SXSW Interactive 2013
Siri, Search & Science Need Semantic Opinion Data
Companies are realizing the value of extracting semantic data from text. Using databases of semantic facts, Siri leverages Wolfram Alpha to answer factual questions, and Google's knowledge graph understands "things, not strings" in order to provide facts about search results. However, an increasing proportion of web searches are not about facts, but rather about opinions, or more broadly, about information without a single source answer. People increasingly want to know facts about opinions, such as the best rated movie, whether Obama is leading the presidential race, or what purchases often lead to greater fulfillment. The semantic web will deliver increasing value by allowing mashups of opinions, not facts, allowing applications to deliver facts about opinions, leveraging the wisdom of crowds. We show how semantic opinion data can, in aggregate, answer questions of interest to market researchers, advertisers, politicians, and psychologists, as well as the average iPhone user.
Share this idea
- What is the difference between a fact, an opinion, and a fact about opinions and what evidence is there that people are increasingly most interested in facts about opinions?
- How will the impact of mobile technologies like Siri make semantic data essential?
- How can aggregated semantic opinions answer questions in diverse domains such as entertainment, politics, science, and advertising?
- Where can I get semantic data that I can use in my applications?
- How can I use aggregated semantic opinion data to answer questions of interest to my organization?
Ravi Iyer, Data Scientist / PostDoctoral Researcher, University of Southern California