Broadcast Transcript Feed with Sentiment Analysis (GBTS)

Broadcast TV and radio transcripts are collected from cable providers, satellite, streaming or terrestrial broadcast from multiple global capture sites that record video and audio 24/7. TVEyes and Finch Computing have partnered to perform a variety of information extraction and enrichment tasks on the associated transcripts that include: insertion of broadcast metadata, sentiment and salience scores, as well as fully-tagged and resolved companies with their ticker symbols. THERE IS OPPORTUNITY AND ALPHA EMBEDDED IN BROADCAST LANGUAGE Unique: A machine-readable source of aggregated content in context not available in a single feed elsewhere. Trusted: News and events become validated when they are aired on television, relative to social media. Additive: Fully disambiguated exogenous content -- in context -- is a generator of incremental alpha FINCH COMPUTING'S SENTIMENT CAPABILITIES UNEARTH NUANCES THAT DRIVE ALPHA • Trained on a large and diverse corpus of news, conversational and narrative datasets to capture nuances in language. • Understand sentiment at the entity level, rather than just at the sentence or document level. • Understand where entity-level sentiment emanates from and to. • Employs a proprietary variation of deep transfer learning to fine-tune language models and optimize contextual word embeddings to predict sentiment in key topic areas of financial and business news. • Can interpret various degrees of sentiment, whether someone is more positive, less positive; more aloof or less aloof, etc. These comparisons can be made from one source to another or from day to day to determine how an entity’s sentiment has changed or evolved over time. • Apply our algorithmic, entity level approach to sentiment to specific topics discussed. To do it, we examine a temporal window of text and employ our inference-trained models to decipher the topics about which the language is referring.