What type of study is this?

September 5, 2025Open Access

A hereditary attentive question answering framework for knowledge bases

RMRômulo C. de MelloUniversidade Federal de Juiz de Fora JGJorão GomesVienna University of Economics and Business JSJairo Francisco de SouzaUniversidade Federal de Juiz de Fora

Key Points

Multi-tool entity recognition significantly outperforms single-tool methods, enhancing query handling.
SPARQL correctness for mid-complex queries reaches up to 80%, demonstrating framework effectiveness.
The integration of NLP tools with dummy templates increases coverage for challenging KBQA scenarios.
Further improvements in morphological handling and specialized embeddings are necessary for niche query support.

Abstract

Background. The rapid growth of online data has made retrieving relevant information a challenging task, prompting the rise of Knowledge Base Question Answering (KBQA) systems that handle complex, multi-hop queries. Purpose. This extended work refines our previous pipeline by introducing structured dummy templates, a Hereditary Tree-LSTM (HTL) for classification, and more comprehensive analyses of entity recognition, property extraction, and SPARQL assembly. Methods. We enhanced the LC-QUAD 2.1 dataset with standardized templates and evaluated a flexible pipeline that integrates DeepPavlov, Falcon, SpaCy, qualifiers constraints, and reverse lookups. Results. Our experiments reveal that multi-tool entity recognition outperforms single-tool methods, while property extraction benefits from extended property sets and refined ranking strategies. Overall SPARQL correctness reaches up to 70–80% in mid-complex queries but remains lower in domain-specific subsets. Conclusion. The proposed synergy of NLP tools and refined dummy templates increases coverage for complex KBQA, though further improvements in morphological handling and specialized embeddings may be needed to address challenging multi-hop or niche queries comprehensively.

Ask AI

Helpful

Bookmark

View Full Paper