Untitled

Abstract

In this paper, we consider advancing webscale knowledge extraction and alignment by integrating OpenIE extractions in the form of (subject, predicate, object) triples with Knowledge Bases (KB). Traditional techniques from universal schema and from schema mapping fall in two extremes: either they perform instance-level inference relying on embedding for (subject, object) pairs, thus cannot handle pairs absent in any existing triples; or they perform predicate-level mapping and completely ignore background evidence from individual entities, thus cannot achieve satisfying quality. We propose OpenKI to handle sparsity of Ope-nIE extractions by performing instance-level inference: for each entity, we encode the rich information in its neighborhood in both KB and OpenIE extractions, and leverage this information in relation inference by exploring different methods of aggregation and attention. In order to handle unseen entities, our model is designed without creating entity-specific parameters. Extensive experiments show that this method not only significantly improves state-of-the-art for conventional OpenIE extractions like ReVerb, but also boosts the performance on OpenIE from semi-structured data, where new entity pairs are abundant and data are fairly sparse.

Integrating Lexical Information into Entity Neighbourhood Representations for Relation Prediction

“…These approaches do not utilise graph information in the form of connections between the texts and can only extract relations explicitly mentioned in the texts. We note that the OpenKI model (Zhang et al, 2019), which we use as a baseline, outperforms these models (see Table 3).…”

“…We build on the Entity Neighbourhood Encoding (ENE) model proposed by Zhang et al (2019). We then combine our enhanced neighbourhood encodings with the more complex "dual attention" model coined as "OpenKI".…”

“…Zhang et.al. (Zhang et al, 2019) posit that the aggregated representations v agg subj (e) and v agg obj (e) provide ultra-fine grained type information about entities when playing the respective roles and observe that including entity type information into their models does not notably improve performance, suggesting that type information is already present. Taking inspiration from that, we propose combining these aggregated representations with text based entity embeddings v text e ∈ R T derived from entity names and descriptions.…”

“…Previous work following this approach draws on patterns in the curated knowledge graph in combination with the graph of entity mentions in texts, allowing prediction of new knowledge base relations (Riedel et al, 2013;Verga et al, 2015Verga et al, , 2017. Zhang et al (2019) extend this work by incorporating text predicates connecting entity mentions extracted using OpenIE tools (Fader et al, 2011; and introducing the concept of "entity neighbourhoods" consisting of the binary OpenIE predicates and knowledge base relations 1 that occur with a given entity as their subject or object. Drawing on the success of text based representations incorporated into entity recognition tasks (Gillick et al, 2019), we extend Zhang et.al.…”

“…A system that can perform this task has the ability to extend an arbitrary set of relational database tables with information extracted from a document corpus. OpenKi (Zhang et al, 2019) addresses this task through extraction of named entities and predicates via OpenIE tools then learning relation embeddings from the resulting entityrelation graph for relation prediction, outperforming previous approaches. We present an extension of OpenKi that incorporates embeddings of text-based representations of the entities and the relations.…”