Home > Development > Quotes Over Time

Quotes Over Time

The Quotes Over Time (QoverT) system analyzes streaming news feeds, extracts quotes from public figures, develops unique social and context-based networks, and visualizes changes in public figure sentiment across multiple, unfolding news events.

Researched in conjunction with The Geospace and Media Tool, the Quotes Over Time system studies the connection between public figures, published media, and public opinion. Over a time-based user interface, the system automatically scours and archives news from the Reuters AlertNet service extracting quotes and social networks data from those news articles.

A thorough analysis of news articles from around the world was initially conducted to identify key attributes and indicators within news articles. This information, once mapped onto an analysis of public sentiment during and after these articles' publication, revealed a striking yet simple conclusion: Public sentiment can be directly tied to the spoken words of public figures — not the words of journalists and editors.

Integrating PIIM's semantic technologies with opensource packages, ingested news articles are dynamically aggregated and categorized. With social network and other information a complex "news file" is developed for each article and topic.

PIIM employed several data visualization packages to graph and cross reference changes in reporting and responses to events and public figures. The Quotes Over Time system provides a clear, visual explanation of the relationships between news, public opinion, and public figures.

The Quotes Over Time system provides deep insight across divergent topics through the following features:

• 24 /7 news ingestion and archival process connected to the Reuters AlertNet system
• Quote Extraction technology developed by PIIM
• Social networking identification and visualization technologies
• Semantic-based clustering software to dynamically and automatically tag news articles
• A suite of visualization modules to graph time- and context-based information