Sampling, Network Graphs and Link Analysis come to PYLON

Kester Ford
19th April 2016 0 Comments

It is with great pleasure that we announce the latest batch of improvements to our PYLON for Facebook Topic Data platform. As of today, we are offering our partners three new ways of analyzing Facebook topic data, making it even easier for them to provide marketers with the unprecedented insights that Facebook topic data offers.


The first of these new features is sampling. Sampling is an analysis technique that allows the characteristics of a large dataset to be understood through study of a subset of that data. In PYLON for Facebook Topic Data it is a premium feature that allows analysis to be performed on a proportion of stories or engagements that meet your filtering criteria. Adding sampling to their account will allow DataSift partners to create advanced features such as industry baselines, gain insights on global trends and avoid loss of data quality in unpredictable datasets.

Topic Network Graphs

We are also making it much easier to produce network graphs from Facebook topic data. Network graphs are an effective way of visualizing the relationships between topics in your Facebook topic data index. We have introduced new analysis targets that make it easy for our partners to create these graphs in their products. They are used to explore the relationships between topics in an index by representing each topic as a node whose size is proportionate to the number of times the topic appears in the index and linking the nodes together based on how often they appear in posts together. Please explore our sample network graph of real topic data to get a feel for what is possible.

Link Title Analysis

At the end of last year we introduced the capability to filter on link titles. This means that posts that link to web pages that contained relevant terms can be included in analysis even if the terms are not mentioned in the post itself. Now we are adding the ability to perform frequency distribution analysis on link titles. This gives DataSift’s partners the capability of presenting the most popular web content in an index in a more easily readable form than a url and without the dilution of results caused by link shortening.

For detailed information on the new features, check out our developer blog. To find out how to get access to unprecedented insights from the world’s largest social platform visit our partner portal to find the best solution for you. Or if you would like to become a partner and deliver these insights to your customers, then please get in touch.

Kester Ford

Written by Kester Ford

Kester Ford is DataSift's Director of Product Marketing

Share This