Audience Segmentation: Data-Science, Big-Data Architecture & Solution
Submitted by prabhakar srinivasan (@prabhacar7) on Wednesday, 5 June 2013
Analytics and Visualization
The objective of this talk is to show how analytics techniques can be used to answer fundamental questions in the area of Audience Segmentation for the Broadcast domain. Various stakeholders like planners, channels, operators, media agencies, advertisers are very keen to know ‘who is watching TV’ so that the products, services, content and advertisements can be better customized and tailored to the needs and interests of the viewers. We will cover the subject of audience measurement, data-science aspects of audience segmentation and the BigData tools, technologies and architecture used to address the scalability considerations. Big-data Architects, Big-data development and infrastructure engineers, Data science specialists, Product Managers, Business partners in Broadcasting all have some useful knowledge to take away from this presentation
How do you determine the number of people in a household from a set of anonymous viewing activity? How do you find out attributes like the age, gender, etc., of various people who are watching TV in a household? How do you validate the audience segments? These and many more such interesting challenges are in the scope of audience segmentation and we describe the means and technologies to answer these questions. After this presentation attendees will be better informed about the challenges of analysing anonymous viewing data collected from households and determining audience segments automatically. The data science and the technology used to achieve this goal make for a compelling story to be shared with the larger BigData community.
I am a member of the New Initiatives team which is part of Cisco’s SPVTG (Service Providers Video Technology Group). I work full-time for the Corona project which is a team focussed on BigData and Data-science challenges. With a specialization in Broadcasting, Internet technologies and Analytics, I have the unique opportunity and knowledge to deal with challenges specific to the Broadcasting domain, which are also BigData problems and the solutions to these problems which are currently being developed and deployed