Current context
MBL / machine learning in computational linguistics
First applications of MBL to language:
-
NETtalk (Sejnowski and Rosenberg, 1986; Stanfill and Waltz, 1986; Lehnert, 1987; Wolpert, 1990; Weijters, 1991)
-
Machine translation (Nagao, 1985)
More recent applications of memory-based learning:
-
Shallow/partial parsing (Argamon, Dagan, and Krymolowski, 1998; Cardie
and Pierce, 1998; Daelemans, Veenstra, and Buchholz, 1999)
-
Word sense disambiguation (Ng and Lee, 1996; Mooney, 1996)
-
Domain-specific lexical tagging (Cardie, 1993, 1994)
-
Machine translation (Jones, 1996)
-
Text categorisation (Riloff and Lehnert, 1994)
-
Word pronunciation (Van den Bosch and Daelemans, 1993; Daelemans,
Gillis, and Durieux, 1994; Van den Bosch, 1997)
-
Word morphology (Van den Bosch, Daelemans, and Weijters, 1996; Van den
Bosch and Daelemans, 1999)
-
Part-of-speech tagging (Daelemans, Zavrel, Berck, and Gillis, 1996;
Zavrel and Daelemans, 1999)
-
Attachment and subcategorisation (Zavrel, Daelemans and Veenstra,
1998; Buchholz, 1998)
-
Phrase chunking (Veenstra, 1998; Tjong Kim Sang and Veenstra, 1999)
Statistical NLP:
-
Maximum entropy (Ratnaparkhi, 1998) - full shift-reduce parsing on top
of tagging and chunking
-
Winnow (Golding and Roth, 1997) - spelling correction
-
DOP (Bod, 1997) - full parsing
-
HMM (Brants, Skut, and Krenn)
-
PCFG
-
...
Other inductive learning (cf. Cardie and Mooney tutorial):
-
Parsing with decision trees (Schmid; Magerman; Hermjakob)
-
Machine translation with decision trees (Hermjakob)
-
Dialogue act classification with decision trees (Litman;
Andernach)
-
Semantic parsing with inductive logic programming (Zelle and Mooney; Muggleton et al.)
-
...
References