Knowledge Graph Identification

By Jay Pujara, Hui Miao, Lise Getoor, William W. Cohen. International Workshop on the Semantic Web, 2013

338 citations recorded by the source corpus.

Large-scale information processing systems are able to extract massive collections of interrelated facts, but unfortunately transforming these candidate facts into useful knowledge is a formidable challenge. In this paper, we show how uncertain extractions about entities and their relations can be transformed into a knowledge graph. The extractions form an extraction graph and we refer to the task of removing noise, inferring missing information, and determining which candidate facts should be included into a knowledge graph as knowledge graph identification. In order to perform this task, we must reason jointly about candidate facts and their associated extraction confidences, identify co-referent entities, and incorporate ontological constraints. Our proposed approach uses probabilistic soft logic (PSL), a recently introduced probabilistic modeling framework which easily scales to millions of facts. We demonstrate the power of our method on a synthetic Linked Data corpus derived from the MusicBrainz music community and a real-world set of extractions from the NELL project containing over 1M extractions and 70K ontological relations. We show that compared to existing methods, our approach is able to achieve improved AUC and F1 with significantly lower running time.

Subfields: Artificial Intelligence, Management Science and Operations Research

View Paper: https://www.semanticscholar.org/paper/b2a090506264bc9706dc9bcc5d61b4965ae919e7

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Artificial Intelligence

In Management Science and Operations Research

Citations in the corpus

Estimating Accuracy from Unlabeled Data (2014, methodology): “In this context, we could try using our estimates in order to develop a more robust co-learning framework, or in the case of NELL, use the estimated error rates in combination with a framework such as probabilistic soft logic (PSL) (Bröcheler et al., 2010; Pujara et al., 2013) in order to improve th”
Hinge-Loss Markov Random Fields and Probabilistic Soft Logic (2015, background): “Finally, we note that HL-MRFs and PSL have already been effectively applied to a wide variety of domains, including automatic knowledge base construction (Pujara et al., 2013), high-level computer vision (London et al.”
A Review of Relational Machine Learning for Knowledge Graphs (2015, background): “The resulting system can be scaled to fairly large knowledge bases, as shown in [111].”
BeliefBank: Adding Memory to a Pre-Trained Language Model for a Systematic Notion of Belief (2021, background): “Pujara et al. (2013) deﬁne “knowledge graph identiﬁcation” as the task of building a maximally consistent KG given noisy candidate facts and their extraction conﬁ-dences, and constraints between them.”
Robust Discovery of Positive and Negative Rules in Knowledge Bases (2018, background): “In order to be executed over a KB, or plugged into an existing inference system [17], rules must be manually crafted, a task that can be difficult for domain experts without a CS background.”
Hinge-Loss Markov Random Fields and Probabilistic Soft Logic (2015, background): “…and PSL has been demonstrated on many problems, including information extraction (Liu et al., 2016) and automatic knowledge base construction (Pujara et al., 2013), extracting and evaluating natural-language arguments on the Web (Samadi et al., 2016), high-level computer vision (London et…”
Never-Ending Learning (2015, methodology): “Recently, (Pujara et al. 2013) has demonstrated a more effective algorithm for the joint inference problem faced by the KI; we are now in the process of upgrading NELL’s KI to use this implementation.”
Enriching a Model's Notion of Belief using a Persistent Memory (2021, background): “In the area of knowledge graph construction, Pujara et al. (2013) define “knowledge graph identification” as the task of building a maximally consistent knowledge graph given noisy facts and their extraction confidences, and ontological constraints between them.”
Knowledge Graph Alignment Network with Gated Multi-hop Neighborhood Aggregation (2019, background): “As different KGs usually have heterogeneous schemas and data incompleteness (Pujara et al. 2013), the counterpart entities usually have dissimilar neighborhood structures.”
Towards a Definition of Knowledge Graphs (2016, background): “[17, 19]), and (b) the terms knowledge graph and knowledge base are used interchangeably (cf.”
Clinical Temporal Relation Extraction with Probabilistic Soft Logic Regularization and Global Inference (2020, neutral): “In recent years, PSL rules have been applied to various machine learning topics such as Fairness (Farnadi, Babaki, and Getoor 2019), Model Interpretability (Hu et al. 2016), Probabilistic Reasoning (Augustine, Rekatsinas, and Getoor 2019; Dellert 2020), Knowledge Graph Construction (Pujara et al. 20”
Domain-specific Knowledge Extraction from the Web of Data (2018, background): “PSL is proven to work well in such environments and has generated state-of-the-art results [Kouki et al. 2015] [Pujara et al. 2013].”
Knowledge Graphs (2020, neutral): “The growing industrial uptake of the concept proved difficult for academia to ignore, with more and more scientific literature being published on knowledge graphs in recent years [32, 77, 100, 105, 106, 140, 144].”
Knowledge vault: a web-scale approach to probabilistic knowledge fusion (2014, neutral): “, [34]); (2) methods that use latent variables to model the correlations indirectly, using either discrete factors (e.”