Background

Sentiment analysis of natural language texts is a large and growing field. Some methods for generating sentiment lexicons assume positive and negative sentiment using synonyms and antonyms. Such methods may not accurately capture the sentiment of a word. Other methods use semantics, such as "and" and "but", or tone/orientation to determine a sentiment of a word, but such methods may have low accuracy. Current methods for analyzing sentiment treat only single complete documents, for example, to determine if a movie review is good or bad or quantify opinion from a product review. Therefore, there is a need for a method of generating a more accurate sentiment lexicon and for determining a sentiment over a plurality of texts.

Technology

This technology uses statistical analysis of text streams to simultaneously monitor changes in reputation to thousands of distinct news entries. Commercial applications of this technology include 1) market research - the technology can analyze the reputation of people, products, and companies without the need for expensive surveys or polling, 2) financial analysis - the conversion of news data to time-series facilitates automated investment analysis, e.g. strengthening pair trading investment strategies by identifying companies without the need for expensive surveys/polling, 2) internet search engines - augmenting results by providing sentiment data on articles.

Advantages

Gives the ability to monitor entity sentiment as a time-series in any text stream, such as news or blogs, even if written in different languages and from different news sources.

Application

Commercial applications include: - Market Research- This technology can analyze the reputation of people, products, and companies without the need for expensive surveys or polling - Financial Analysis- The conversion of news data to time-series facilitates automated investment analysis - Internet Search Engines- augmenting results by providing sentiment data on articles