Companies don’t rely solely on sentiment analysis to inform strategic decisions, but it is a powerful complement to traditional market intelligence. Like many things in the social data ecosystem, sentiment is a rapidly evolving tool. While challenges like accurately identifying and classifying irony, sarcasm, and emoticons exist, companies are meeting that challenge with increasingly sophisticated, Twitter syntax-specific tools.
IHS, a leading analysis and information provider–and one we’re excited to announce as a new Plugged In partner–built a sentiment intelligence tool to facilitate its clients’ use of social data. This past summer, IHS released the U.S. Sentiment Index, a tool that assess realtime Tweets, providing a representation of the average mood of the United States.
We were curious to learn more about how sentiment analysis is being used across industries not typically known for their use of social data. IHS shared an example of how companies, in this case in the oil and gas industry, incorporated sentiment analysis to provide a deeper understanding of public opinion on hydraulic fracturing, commonly referred to as “fracking.” IHS looked at the sentiment of fracking-related Tweets globally, as well as in specific states like Colorado. Analysis determined in which states the most Tweets about fracking originated and what keywords are most commonly associated with the topic. Both of these things contribute to the companies’ understanding of the drivers of public sentiment on the topic of fracking–valuable information.
To expose another layer of insight, IHS used network analysis to understand and measure the virality of messages. One of the takeaways from the research was that the content of a message is not as important as understanding which voices influence the dissemination of that message. Further, the number of followers an influencer has was not as important as whether one of those followers retweeted the message outside the influencer’s immediate circle of followers. These are just a few highlights from a more in-depth paper we wrote.