Article

µ-TBL Lite: A Small, Extendible Transformation-Based Learner

Authors:
Torbjörn Lager at University of Gothenburg
Torbjörn Lager
Download citation
To read the full-text of this research, you can request a copy directly from the author.

Abstract

This short paper describes - and in fact gives the complete source for - a tiny Prolog program implementing a flexible and fairly efficient Transformation-Based Learning (TBL) system.

No full-text available

Request Full-text Paper PDF

To read the full-text of this research,
you can request a copy directly from the author.

... The AMIRA tagset, for instance, uses one tag (RP) to cover a range of particles which are subdivided into eight subclasses (EMPH PART, EXCEPT PART, FOCUS PART, INTERROG PART, RC PART, NEG PART, PART, VERB PART) in the PATB; and it uses several tags to describe different kinds of verbs (VB, VBG, VBD, VBN, VBP) where the PATB just uses three (IV, PV, CV). In order to overcome these problems, we use transformation-based retagging (TBR) [13,14] to recover from the mismatches between the two tagsets. TBR collects statistics about the local context in which erroneous tags have been assigned, and attempts to find rules based on this information to apply after the original tagger has been run. ...
Improved POS-Tagging for Arabic by Combining Diverse Taggers
Conference Paper
Full-text available
  • Sep 2012
A number of POS-taggers for Arabic have been presented in the literature. These taggers are not in general 100% accurate, and any errors in tagging are likely to lead to errors in the next step of natural language processing. The current work shows an investigation of how the best taggers available today can be improved by combining them. Experimental results show that a very simple approach to combining taggers can lead to significant improvements over the best individual tagger. © 2012 IFIP International Federation for Information Processing.
View
Learning Predicate Insertion Rules for Document Abstracting
Conference Paper
  • Feb 2011
  • Lect Notes Comput Sci
The insertion of linguistic material into document sentences to create new sentences is a common activity in document abstracting. We investigate a transformation-based learning method to simulate this type of operation relevant for text summarization. Our work is framed on a theory of transformation-based abstracting where an initial text summary is transformed into an abstract by the application of a number of rules learnt from a corpus of examples. Our results are as good as recent work on classification-based predicate insertion.
View
ResearchGate has not been able to resolve any references for this publication.