Strategies for lifelong knowledge extraction from the web

By Michele Banko, Oren Etzioni. 2007

82 citations recorded by the source corpus.

Subfields: Artificial Intelligence, Information Systems

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Artificial Intelligence

In Information Systems

Citations in the corpus

Continual Lifelong Learning in Natural Language Processing: A Survey (2020, background): “, 2010) and an unsupervised ALICE (Banko and Etzioni, 2007) systems, which iteratively extract information and build general domain knowledge, were at the forefront of CL in NLP.”
Ordinal Common-sense Inference (2016, background): “ing from instance-level predications discovered in descriptive texts (Schubert, 2002; Liakata and Pulman, 2002; Clark et al., 2003; Banko and Etzioni, 2007).”
O N M EMORY IN H UMAN AND A RTIFICIAL L ANGUAGE P ROCESSING S YSTEMS (2020, unknown): “Work on using knowledge bases to improve AI systems has progressed from the early decades of AI with the design of rule-based expert systems (Jackson, 1990) to the modern era prior to the deep learning revolution (Banko & Etzioni, 2007; Mitchell et al., 2015).”
DELIS Dynamically Evolving , Large-scale Information Systems Integrated Project Member of the FET Proactive Initiative Complex Systems DELIS-TR-594 Yago : A Large Ontology from Wikipedia and WordNet (2007, methodology): “[15] use different domain search strategies for fact extraction and show an precision of around 80%.”
Toward an Architecture for Never-Ending Language Learning (2010, unknown): “In our experiments, CPL was given as input a corpus of 2 billion sentences, which was generated by using the OpenNLP package 1 to extract, tokenize , and POS-tag sentences from the 500 million web page English portion of the ClueWeb09 data set (Callan and Hoy 2009).”
Extracting implicit knowledge from text (2009, methodology): “(2007), Banko and Etzioni (2007), Paşca (2008)) have relied on the use of manually created word sense hierarchies, such as WordNet.”
YAGO: A Large Ontology from Wikipedia and WordNet (2008, methodology): “dge to drive the strategies for acquiring new facts [5].”
Machine Reading at the University of Washington (2010, methodology): “In general, potential bootstrap sources can be broadly divided into domain knowledge (e.g., patterns and rules) and crowdsourced contents (e.g., linguistic resources, Wikipedia, Amazon Mechanical Turk, the ESP game).”