Facebook Topic Data Just Got an Upgrade with Super Public Text Samples

Kester Ford
18th November 2015 0 Comments

Super Public text samples are available from November 18th for all PYLON for Facebook Topic Data customers at no extra cost.

Today, we are pleased to announce the availability of Super Public text samples, a new feature of PYLON for Facebook Topic Data that allows customers to retrieve public Facebook posts that match their filtering rules. Organizations will be able to validate their filters against these Super Public posts to gain additional confidence in their query results.

Screen Shot 2015-11-16 at 11.12.35 AM

With the availability of Facebook topic data – announced earlier this year – DataSift’s customers have been able to provide brands and marketers with unprecedented insights into audiences on Facebook. In order to protect the privacy of the people on Facebook, these insights are all anonymized and aggregated. The introduction of Super Public text samples means that these companies – and the brands they serve – can now  have greater confidence in the insights derived from topic data. They will also be able to use the samples to generate deeper insights through machine learning.

Super Public text samples are intended to be used in three main ways:

  • Verifying and improving filter results – Access to text in Super Public posts helps you verify that your filters are working properly and improve them over time. Super Public posts help distinguish false positives and identify new exclusion rules, e.g. an exclusion rule should be set up for posts mentioning Harrison Ford to prevent showing up as false positive for Ford Motors.
  • Validating insights – Verbatims from Super Public posts can also be used to double check that understanding of the stories driving engagement derived from link, hashtag and topic analysis is correct.
  • Building better classifiers – Super Public text samples can also be used to supercharge your classifiers, by using them to train machine-learning algorithms to produce text classifiers in order to garner deeper audience insights at scale.

Screen Shot 2015-11-16 at 11.12.42 AM

Super Public Text Samples helps identify false positives and improves filters.

What exactly are Super Public posts?
Super Public posts are those that match all of the following criteria:
1) They are posted by people who have a “Follow” setting enabled in their profile.

2) And the post is posted with the privacy setting set as public.

3) And the post is not on the timeline of another person.

These criteria are designed to ensure that the privacy of the people on Facebook is respected. Posts that satisfy the Super Public definition are really intended for the public to see. As with all topic data, these text samples are anonymized before being made available to DataSift.

To find out more about Super Public text samples (and nine other reasons Facebook topic data will change your world) check out our latest webinar! You can discover how to work with Super Public text samples, by checking out our developer blog.

Kester Ford

Written by Kester Ford

Kester Ford is DataSift's Director of Product Marketing

Share This