作者: Diego Molla , Menno van Zaanen , Daniel Smith
DOI:
关键词:
摘要: Current text-based question answering (QA) systems usually contain a named entity recogniser (NER) as core component. Named recognition has traditionally been developed component for information extraction systems, and current techniques are focused on this end use. However, no formal assessment done the characteristics of NER within task answering. In paper we present that aims at higher recall by allowing multiple labels to strings. The is embedded in system overall QA performance compared one with traditional variation only allows single labels. It shown added noise produced introduced additional offset gained, therefore enabling have better chance find answer.