Polaris Semantic Parser
Lymba's Polaris is a powerful semantic parser that takes free text and extracts a rich set of semantic relations between words or concepts in each sentence. Unlike traditional verb driven semantic parsers, Polaris extracts semantics from the verbal constituents, noun phrases, prepositional phrases, and many more.
There are more than 40 semantic relations detected in text.
A few examples include temporal, spatial, part-whole, agent, is-a, purpose, manner, result, and topic. A typical semantic relation tagging might look like this:
Migratory birds fly south every year to avoid the cold weather.
IS-A(migratory birds, birds)
AGENT(migratory birds, fly)
LOCATION(south, fly)
TEMPORAL(every year, fly)
PURPOSE(to avoid the cold weather, fly)
THEME(the cold weather, avoid)
PROPERTY(cold, weather)