Sentiment Scoring using Fine Grain Scoring Versus Bucketing

My last blog generated some questions about the three types of Tweet scoring.  Two Bucket Rough Grain Scoring classifies tweets as positive or negative.  This scoring method over weights neutral Tweets and under weights extreme Tweets.  Three Bucket Rough Grain scoring adds a bucket for neutral Tweets but still suffers from under weighting extreme positive or negative Tweets.  The most accurate method is Fine Grain Scoring that assigns a score to each tweet and does not have the over or under weighting problem.   These three methodologies can generate very different aggregate sentiment values.

Below is an example of the impact of the different methods of scoring.

• Alcoa Aluminum (AA) from 8 AM (Central) on 7/8/2013 to 3 PM on 7/9
• 180 Tweets observed, 125 with non-zero scores
• Earnings call at 4 PM on 7/9
• End result was a slight earnings miss

Sentiment represented by the three methods of scoring

Fine grain scoring = -.187 by looking at the tweets prior to the announcement a reasonable expectation would be a slight miss, which is what happened.

Two bucket scoring = 13, a positive number representing a beat of estimates?

Three bucket scoring = -8 a significant miss expectation?

The bucket scoring methods provided a much less accurate view of expectations of this earnings announcement.  Score each tweet independently and aggregate for a more accurate score.