Fast rule mining in ontological knowledge bases with AMIE $$+$$ +

Galárraga, Luis; Teflioudi, Christina; Hose, Katja; Suchanek, Fabian M.

doi:10.1007/s00778-015-0394-1

Cited by 404 publications

(413 citation statements)

References 32 publications

(33 reference statements)

Supporting

Mentioning

375

Contrasting

Order By: Relevance

“…Using MDL, we can approximate the optimal number of rules in a parameter-free, information-theoretic way, leading to fewer but descriptive rules. Table 3: Compression: The small % bits needed (relative to an empty model) and number of rules found by various models demonstrate the effectiveness of KGist variants at finding a concise set of rules in G. AMIE+ [17] finds Horn rules, which cannot be encoded with our model, so we only report the number of rules it finds. Freq and Coverage are baseline models that we introduce by greedily selecting from our candidate set C (without MDL) the top-k rules that (1) correctly apply the most often and (2) cover the most edges, resp.…”

Section: Rule Conciseness and Interpretabilitymentioning

confidence: 99%

“…Here we quantitatively analyze the effectiveness of KGist at identifying a diverse set of anomalies, and demonstrate the interpretability of what it finds. Whereas most approaches focus on exceptional facts [51], erroneous links, erroneous node type information [35], or identification of incomplete information (e.g., link prediction) [17], KGist rules can be used to address multiple of these at once. To evaluate this, we inject anomalies of multiple types into a KG, and see how well KGist identifies them.…”

Section: [Q2] What Is Strange In a Kg?mentioning

confidence: 99%

See 1 more Smart Citation

What is Normal, What is Strange, and What is Missing in a Knowledge Graph: Unified Characterization via Inductive Summarization

Belth

Zheng

Vreeken

et al. 2020

Proceedings of the Web Conference 2020

View full text Add to dashboard Cite

Knowledge graphs (KGs) store highly heterogeneous information about the world in the structure of a graph, and are useful for tasks such as question answering and reasoning. However, they often contain errors and are missing information. Vibrant research in KG refinement has worked to resolve these issues, tailoring techniques to either detect specific types of errors or complete a KG.In this work, we introduce a unified solution to KG characterization by formulating the problem as unsupervised KG summarization with a set of inductive, soft rules, which describe what is normal in a KG, and thus can be used to identify what is abnormal, whether it be strange or missing. Unlike first-order logic rules, our rules are labeled, rooted graphs, i.e., patterns that describe the expected neighborhood around a (seen or unseen) node, based on its type, and information in the KG. Stepping away from the traditional support/confidence-based rule mining techniques, we propose KGist, Knowledge Graph Inductive SummarizaTion, which learns a summary of inductive rules that best compress the KG according to the Minimum Description Length principle-a formulation that we are the first to use in the context of KG rule mining. We apply our rules to three large KGs (NELL, DBpedia, and Yago), and tasks such as compression, various types of error detection, and identification of incomplete information. We show that KGist outperforms task-specific, supervised and unsupervised baselines in error detection and incompleteness identification, (identifying the location of up to 93% of missing entities-over 10% more than baselines), while also being efficient for large knowledge graphs.

show abstract

Section: Rule Conciseness and Interpretabilitymentioning

confidence: 99%

Section: [Q2] What Is Strange In a Kg?mentioning

confidence: 99%

What is Normal, What is Strange, and What is Missing in a Knowledge Graph: Unified Characterization via Inductive Summarization

Belth

Zheng

Vreeken

et al. 2020

Proceedings of the Web Conference 2020

View full text Add to dashboard Cite

show abstract

“…Our work is also related to AMIE+ [9], a system for mining Horn rules in knowledge bases, where in each rule the body is a path and the head is a relation. Our work is orthogonal to [9] since we do not focus on mining rules, but rather on ranking the most prominent equivalences for a given relation path. The application of our method for rule mining, considering more generic rules than Horn rules, is part of our future work.…”

Section: Related Workmentioning

confidence: 99%

Identifying Equivalent Relation Paths in Knowledge Graphs

Mohamed

Muñoz

Nováček

et al. 2017

Lecture Notes in Computer Science

View full text Add to dashboard Cite

Abstract. Relation paths are sequences of relations with inverse that allow for complete exploration of knowledge graphs in a two-way unconstrained manner. They are powerful enough to encode complex relationships between entities and are crucial in several contexts, such as knowledge base verification, rule mining, and link prediction. However, fundamental forms of reasoning such as containment and equivalence of relation paths have hitherto been ignored. Intuitively, two relation paths are equivalent if they share the same extension, i.e., set of source and target entity pairs. In this paper, we study the problem of containment as a means to find equivalent relation paths and show that it is very expensive in practice to enumerate paths between entities. We characterize the complexity of containment and equivalence of relation paths and propose a domain-independent and unsupervised method to obtain approximate equivalences ranked by a tri-criteria ranking function. We evaluate our algorithm using test cases over real-world data and show that we are able to find semantically meaningful equivalences efficiently.

show abstract

“…Hence, anyone who is not known to be a child of Obama is not. The validity of the PCA has been evaluated manually [6] on YAGO. For relations with generally high functionality [17], the PCA holds nearly perfectly.…”

Section: Obtaining Completeness Informationmentioning

confidence: 99%

But What Do We Actually Know?

Razniewski¹,

Suchanek²,

Nutt³

2016

Proceedings of the 5th Workshop on Automated Knowledge Base Construction

View full text Add to dashboard Cite

Knowledge bases such as Wikidata, DBpedia, YAGO, or the Google Knowledge Vault collect a vast number of facts about the world. But while quite some facts are known about the world, little is known about how much is unknown. For example, while the knowledge base may tell us that Barack Obama is the father of Malia Obama and Sasha Obama, it does not tell us whether these are all of his children. This is not just an epistemic challenge, but also a practical problem for data producers and consumers. We envision that KBs become annotated with information about their recall on specific topics. We show what such annotations could look like, how they could be obtained, and survey related work.

show abstract

Fast rule mining in ontological knowledge bases with AMIE $$+$$ +

Cited by 404 publications

References 32 publications

What is Normal, What is Strange, and What is Missing in a Knowledge Graph: Unified Characterization via Inductive Summarization

What is Normal, What is Strange, and What is Missing in a Knowledge Graph: Unified Characterization via Inductive Summarization

Identifying Equivalent Relation Paths in Knowledge Graphs

But What Do We Actually Know?

Contact Info

Product

Resources

About