TIARA: Multi-grained Retrieval for Robust Question Answering over Large Knowledge Bases
About
Pre-trained language models (PLMs) have shown their effectiveness in multiple scenarios. However, KBQA remains challenging, especially regarding coverage and generalization settings. This is due to two main factors: i) understanding the semantics of both questions and relevant knowledge from the KB; ii) generating executable logical forms with both semantic and syntactic correctness. In this paper, we present a new KBQA model, TIARA, which addresses those issues by applying multi-grained retrieval to help the PLM focus on the most relevant KB contexts, viz., entities, exemplary logical forms, and schema items. Moreover, constrained decoding is used to control the output space and reduce generation errors. Experiments over important benchmarks demonstrate the effectiveness of our approach. TIARA outperforms previous SOTA, including those using PLMs or oracle entity annotations, by at least 4.1 and 1.1 F1 points on GrailQA and WebQuestionsSP, respectively.
Related benchmarks
| Task | Dataset | Result | Rank | |
|---|---|---|---|---|
| Knowledge Base Question Answering | WEBQSP (test) | Hit@175.2 | 143 | |
| Knowledge Base Question Answering | WebQSP Freebase (test) | F1 Score78.9 | 46 | |
| Knowledge Base Question Answering | WebQSP → GrailQA-Tech (test) | F1 Score48.2 | 36 | |
| Knowledge Base Question Answering | GrailQA v1.0 (test) | Overall EM73 | 33 | |
| Knowledge Base Question Answering | GrailQAbility answerable 1.0 (test) | F1 (L)80.03 | 27 | |
| Knowledge Base Question Answering | GrailQA (test) | F152.8 | 27 | |
| Knowledge Graph Question Answering | GrailQA (Overall) | Hits@173 | 20 | |
| Knowledge Base Question Answering | WebQSP → GraphQA-Pop (test) | F137.8 | 20 | |
| Knowledge Base Question Answering | GrailQAbility unanswerable 1.0 (test) | F1 (L)93.43 | 19 | |
| Knowledge Base Question Answering | GrailQA 500-sample (dev) | F1 Score82.6 | 18 |