FactSquared Stock Sentiment Speech Analytics Data USA

FactSquared Analyze offers unique data-driven insights into what public figures are -- and aren’t -- saying in their public comments on market-moving topics. How We Do It FactSquared’s proprietary analytics engine uses unique neural models that assess whether an individual is deviating from their normal speaking patterns. Analyze examines characteristics such as voice level, stress, rate of speech, word choice, visual cues and more than 250 other factors, indexed to speaker’s historical speech patterns. Key topics can then be parsed out, trended over time, and analyzed in a way that highlights deviations from past behavior. Coverage Each year, FactSquared Analyze monitors: - 12,000 earnings calls - 1,400 Congressional hearings - 3,500 Executive branch Tweets - CEO and Analyst commentary on 3,000 publicly-traded stocks Always Learning - Always Growing FactSquared Analyze grows more powerful with each word spoken. The algorithms learn how public figures speak, permitting the neural model to identify when something isn’t quite right in what is being discussed, tailored to the individual, the context and whether it’s scripted or unscripted. Underlying Capabilities FactSquared Analyze sits atop FactSquared Transcribe, which offers accurate, near-instant voice-to-text transcription for critically important public conversations. Transcribe uses the same proprietary AI-based models to identify who is speaking, using unique voice characteristics and vocabulary of a particular individual, and punctuates specific to the person’s grammar, with accuracy that scores 4.5% over leading industry accuracy benchmarks.