LangPro
: Natural Language Theorem Prover
LangPro tries to find out a semantic relation between
a set of premises and a hypothesis.
The semantic relation can be:
entailment/yes,
contradiction/no,
neutral/unknown.
LangPro obtains competetive results on the SICK and FraCaS RTE
(i.e. recognizing textual entailment) datasets.
The code is available at GitHub.