On the Reproducibility of the TAGME Entity Linking System

Hasibi, Faegheh; Balog, Krisztian; Bratsberg, Svein Erik

doi:10.1007/978-3-319-30671-1_32

Cited by 18 publications

(15 citation statements)

References 13 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Despite its usefulness, linked entities extracted from ELS's have issues because of low precision rates (Hasibi et al, 2016) and design challenges in training datasets (Ling et al, 2015). These issues can be summarized into two parts: ambiguity and coarseness.…”

Section: Possible Issuesmentioning

confidence: 99%

Entity Commonsense Representation for Neural Abstractive Summarization

Amplayo

Lim

Hwang

2018

Proceedings of the 2018 Conference of the North American Chapter Of the Association for Computational Linguistics: Hu

View full text Add to dashboard Cite

A major proportion of a text summary includes important entities found in the original text. These entities build up the topic of the summary. Moreover, they hold commonsense information once they are linked to a knowledge base. Based on these observations, this paper investigates the usage of linked entities to guide the decoder of a neural text summarizer to generate concise and better summaries. To this end, we leverage on an off-the-shelf entity linking system (ELS) to extract linked entities and propose Entity2Topic (E2T), a module easily attachable to a sequence-to-sequence model that transforms a list of entities into a vector representation of the topic of the summary. Current available ELS's are still not sufficiently effective, possibly introducing unresolved ambiguities and irrelevant entities. We resolve the imperfections of the ELS by (a) encoding entities with selective disambiguation, and (b) pooling entity vectors using firm attention. By applying E2T to a simple sequenceto-sequence model with attention mechanism as base model, we see significant improvements of the performance in the Gigaword (sentence to title) and CNN (long document to multi-sentence highlights) summarization datasets by at least 2 ROUGE points.

show abstract

Section: Possible Issuesmentioning

confidence: 99%

Entity Commonsense Representation for Neural Abstractive Summarization

Amplayo

Lim

Hwang

2018

Proceedings of the 2018 Conference of the North American Chapter Of the Association for Computational Linguistics: Hu

View full text Add to dashboard Cite

show abstract

“…While there is much modern work on entity linking, constructing an entity linker for novel types of entities or novel languages still requires non-trivial engineering and research effort [19,15]. In particular, although recent TAC entity linking work has focused on multilingual efforts, the organizers of this challenge observe that a 90% linking accuracy requires that about 20,000 query mentions are labeled for training.…”

Section: Related Workmentioning

confidence: 99%

“…This is important because although there has been substantial research on entity linking [23,13], constructing an entity linker for novel types of entities or novel languages still requires non-trivial engineering and research effort [19,15].…”

Section: Entity-taggingmentioning

confidence: 99%

“…We found that running the recommended entity linker was computationally prohibitive for our needs. While other entity linkers might be more efficient, generalizing to new languages is an open research area [18,19,15]. In particular: the organizers of TAC-KBP observe that a 90% linking accuracy requires that about 20,000 query mentions be labeled for training [19].…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Improving Entity Ranking for Keyword Queries

Foley

O’Connor

Allan

2016

Proceedings of the 25th ACM International on Conference on Information and Knowledge Management

View full text Add to dashboard Cite

Knowledge bases about entities are an important part of modern information retrieval systems. A strong ranking of entities can be used to enhance query understanding and document retrieval or can be presented as another vertical to the user.Given a keyword query, our task is to provide a ranking of the entities present in the collection of interest. We are particularly interested in approaches to this problem that generalize to different knowledge bases and different collections. In the past, this kind of problem has been explored in the enterprise domain through Expert Search. Recently, a dataset was introduced for entity ranking from news and web queries from more general TREC collections.Approaches from prior work leverage a wide variety of lexical resources: e.g., natural language processing and relations in the knowledge base. We address the question of whether we can achieve competitive performance with minimal linguistic resources.We propose a set of features that do not require index-time entity linking, and demonstrate competitive performance on the new dataset. As this paper is the first non-introductory work to leverage this new dataset, we also find and correct certain aspects of the benchmark. To support a fair evaluation, we collect 38% more judgments and contribute annotator agreement information.

show abstract

“…Finally, two different disambiguation algorithms are employed to link the correct Wikipedia page with the entity. In a similar way, Tagme and Spotlight extract and link entities to a knowledge base [11][12][13]. The major difference is that Spotlight uses DBpedia as its knowledge base.…”

Section: Entity Extractionmentioning

confidence: 99%

Construction of an Industrial Knowledge Graph for Unstructured Chinese Text Learning

et al. 2019

View full text Add to dashboard Cite

The industrial 4.0 era is the fourth industrial revolution and is characterized by network penetration; therefore, traditional manufacturing and value creation will undergo revolutionary changes. Artificial intelligence will drive the next industrial technology revolution, and knowledge graphs comprise the main foundation of this revolution. The intellectualization of industrial information is an important part of industry 4.0, and we can efficiently integrate multisource heterogeneous industrial data and realize the intellectualization of information through the powerful semantic association of knowledge graphs. Knowledge graphs have been increasingly applied in the fields of deep learning, social network, intelligent control and other artificial intelligence areas. The objective of this present study is to combine traditional NLP (natural language processing) and deep learning methods to automatically extract triples from large unstructured Chinese text and construct an industrial knowledge graph in the automobile field.

show abstract

On the Reproducibility of the TAGME Entity Linking System

Cited by 18 publications

References 13 publications

Entity Commonsense Representation for Neural Abstractive Summarization

Entity Commonsense Representation for Neural Abstractive Summarization

Improving Entity Ranking for Keyword Queries

Construction of an Industrial Knowledge Graph for Unstructured Chinese Text Learning

Contact Info

Product

Resources

About