one for general news. For now, our AnalyticBridge real time analytic news (see twitter.com/analyticbridge/analytics, see also the scrollable RT feed box at http://www.analyticbridge.com/group/miningterabytesofdata) is made of

  • Feeds from 80 Twitter blogs
  • Feeds from my own Twitter blog

Some of these blogs (at least mine, but I believe another one as well) have two feed sources

  • content produced internally
  • content produced externally: in my case: featured AB posts, featured AB jobs, featured AB conferences, and Google news feeds for a number of pre-selected keywords (all these sources - external and internal, from everyone, are blended together in one single feed, and then displayed on our Data Mining group, but some other AB groups have customized, partial feeds only)

How will this create new jobs?

News feed optimization will require text mining experts to work on the following tasks:

  • Reporting - identify news that have a high CTR (click per impression), based on position
  • Analyze - identify metrics (KPI's) that are correlated with higher CTR (e.g. keywords found in the news title, number of words in the title, association or lack of association between the keywords found in the news title and the page category, user segment)
  • Model - the CTR is the fundamental response for these predictive models, but sign up rate is also critical
  • Optimize - On any landing page, select and rank 20 news titles out of 5,000 from your news feed, and display these 20 titles in the right order to achieve optimal response. Two different users (on the same page) might be served two different set of news. This will require sophisticated scoring algorithms.

Views: 224

Comment

You need to be a member of AnalyticBridge to add comments!

Join AnalyticBridge

On Data Science Central

© 2019   AnalyticBridge.com is a subsidiary and dedicated channel of Data Science Central LLC   Powered by

Badges  |  Report an Issue  |  Privacy Policy  |  Terms of Service