Type evaluation in a question-answering system

摘要

A system and method for automatically mapping LATs and candidate answers to multiple taxonomies without a need to merge these taxonomies. The method includes using a syntactic analysis of a corpus to extract all type instances of the LAT. The extracted instances are then mapped to a given taxonomy and clustered in a set of supertypes. Each supertype receives a score based on the coverage of LAT instances in the corpus. The method includes mapping the candidate answer to the same taxonomy to determine if the candidate answer is an instance of a significant supertype. Then the score of a candidate answer is obtained by aggregating or taking a maximum of the score of the matched significant supertypes. This score evaluates the type match between the LAT and candidate answer for a taxonomy. Multiple taxonomies can be used to increase the chance of LAT and candidate answer mapping.

主权项

1. A computer-implemented method for scoring candidate answers to questions, said method comprising:
receiving a natural language question into a computer-implemented question answering system; applying syntactic analysis on said question to determine a lexical answer type (LAT) for said question; generating a second question based on said LAT; identifying one or more LAT instances by inputting said generated second question through a search engine; mapping said one or more LAT instances to one or more taxonomies; identifying, in said one or more taxonomies, one or more supertype entities associated with said mapped LAT instances; generating a type evaluation score for a candidate answer based on a match between said candidate answer and a supertype entity within a taxonomy; generating one or more answers to said question, wherein each answer has an associated confidence score based at least in part on said type evaluation score, wherein said one or more supertype entities are significant supertype entities, a significant supertype entity determined by:
calculating a recall loss between a parent node and child node within a taxonomy, andidentifying the parent node having the smallest recall loss exceeding a threshold as the significant supertype entity.