PERC: A Personal Email Classifier

Genre Classification in Automated Ingest and Appraisal Metadata

“…ERPANET [11]) and Automatic Metadata Generation (AMG) at the Catholic University of Leuven( [2]), and the extraction of bibliographic information from medical articles, based on the detection of contiguous blocks and fuzzy pattern matching, is available from Medical Article Record System (MARS) ( [42]) developed at the US National Library of Medicine (NLM) ([30]). There have also been previous work on metadata extraction from scientific articles in postscript using a knowledge base of stylistic cues ( [19], [20]) and, from the language processing community, there have been results in automatic categorisation of emails ( [6], [24]), text categorisation ( [39]) and document content summarisation ( [43]). Other communities have used image analysis for information extraction from the Internet ( [3]), document white space analysis ( [9]), graphics recognition in PDF files ( [41]), and algorithms for page segmentation ( [40]).…”

Building a document genre corpus: a profile of the KRYS I corpus

“…We are also hoping to gather some information from users of the Corpus online 6 . Further labelling performed by volunteer classifiers from other background using the classification system available online 7 may also help to understand the extensibility of the results in this paper.…”

“…3 Past efforts in automated metadata extraction (e.g. [4], [6], [16], dc-dot metadata editor;4 [1], [7]) employ methods that often rely on structural elements or presentation styles found to be common among the documents. These structural elements or styles are closely bound to the genre of the document, hence, it seems reasonable that a better understanding of the genre of documents and how they are used in information search would be a key step in developing a broadly effective metadata extraction tool.…”

Detecting Family Resemblance: Automated Genre Classification

“…Previous work exists on the extraction of descriptive metadata extraction within specific domains or genres (e.g. MetadataExtractor, DC-dot, Automatic Metadata Generation, Thoma, 2001;Giuffrida, Shek, & Yang, 2000;Han, Giles, Manavoglu, Zha, Zhang, & Fox, 2000;Bekkerman, McCallum, & Huang, 2004;Ke, Bowerman, & Oakes, 2006;Sebastiani, 2002;and Witte, Krestel, & Bergler, 2005). However, a general tool has yet to be developed to extract metadata from documents of varied forms and subjects.…”