Fast Pathfinding in Knowledge Graphs Using Word Embeddings

Martin, Leon; Boockmann, Jan H.; Henrich, Andreas

doi:10.1007/978-3-030-58285-2_27

Cited by 1 publication

(21 citation statements)

References 7 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Startpage shows a knowledge panel with information about a UK student exchange program named after Turing. While Startpage's result is a good attempt, the authors of [2] argue that knowledge panels for dual-entity queries could explain the relationship between the two mentioned entities, thereby potentially satisfying the users' information need without requiring them to consult the ranked search entries. Especially when the two entities are semantically distant, the way they are connected via a path in a KG, can provide valuable information for the users.…”

Section: Introductionmentioning

confidence: 99%

“…The authors of [2] interpreted this task as a pathfinding problem, where a path found between e and e in a KG serves as a means of describing the entity relationship. Web search engines like those mentioned before allow users to issue arbitrary textual queries with entities from basically any domain.…”

Section: Introductionmentioning

confidence: 99%

“…Therefore, the edges of a KG can be considered as bidirectional [3]. The vast size and generality of KGs like Wikidata pose several concrete problems for pathfinding [2]:…”

Section: Introductionmentioning

confidence: 99%

“…To tackle these problems, [2] proposes a bidirectional A* search algorithm [6] that considers the semantic distance between entities estimated via word embeddings of their labels to guide the search. The algorithm is parameterized with three hyperparameters α, β , and γ that weight the individual components of the employed cost function (s. Section 2).…”

Section: Introductionmentioning

confidence: 99%

“…The remainder of the present paper is organized as follows: Section 2 investigates foundations and related work, including a thorough recapitulation of [2]. Building upon this, Section 3 focuses on answering RQ1 through RQ3.…”

Section: Introductionmentioning

confidence: 99%

See 4 more Smart Citations

BiPaSs: Further Investigation of Fast Pathfinding in Wikidata

Martin

2023

Knowledge Graphs: Semantics, Machine Learning, and Languages

View full text Add to dashboard Cite

Purpose: A previous paper proposed a bidirectional A* search algorithm for quickly finding meaningful paths in Wikidata that leverages semantic distances between entities as part of the search heuristics. However, the work lacks an optimization of the algorithm’s hyperparameters and an evaluation on a large dataset among others. The purpose of the present paper is to address these open points. Methodology: Approaches aimed at enhancing the accuracy of the semantic distances are discussed. Furthermore, different options for constructing a dataset of dual-entity queries for pathfinding in Wikidata are explored. 20% of the compiled dataset are utilized to fine-tune the algorithm’s hyperparameters using the Simple optimizer. The optimized configuration is subsequently evaluated against alternative configurations, including a baseline, using the remaining 80% of the dataset. Findings: The additional consideration of entity descriptions increases the accuracy of the semantic distances. A dual-entity query dataset with 1,196 entity pairs is derived from the TREC 2007 Million Query Track dataset. The optimization yields the values 0.699/0.109/0.823 for the hyperparameters. This configuration achieves a higher coverage of the test set (79.2%) with few entity visits (24.7 on average) and moderate path lengths (4.4 on average). For reproducibility, the implementation called BiPaSs, the query dataset, and the benchmark results are provided. Value: Web search engines reliably generate knowledge panels with summarizing information only in response to queries mentioning a single entity. This paper shows that quickly finding paths between unseen entities in Wikidata is feasible. Based on these paths, knowledge panels for dual-entity queries can be generated that provide an explanation of the mentioned entities’ relationship, potentially satisfying the users’ information need.

show abstract

Section: Introductionmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

“…Therefore, the edges of a KG can be considered as bidirectional [3]. The vast size and generality of KGs like Wikidata pose several concrete problems for pathfinding [2]:…”

Section: Introductionmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

See 3 more Smart Citations