A Novel Path-Based Entity Relatedness Measure for Efficient Collective Entity Linking

Abstract

Collective entity linking is a core natural language processing task, which consists in jointly identifying the entities of a knowledge base (KB) that are mentioned in a text exploiting existing relations between entities within the KB. State-of-the-art methods typically combine local scores accounting for the similarity between mentions and entities, with a global score measuring the coherence of the set of selected entities. The latter relies on the structure of a KB: the hyperlink graph of Wikipedia in most cases or the graph of an RDF KB, e.g., BaseKB or Yago, to benefit from the precise semantics of relationships between entities. In this paper, we devise a novel RDF-based entity relatedness measure for global scores with important properties: (i) it has a clear semantics, (ii) it can be calculated at reasonable computational cost, and (iii) it accounts for the transitive aspects of entity relatedness through existing (bounded length) property paths between entities in an RDF KB. Further, we experimentally show on the TAC-KBP2017 dataset, both with BaseKB and Yago, that it provides significant improvement over state-of-the-art entity relatedness measures for the collective entity linking task.

Semantic relatedness in DBpedia: A comparative and experimental assessment

“…• Analyzing the first 40 pages of results (400 in total), and by selecting the articles about semantic relatedness methods, according to the above criteria. This search allowed us to identify 4 methods, namely, Linked Data Semantic Distance with Global Normalization (here referred to as LDSDGN) [44], Propagated Linked Data Semantic Distance (PLDSD) [4], Exclusivity-based measure (here referred to as ExclM) [27], and ASRMP m [13].…”

“…ASRMP m . In [13], El Vaigh et al propose the ASRM P m family of relatedness measures, originating from a previous proposal of the authors, referred to as Weighted Semantic Relatedness Measure (W SRM ) [12]. They state that a well-founded relatedness measure should meet the following three requirements: (i) to have a formal semantics in order to be defined on a knowledge graph such as RDF or OWL (as opposed to Wikipedia), (ii) to have a reasonable computational cost, (iii) to be transitive, in order to capture directly or indirectly related resources, and symmetric.…”

“…Therefore, for each Wikipedia article a corresponding RDF resource exists. Note that, a significant part of DBpedia comes from the information in the infoboxes of the Wikipedia articles 13 . The infobox contains data represented as property-value pairs provided by articles' editors that are, in general, an excerpt of the relevant information of a DBpedia resource.…”

“…How much are two given words related? In general, the way of automatically computing a degree of relatedness between words falls into one of the following categories of methods [22]: corpus-based methods, which use large corpora of natural language texts, and exploit co-occurrences of words, as for instance [47]; knowledge-based methods, which rely on structured resources, as for instance [13]; and hybrid methods, which are a mix of the two, as for instance [38]. Corpus-based methods benefit from the huge availability of textual documents and the advancements in the field of natural language processing and, for this reason, they have been widely investigated in the literature for a long time.…”

A simple semantic ranking approach for entity linking

“…However, these methods resolve mentions independently so that ignore the interaction among mentions [22,24] . Hence, another line of work turn to the collective methods, which disambiguate all mentions within the same document simultaneously to obtain a set of topically coherent target entities, and they have been presented more effective than the former ones [3,6,17,21] . Unfortunately, the collective models usually undergo a complex process and have a high time complexity.…”