Media Sentiment

Science4Data’s Media Sentiment (sometimes just called Sentiment) is a score calculated based on a proprietary Natural Language Processing (NLP) algorithm that is tuned to identify the author’s attitude as positive, negative or neutral. Science4Data’s algorithm exceeds traditional sentiment analysis (i.e.: happy = positive; angry = negative; etc) by identifying, including and weighing significant geopolitical sentiment terms based on semantic analysis of thousands of news stories. As a result, the Science4Data Media Sentiment Score demonstrates targeted sentiment with a focus on Geopolitical Risk, Financial and Business content.

* In addition to Media Sentiment, Science4Data measures and tracks additional sentiment metrics including Scan/Ticker Sentiment, Sentiment Magnitude, Google Sentiment, Entity Level Sentiment and Overall Positivity and Negativity of an article.

Other Metrics

Science4Data employs the same semantic study based methodology to define metrics quantifying the Risk; Predictive and Temporal; and PESTEL (Politics, Economy, Society, Trade, Technology, Environment and Legal) dimensions of an article.

Science4Data continues to annotate the corpus of language. Changes and improvements to these measures should be expected and will be incorporated into analysis results on an ongoing basis.