Sentiment: This button toggles the sentiment analysis display for each article. 'NEUTRAL' is a standalone value. It is typically dominant, but if the value is under 0.5, then the text is either 'POS' or 'NEG'. This data is provided courtesy of the NLTK Sentiemnt API.
Frequency: This changes the 'ARTICLE MAPS' into 'heat maps' for the frequency of important terms. The greater the (red) saturation, the more frequent the usage.
Quotes: This toggles the quote chart feature — a scatter plot of extracted quotations from the article. Each one is represented by circle, color coded to it's original publication. The metric used for the 'x'axis is a score based on the a count for the frequency of each work in the quote as compared with the other quotes. This is then divided by a fraction of the total number of words in this quote. The same is done for the 'y'axis, but compared to all of the words in every article. The hope is for densely meaningful quotes to score highly, and for similar quotes, used by different publishers, to be grouped together. Hover over the circles to see the text.
Entities: This toggles the 'NAMED ENTITIES' display. These concentric arcs represent the levels of connectivity between the people, places, organizations, etc., mentioned. The center arc has entities that appeared in all of the articles, while the outer, shows what was unique.
Grouping: Grouping is a sub-function of Entities. This toggles a hover action for the arcs that shows what other entities appeared along with the one you are currently over. At the outer arcs, this will reveal which entities were in each individual article.
Terms: Terms is also a sub-function of Entities. This toggles the display of 'KEY TERMS'. These are important words that appeared in sentences with the named entities, show with a bar chart to visualize their frequency in the text. Hovering on the chart or the rectangle nodes next to the entities will reveal the links binding term to entity.
Search: Enter a single term into this field and it will highlight anywhere it appears on the visualization (entity arc, or word box). The search is case sensitive and does not accept punctuation.