Question Answering Over Spatio-Temporal Knowledge Graph

2024-02-18

arXiv.org

AI Summary

Existing KGQA methods struggle to model spatiotemporal constraints and lack large-scale, spatiotemporally grounded question-answering benchmarks. To address this, we introduce STQAD—the first large-scale benchmark for Spatiotemporal Knowledge Graph Question Answering (STKGQA)—comprising 10,000 natural-language questions with fine-grained spatiotemporal annotations. We propose STComplEx, a spatiotemporally aware knowledge graph embedding method that jointly models semantic, spatial, and temporal information. Furthermore, we design STCQA, an end-to-end framework integrating spatiotemporal question parsing, multi-hop reasoning over spatiotemporal KGs, and semantic matching-based answer retrieval. Extensive experiments demonstrate that STCQA significantly outperforms state-of-the-art KGQA models on STQAD, validating both the benchmark’s quality and the efficacy of our methodology. STQAD and STCQA collectively establish foundational infrastructure and a principled technical paradigm for advancing STKGQA research.

Technology Category

• Data Mining & Knowledge Management: Mining of Spatial, Temporal or Spatio-Temporal Data
• Knowledge Representation and Reasoning: Geometric, Spatial, and Temporal Reasoning
• Natural Language Processing: Question Answering

Application Category

• Semantics and Knowledge: Methods to enhance, augment, integrate or synergize semantic models such as knowledge graphs and LLMs
• Search and Retrieval-Augmented AI: Web question answering
• Graph Algorithms and Modeling for the Web: Querying, indexing, and retrieval in Web-related graphs

Abstract

Spatio-temporal knowledge graphs (STKGs) extend the concept of knowledge graphs (KGs) by incorporating time and location information. While the research community's focus on Knowledge Graph Question Answering (KGQA), the field of answering questions incorporating both spatio-temporal information based on STKGs remains largely unexplored. Furthermore, a lack of comprehensive datasets also has hindered progress in this area. To address this issue, we present STQAD, a dataset comprising 10,000 natural language questions for spatio-temporal knowledge graph question answering (STKGQA). Unfortunately, various state-of-the-art KGQA approaches fall far short of achieving satisfactory performance on our dataset. In response, we propose STCQA, a new spatio-temporal KGQA approach that utilizes a novel STKG embedding method named STComplEx. By extracting temporal and spatial information from a question, our QA model can better comprehend the question and retrieve accurate answers from the STKG. Through extensive experiments, we demonstrate the quality of our dataset and the effectiveness of our STKGQA method.

Problem

• Addresses lack of spatio-temporal KGQA datasets
• Improves reasoning over implicit spatio-temporal dependencies
• Enhances KGQA methods for spatio-temporal interactions

Innovation

• Joint embedding of temporal and spatial features
• Dynamic filtering through constraint-aware reasoning
• Creation of comprehensive spatio-temporal benchmark dataset