Third Workshop on African Language Technology (AfLaT 2011) - Report

AfLaT2011, the Third Workshop on African Language Technology, was organized as a breakout session of the AGIS11 conference (Action Week for Global Information Sharing) in Addis Ababa, Ethiopia. It marked the first time an AfLaT workshop took place on the African continent. AfLaT2011 featured ten presentations on a variety of topics and languages.
We want to thank the presenters and participants for their contributions to the workshop, as well as the organizers of AGIS11 to allow us to collocate the AfLaT workshop with their wonderful conference.

Bilingual Data Mining for the English-Amharic Statistical Machine Translation (EASMT)

Mulu Gebreegziabher

Mulu Gebreegziabher and colleagues describe the collection of a parallel corpus English - Amharic. Difference sources for parallel data are described, as well as accuracy scores on document and sentence alignment. Very encouraging preliminary scores on statistical machine translation for this language pair are reported as well.
[Full Abstract]

Document Clustering in Amharic

Yalemisew Mintesinot Abgaz

Document classification for Amharic is described by Yalemisew Abgaz, using an unsupervised technique, based on TF/IDF metrics. Normalization, filtering and stemming are described as necessary preprocessing steps for this technique and the quality of the approach is empirically validated.
[Full Abstract]

IFE-MT: An English-to-Yorùbá Machine Translation System

Tunji Odejobi

Based on a limited, but important subset of the Yorùbá language, this machine translation system uses syntactic tree structures to enable the mapping from English into Yoruba. A slick and user-extendible interface for this system is presented as well.
[Full Abstract]