Abstract
Knowledge Graph (KG) completion has been widely studied to tackle the incompleteness issue (i.e., missing facts) in modern KGs. A fact in a KG is represented as a triplet (h, r, t) linking two entities h and t via a relation r. Existing work mostly consider link prediction to solve this problem, i.e., given two elements of a triplet predicting the missing one, such as (h, r, ?). This task has, however, a strong assumption on the two given elements in a triplet, which have to be correlated, resulting otherwise in meaningless predictions, such as (Marie Curie, headquarters location, ?). In addition, the KG completion problem has also been formulated as a relation prediction task, i.e., when predicting relations r for a given entity h. Without predicting t, this task is however a step away from the ultimate goal of KG completion. Against this background, this paper studies an instance completion task suggesting r-t pairs for a given h, i.e., (h, ?, ?). We propose an end-to-end solution called RETA (as it suggests the Relation and Tail for a given head entity) consisting of two components: a RETA-Filter and RETA-Grader. More precisely, our RETA-Filter first generates candidate r-t pairs for a given h by extracting and leveraging the schema of a KG; our RETA-Grader then evaluates and ranks the candidate r-t pairs considering the plausibility of both the candidate triplet and its corresponding schema using a newly-designed KG embedding model. We evaluate our methods against a sizable collection of state-of-the-art techniques on three real-world KG datasets. Results show that our RETA-Filter generates of high-quality candidate r-t pairs, outperforming the best baseline techniques while reducing by 10.61%-84.75% the candidate size under the same candidate quality guarantees. Moreover, our RETA-Grader also significantly outperforms state-of-the-art link prediction techniques on the instance completion task by 16.25%-65.92% across different datasets.
Highlights
Knowledge Graphs (KGs), such as Freebase [5], Wikidata1 or Google’s Knowledge Graph2, have become a key resource powering a broad spectrum of Web applications, such as semantic search [48], questionanswering [51], or recommender systems [54]
To implement our instance completion task, for a test h, we first generate a set of candidate r -t pairs, and score and rank them
We first take the top N relations generated by a relation prediction technique and use one tail candidate refinement technique to generate a set of candidate r -t pairs
Summary
Knowledge Graphs (KGs), such as Freebase [5], Wikidata or Google’s Knowledge Graph, have become a key resource powering a broad spectrum of Web applications, such as semantic search [48], questionanswering [51], or recommender systems [54]. With a small set of predicted relations, the number of candidate r -t pairs fed to the link prediction technique can be significantly reduced Such an approach still shows subpar performance, as it fails to fully consider the triplewise correlation of the three elements in a triplet, in particular the schema information encoded in the entity-typed triplet (h_type, r, t_type). If we have the schema information represented as entity-typed triplets (h_type, r, t_type) —(enterprise, headquarters location, city) and (enterprise, industry, economic branch), we could filter out such noisy r -t pairs that do not match the schema of the KG (to the given h) Against this background and to effectively solve our instance completion problem over KGs (h, ?, ?), we propose an end-to-end solution fully leveraging schema information encoded in triplets. Our RETA-Grader significantly outperforms state-of-the-art link prediction techniques on the instance completion task by 16.25%-65.92% across different datasets
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a callDisclaimer: All third-party content on this website/platform is and will remain the property of their respective owners and is provided on "as is" basis without any warranties, express or implied. Use of third-party content does not indicate any affiliation, sponsorship with or endorsement by them. Any references to third-party content is to identify the corresponding services and shall be considered fair use under The CopyrightLaw.
