Finding Top Message Threads for a Filter

This is the second post in a series where we explore Entities in the Social Analytics Lab API. Through this series of posts, we will provide details on how the Entities can be used to accomplish some basic scenarios in the Social Analytics lab.

In this scenario we’ll look at identifying top threads for one of the lab datasets. To accomplish this, you will need three Entities in our API (pictured below):

Filters

Filters are a central concept in storing and retrieving data in the Social Analytics Lab. You could think of the set of filters as a list of topics available in an Social Analytics Instance. For example, the Windows 8 lab includes the following filters:

Notice that three of the Filters are labeled as Data Acquisition Filters. Filters can control what activity from social media channels is tracked in an instance of Social Analytics or it can be simply a materialized view over the social data tracked in an instance. With our permissions model for filter definitions, control over what Social Data is tracked in an instance can be controlled while giving many users permissions to define filters that are views over the data.

MessageThreads

We define a MessageThread as a post and all of its replies within a channel. MessageThreads are the primary unit of activity we use to track social channel activity. If you examine the MessageThreads entity, you’ll find a collection of measures described an aggregated view of the thread, including:

• Replies
• Answers
• Likes
• Retweets
• IsQuestion
• IsAnswered
• PositiveSentiment
• NegativeSentiment

The Replies field is a total of all replies of all types in the thread, so you don’t want to sum Replies and Answers etc., or you’ll be double-counting.

FilterResultCaches

FilterResultCaches represents what MessageThreads relate to all filters, and is one of the most useful Entities for analyzing Social Analytics data. Using the following LINQ Query in LinqPad:

 (
 from f in FilterResultCaches
 orderby f.MessageThread.Replies descending
 where f.FilterId == 6
 && f.MessageThread.Replies > 1
 && f.LastUpdatedOn > DateTime.Now.AddDays(-7)
 select new
 {
 f.MessageThread.Replies,
 f.Site.Name,
 f.MessageThread.PositiveSentiment,
 f.MessageThread.NegativeSentiment,
 f.MessageThread.LastReplyContentItem.Title,
 f.MessageThread.LastReplyContentItem.HtmlUrl,
 }
).Take(10)
 

we can find the Messagethreads with the most replies during the last 7 days in the “Windows 8 and Tablet PCs” filter:

The results include our calculated Positive and Negative Sentiment. The values indicate the number of content items in the MessageThread with 80% or more confidence that the tone is either positive or negative.

Check out our next blog post to find out more about what you can do with the entities in the Social Analytics Lab.

Also, please let us know what questions you have or other interesting discoveries you find in the labs.