An important component of text analytics today is sentiment analysis. Existing tools that analyze “sentiment” are primarily based on words that represent generic positive or negative attitudes. These can give an idea of a very strong sentiment (terrible, awful, marvelous, good or bad) but cannot drill down to a sentiment that is conditional on the domain or object of the word. Such lexical instances are domain and culture-specific and do not hold the same connotation when they appear in all texts or when expressed by authors from different backgrounds or ideological inclinations.

​

Moreover, the identification of a sentiment as “positive”, “negative” or “neutral” without drilling down to the actual sentiment is problematic. Smart, pretty, honest, effective are all positive towards a person whereas stupid, ugly, corrupt and ineffective are all negative, but they each express a sentiment towards a different attribute of the person. The same word may also have a different “sentiment meaning” when applied to a different entity type, e.g. a smart (intelligent) person vs. a smart (pretty) dress, a corrupt (dishonest) person vs. a corrupt (damaged) file.

​

The IntuView technology integrates ontological, linguistic, and extra-linguistic knowledge. It differentiates between language registers and styles and extrapolates from sentiment towards a number of objects to a general sentiment towards the “ontological parents” of those objects. Hence, the IntuView technology not only identifies the general sentiment of the text (positive or negative) but drills down to identify:

​

• The Sentiment “holder” (who is expressing the sentiment).

• The sentiment “target” (towards whom it is expressed).

•  The attribute of the object towards which the sentiment is expressed (e.g. “honesty”, “effectiveness”, “Intelligence”).

• The nuance and degree of the positive or negative sentiment (highly positive or negative, medium or low);

•  The “aggregated sentiment” towards the entity by different sentiment “holders”.

•  The context of the sentiment or the theme of the discourse in which it occurs.

• The inferred sentiment towards entities who are not mentioned explicitly in the text but alluded to as “parent” entities through those entities that are mentioned explicitly in the text.