PASCAL - Pattern Analysis, Statistical Modelling and Computational Learning

Overview and results of Morpho Challenge 2010
Mikko Kurimo, Sami Virpioja and Ville Turunen
In: Proceedings of the Morpho Challenge 2010 Workshop Technical Report TKK-ICS-R37, Helsinki University of Technology, Department of Information and Computer Science . (2010) Aalto University , Espoo, Finland .

Abstract

In Morpho Challenge 2010 unsupervised and semi-supervised algorithms are sug- gested to provide morpheme analyses for words in different languages that can be used in various practical applications. Morpheme analysis is particularly use- ful in speech recognition, information re- trieval and machine translation for mor- phologically rich languages where the amount of different word forms is very large. The evaluations in Morpho Chal- lenge consist of: 1. comparisons to gram- matical morphemes, 2. information re- trieval experiments based on morphemes instead of words, and 3. machine trans- lation experiments where morpheme and word based systems are combined. The evaluation languages are: Finnish, Turk- ish, German, and English. This overview paper describes the goals, data, tasks, par- ticipants, evaluations, and obtained re- sults. The Morpho Challenge is part of the EU Network of Excellence PASCAL2 Challenge Program.

EPrint Type:Book Section
Project Keyword:Project Keyword UNSPECIFIED
Subjects:Computational, Information-Theoretic Learning with Statistics
Natural Language Processing
Speech
Theory & Algorithms
Information Retrieval & Textual Information Access
ID Code:7283
Deposited By:Mikko Kurimo
Deposited On:16 March 2011