Confidential — Stefan Michaelcheck Only

An Experimental Analysis of Geographic Knowledge Graph Completion Methods

2026empirical benchmarkingevaluativeevaluation

Association for Artificial Intelligence 2026

Underline Science Inc.

https://doi.org/10.48448/b9hc-tc33OpenAlex: W7137837247
1
URLs Found
0
Internal Citations
1
Authors
usable
Abstract Quality
GPT-5.5 Abstract Analysis

Problems Identified (5)

Incomplete geographic knowledge graphs: Geographic knowledge graphs are often incomplete because they are typically derived from crowd-sourced data, making completion important.

Spatially unaware KGC methods: Most current knowledge graph completion methods are generic and do not account for the spatial nature of geographic knowledge graphs.

Impractical geo-specific KGC methods: Existing methods tailored to geographic knowledge graphs are described as computationally expensive or designed for an impractical closed-world setting.

Incomplete geographic knowledge graphs: Geographic knowledge graphs are often incomplete because they are typically derived from crowd-sourced data, making completion important.

Spatially unaware KGC methods: Most current knowledge graph completion methods are generic and do not account for the spatial nature of geographic knowledge graphs.

Proposed Solutions (4)

Benchmark evaluation of KGC methods: The paper evaluates existing state-of-the-art standard and geo-specific knowledge graph completion methods on a large dataset of geographic knowledge graphs.

Research direction recommendations: The paper suggests possible areas of future work for geographic knowledge graph completion.

Benchmark evaluation of KGC methods: The paper evaluates existing state-of-the-art standard and geo-specific knowledge graph completion methods on a large dataset of geographic knowledge graphs.

Research direction recommendations: The paper suggests possible areas of future work for geographic knowledge graph completion.

Results (3)

Poor KGC method performance:

Open research problem remains:

Poor KGC method performance:

Research Domain

Geographic knowledge graph completion

← Back to all papers