LearningPinocchio: adaptive information extraction for real world applications

Ciravegna, Fabio; Lavelli, Alberto

doi:10.1017/s135132490400333x

Cited by 23 publications

(20 citation statements)

References 7 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…The use of machine learning has been envisaged to provide a solution capable of overcoming the potential domain-dependencies of rule-based IE systems (Moens 2006;Ciravegna and Lavelli 2004). Machine Learning developed out of Artificial Intelligence research, by which algorithms are designed that enable computers to 'adapt' to external conditions.…”

Section: Machine Learning Information Extraction Systemsmentioning

confidence: 99%

Archaeological Discriminant Applications of the Lubischew Test

Guerrero¹,

Serrano²,

Jiménez‐Arenas³

2015

Mathematics and Archaeology

View full text Add to dashboard Cite

IntroductionArchaeologists generate large quantities of text, ranging from unpublished technical fieldwork reports (the 'grey literature') to synthetic journal articles. However, the indexing and analysis of these documents can be time consuming and lacks consistency when done by hand. It is also rarely integrated with the wider archaeological information domain, with bibliographic searches having to be undertaken independently of database queries, for example. Text mining offers a means of extracting information from large volumes of text, providing researchers with an easy way of locating relevant texts and also of identifying patterns in the literature. In recent years techniques of Natural Language Processing (NLP) and its subfield, Information Extraction (IE), have been adopted to allow researchers to find, compare and analyse relevant documents, and to link them to other types of data. This chapter introduces the underpinning mathematics and provides a short presentation of the algorithms and distance measures used, from the point of view of artificial intelligence and computational logic. It describes the different NLP schools of thought and compares the pros and cons of rule-based vs machine learning approaches to information extraction. The role of ontologies and named entity recognition will be discussed and the chapter demonstrates how IE can provide the basis for semantic annotation and how it contributes to the construction of a semantic web for archaeology. The authors have worked on a number of projects that have employed techniques from NLP and IE in Archaeology, including Archaeotools, STAR and STELLAR. The chapter describes the archaeological user needs requirement, drawing examples from several countries, and the authors present examples drawn from their own projects, and previous work by others, of how NLP and IE can contribute to addressing this need. The problems and challenges of employing text mining in the archaeological domain are discussed, as well as the potential benefits.

show abstract

Section: Machine Learning Information Extraction Systemsmentioning

confidence: 99%

Archaeological Discriminant Applications of the Lubischew Test

Guerrero¹,

Serrano²,

Jiménez‐Arenas³

2015

Mathematics and Archaeology

View full text Add to dashboard Cite

show abstract

“…Knowledge-driven systems (e.g. LearningPinocchio [13], PANKOW [12], KIM [27]) use unstructured, i.e., gazetted, or structured, i.e., ontological, background knowledge to locate the entities of interest in a text document.…”

Section: Related Workmentioning

confidence: 99%

Aggregating semantic annotators

2013

View full text Add to dashboard Cite

A growing number of resources are available for enriching documents with semantic annotations. While originally focused on a few standard classes of annotations, the ecosystem of annotators is now becoming increasingly diverse. Although annotators often have very different vocabularies, with both high-level and specialist concepts, they also have many semantic interconnections. We will show that both the overlap and the diversity in annotator vocabularies motivate the need for semantic annotation integration: middleware that produces a unified annotation on top of diverse semantic annotators. On the one hand, the diversity of vocabulary allows applications to benefit from the much richer vocabulary available in an integrated vocabulary. On the other hand, we present evidence that the most widely-used annotators on the web suffer from serious accuracy deficiencies: the overlap in vocabularies from individual annotators allows an integrated annotator to boost accuracy by exploiting inter-annotator agreement and disagreement.The integration of semantic annotations leads to new challenges, both compared to usual data integration scenarios and to standard aggregation of machine learning tools. We overview an approach to these challenges that performs ontology-aware aggregation. We introduce an approach that requires no training data, making use of ideas from database repair. We experimentally compare this with a supervised approach, which adapts maximal entropy Markov models to the setting of ontology-based annotations. We further experimentally compare both these approaches with respect to ontologyunaware supervised approaches, and to individual annotators.

show abstract

“…Technologies suitable for large scale processing of text in knowledge management environments (e.g., Iria & Ciravegna, 2006;Ciravegna & Lavelli, 2004) will be enhanced and adapted to the social web requirements of the current era.…”

Section: Progress Over Related Scientific Workmentioning

confidence: 99%

“…Information extraction over sparse distributed documents (e.g., the Web), integrating information from documents with information from metadata and Web 2.0 tags as well. Technologies for large scale processing of text in knowledge management environments (e.g., Iria & Ciravegna, 2006;Ciravegna & Lavelli, 2004) will be enhanced and adapted to the current social web requirements. 2.…”

Section: Text Analysismentioning

confidence: 99%

Efficient Media Exploitation Towards Collective Intelligence

Mylonas

Solachidis²,

Geyer-Schulz³

et al. 2009

Advances in Data Analysis, Data Handling and Business Intelligence

View full text Add to dashboard Cite

In this work we propose intelligent, automated content analysis techniques for different media to extract knowledge from the multimedia content. Information derived from different sources/modalities will be analyzed and fused, in terms of spatiotemporal, personal and even social contextual information. In order to achieve this goal, semantic analysis will be applied to the content items, taking into account the content itself (e.g., text, images and video), as well as existing personal, social and contextual information (e.g., semantic and machine-processable metadata and tags). The above process exploits the so-called "Media Intelligence" towards the ultimate goal of identifying "Collective Intelligence", emerging from the collaboration and competition among people, empowering innovative services and user interactions. The utilization of "Media Intelligence" constitutes a departure from traditional methods for information sharing, since semantic multimedia analysis has to fuse information from both the content itself and the social context, while at the same time the social dynamics have to be taken into account. Such intelligence provides added-value to the available multimedia content and renders existing procedures and research efforts more efficient.

show abstract

LearningPinocchio: adaptive information extraction for real world applications

Cited by 23 publications

References 7 publications

Archaeological Discriminant Applications of the Lubischew Test

Archaeological Discriminant Applications of the Lubischew Test

Aggregating semantic annotators

Efficient Media Exploitation Towards Collective Intelligence

Contact Info

Product

Resources

About