Using the Semantic Web as a Source of Training Data

Bizer, Christian; Primpeli, Anna; Peeters, Ralph

doi:10.1007/s13222-019-00313-y

Cited by 8 publications

(15 citation statements)

References 16 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…Among the over 98 billion RDF n-quads mentioned above, nearly 10% are related to products, and are described by schema.org vocabularies. As examples, previous studies (Meusel et al (2015); Bizer et al (2019)) showed that among all product offers, 95% had an n-quad related to their names, 65% had one for their description, 35% had one for their brand, and less than 10% had one for their category. We argue that such data can be used to create language resources that are potentially useful for a wide range of product data mining tasks.…”

Section: Introductionmentioning

confidence: 93%

“…Therefore, an RDF n-quad can be considered as describing a 'fact' that are instantiations of concepts, properties and relationships. Studies have shown that such data can be used to train models for various NLP tasks, such as event extraction (Foley et al (2015)) and entity linking (Bizer et al (2019)).…”

Section: Introductionmentioning

confidence: 99%

“…We argue that such data can be used to create language resources that are potentially useful for a wide range of product data mining tasks. However, despite a limited number of sporadic studies in this area (Meusel et al (2015); Bizer et al (2019); Zhang et al (2020)), there is a lack of concerted effort that studies 1) how can such data be used to create language resources to support other downstream tasks; and 2) what kind of downstream tasks can benefit from them (and to what extent).…”

Section: Introductionmentioning

confidence: 99%

See 2 more Smart Citations

An Exploratory Study on Utilising the Web of Linked Data for Product Data Mining

Zhang¹,

Song²

2021

Preprint

View full text Add to dashboard Cite

The Linked Open Data practice has led to a significant growth of structured data on the Web in the last decade. Such structured data describe real-world entities in a machine-readable way, and have created an unprecedented opportunity for research in the field of Natural Language Processing. However, there is a lack of studies on how such data can be used, for what kind of tasks, and to what extent they can be useful for these tasks. This work focuses on the e-commerce domain to explore methods of utilising such structured data to create language resources that may be used for product classification and linking. We process billions of structured data points in the form of RDF n-quads, to create multi-million words of product-related corpora that are later used in three different ways for creating of language resources: training word embedding models, continued pretraining of BERT-like language models, and training Machine Translation models that are used as a proxy to generate product-related keywords. Our evaluation on an extensive set of benchmarks shows word embeddings to be the most reliable and consistent method to improve the accuracy on both tasks (with up to 6.9 percentage points in macro-average F1 on some datasets). The other two methods however, are not as useful. Our analysis shows that this could be due to a number of reasons, including the biased domain representation in the structured data and lack of vocabulary coverage. We share our datasets and discuss how our lessons learned could be taken forward to inform future research in this direction.Keywords linked data • web of data • schema.org • natural language processing • nlp • data mining • product mining

show abstract

Section: Introductionmentioning

confidence: 93%

Section: Introductionmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

An Exploratory Study on Utilising the Web of Linked Data for Product Data Mining

Zhang¹,

Song²

2021

Preprint

View full text Add to dashboard Cite

show abstract

“…While originally applied in the context of relation extraction from text, it has been used for the task of augmenting a knowledge base from semi-structured web data, including web tables [4,10]. Bizer et al [2] make use of schema.org annotations extracted from 43 thousand e-shops to distantly supervise a deep neural network for product matching. To generate training pairs, they make use of generic product identifies that are often provided along the annotations.…”

Section: Related Workmentioning

confidence: 99%

Using Weak Supervision to Identify Long-Tail Entities for Knowledge Base Completion

Oulabi

Bizer

2019

Lecture Notes in Computer Science

Self Cite

View full text Add to dashboard Cite

Data from relational web tables can be used to augment cross-domain knowledge bases like DBpedia, Wikidata, or the Google Knowledge Graph with descriptions of entities that are not yet part of the knowledge base. Such long-tail entities can include for instance small villages, niche songs, or athletes that play in lower-level leagues. In previous work, we have presented an approach to successfully assemble descriptions of long-tail entities from relational HTML tables using supervised matching methods and manually labeled training data in the form of positive and negative entity matches. Manually labeling training data is a laborious task given knowledge bases covering many different classes. In this work, we investigate reducing the labeling effort for the task of long-tail entity extraction by using weak supervision. We present a bootstrapping approach that requires domain experts to provide a small set of simple, class-specific matching rules, instead of requiring them to label a large set of entity matches, thereby reducing the human supervision effort considerably. We evaluate this weak supervision approach and find that it performs only slightly worse compared to methods that rely on large sets of manually labeled entity matches.

show abstract

“…La propia web de Schema.org cuantifica el despliegue de su uso en 10 millones de sitios web. Esto supone alrededor de menos del 1% de la Web, pero que es usado en cerca del 30% de los dominios(Bizer;Primpeli;Peeters, 2019).…”

unclassified

Del mobile-first al data-first: schema.org, búsquedas zero-click y la incertidumbre sobre los asistentes de voz

Pérez¹,

Pastor-Sánchez²

2020

ThinKEPI

View full text Add to dashboard Cite

Del mobile-first al data-first: schema.org, búsquedas zero-click y la incertidumbre sobre los asistentes de voz From mobile-first to data-first: Schema.org, zero-click searches and the uncertainty about voice assistants Tomás Saorín; Juan-Antonio Pastor-Sánchez Saorín, Tomás; Pastor-Sánchez, Juan-Antonio (2020). "Del mobile-first al data-first: schema.org, búsquedas zero-click y la incertidumbre sobre los asistentes de voz". Anuario ThinkEPI, v. 14, e14f04.

show abstract

Using the Semantic Web as a Source of Training Data

Cited by 8 publications

References 16 publications

An Exploratory Study on Utilising the Web of Linked Data for Product Data Mining

An Exploratory Study on Utilising the Web of Linked Data for Product Data Mining

Using Weak Supervision to Identify Long-Tail Entities for Knowledge Base Completion

Del mobile-first al data-first: schema.org, búsquedas zero-click y la incertidumbre sobre los asistentes de voz

Contact Info

Product

Resources

About