Computation as Analysis of Twitter Streams Redirected from Computation as Analysis of Twitter Streams

Computational Analysis of Twitter Streams (1J) 

Convener:  Sanjay C. Sood, Allvoices 

Notes-taker(s):  

A. Tags for the session - technology discussed/ideas considered:  

Semantic analysis, random sampling, fire hose, garden hose, trends, social graph, lexical normalization, data mining, predictive analysis  

 
B.  Discussion notes, key understandings, outstanding questions, observations, and, if appropriate to this discussion: action items, next steps: 

• Work being done to analyze historical trends to predict future events and spikes 

• Use of statistical random sample of entire twitter stream can yield good results (5-10%) 

• Lack of historical data (10 days) from Twitter makes trend analysis more difficult. 

• Multiple strategies for collecting samples of data via Twitter API and 2rd parties 

• Normalization of data and compete data sets critical for quality analysis.