Method and apparatus for monitoring complex contagion and critical mass in online social media

Abstract

A method for determining social contagion while monitoring social media may be executable via operation of configured processing circuitry. The method may include receiving data indicative of social media activity of a plurality of users, selecting features of interest from the data, building a relationship network indicative of connections between the users and local networks to which various ones of the users belong, analyzing the features of interest to determine candidate features for classification as social contagion, determining a complex social contagion score for the candidate features, and providing an indication regarding the classification as social contagion based on the complex social contagion score.

What is claimed is: 
     
         1 . A contagion monitoring apparatus comprising processing circuitry configured to execute instructions that, when executed, cause the apparatus to:
 receive data indicative of social media activity of a plurality of users;   select features of interest from the data;   build a relationship network indicative of connections between users of the plurality of users and local networks to which various ones of the users belong;   analyze the features of interest to determine candidate features for classification as social contagion;   determine a complex social contagion score for the candidate features; and   provide an indication regarding the classification as social contagion based on the complex social contagion score.   
     
     
         2 . The apparatus of  claim 1 , wherein the processing circuitry is further configured to determine a critical mass score for the candidate features,
 wherein providing the indication regarding the classification as social contagion is further performed based on the critical mass score.   
     
     
         3 . The apparatus of  claim 2 , wherein the critical mass score represents a likelihood that the candidate feature has reached critical mass. 
     
     
         4 . The apparatus of  claim 1 , wherein the complex social contagion score represents a likelihood that the candidate feature is a complex contagion. 
     
     
         5 . The apparatus of  claim 4 , wherein the candidate feature is an expression of a trend or idea that has been adopted by a particular user of the plurality of users responsive to an apparent social pressure from a local network of the particular user, and
 wherein the complex social contagion score is indicative of a requirement for social pressure from at least two neighbors of the particular user from within the local network of the particular user to trigger adoption.   
     
     
         6 . The apparatus of  claim 4 , wherein the candidate feature is an expression of a trend or idea in a time window that has been adopted by a particular user of the plurality of users responsive to an apparent social pressure from a local network of the particular user, and
 wherein the complex social contagion score is indicative of a requirement for social pressure from at least two neighbors of the particular user from within the local network of the particular user to trigger adoption.   
     
     
         7 . The apparatus of  claim 1 , further comprising a multi-stage filter configured to filter the received data to reduce processing load associated with analyzing the features of interest. 
     
     
         8 . The apparatus of  claim 7 , wherein the multi-stage filter comprises a sampler configured to randomly sample the received data, and a sliding window monitor configured to define a format for the features of interest. 
     
     
         9 . The apparatus of  claim 8 , wherein the multi-stage filter further comprises a third stage filter to filter features of interest by type criteria to identify the candidate features. 
     
     
         10 . The apparatus of  claim 1 , comprising an analytics engine configured to analyze the features of interest by determining candidate features based on:
 recency of achievement of a specified level of popularity of the feature of interest, or   spread of the feature of interest to other networks beyond a local network of a particular user of the plurality of users if the feature of interest has not achieved the specified level of popularity recently.   
     
     
         11 . The apparatus of  claim 1 , wherein the features of interest are social media hashtags, and wherein the candidate features are hashtags that have entered the top 300 most popular hashtags within a time window, or hashtags that appear to have broken out of the local network of a particular user of the plurality of users. 
     
     
         12 . The apparatus of  claim 1 , wherein the processing circuitry is further configured to build the relationship network indicative of connections in a time window between the users and the local networks to which the various ones of the users belong. 
     
     
         13 . A method executable via operation of configured processing circuitry, the method comprising:
 receiving data indicative of social media activity of a plurality of users;   selecting features of interest from the data;   building a relationship network indicative of connections between users of the plurality of users and local networks to which various ones of the users belong;   analyzing the features of interest to determine candidate features for classification as social contagion;   determining a complex social contagion score for the candidate features; and   providing an indication regarding the classification as social contagion based on the complex social contagion score.   
     
     
         14 . The method of  claim 13 , wherein the analyzing the features of interest to determine the candidate features for classification as social contagion comprises analyzing the features of interest over time. 
     
     
         15 . The method of  claim 13 , further comprising determining a critical mass score for the candidate features,
 wherein providing the indication regarding the classification as social contagion is further performed based on the critical mass score.   
     
     
         16 . The method of  claim 15 , wherein the critical mass score represents a likelihood that the candidate feature has reached critical mass. 
     
     
         17 . The method of  claim 13 , wherein the complex social contagion score represents a likelihood that the candidate feature is a complex contagion. 
     
     
         18 . The method of  claim 17 , wherein the candidate feature is an expression of a trend or idea that has been adopted by a particular user of the plurality of users responsive to an apparent social pressure from a local network of the particular user, and
 wherein the complex social contagion score is indicative of a requirement for social pressure from at least two neighbors of the particular user from within the local network of the particular user to trigger adoption.   
     
     
         19 . The method of  claim 13 , further comprising applying a multi-stage filter to the received data to reduce processing load associated with analyzing the features of interest. 
     
     
         20 . The method of  claim 19 , wherein the applying the multi-stage filter comprises applying a first stage filter to randomly sample the received data, and applying a second stage filter to define a format for the features of interest. 
     
     
         21 . The method of  claim 20 , wherein applying the multi-stage filter further comprises applying a third stage filter to filter features of interest by type criteria to identify the candidate features. 
     
     
         22 . The method of  claim 13 , wherein analyzing the features of interest comprises determining candidate features based on:
 recency of achievement of a specified level of popularity of the feature of interest, or   spread of the feature of interest to other networks beyond a local network of a particular user of the plurality of users if the feature of interest has not achieved the specified level of popularity recently.   
     
     
         23 . The method of  claim 13 , wherein the features of interest are social media hashtags, and wherein the candidate features are hashtags that have entered the top 300 most popular hashtags within a time window, or hashtags that appear to have broken out of the local network of a particular user of the plurality of users.