Ad hoc compounds in English to Swahili machine translation

Research output: Working paperScientific

Abstract

English compounds are normally composed of two or more consecutive words. It is also possible to form compounds using dash as a connector between words. Such constructions are difficult to handle, because they are not in the lexicon, and their interpretation must be carried out using heuristic guessing. However, it is possible to handle the members of the compound separately, and guessing can be avoided. In the Technical Report No. 57 I discussed the problem in English to Finnish machine translation. In this report I will discuss it from the viewpoint of English to Swahili machine translation. Although the basic approach in both is the same, the implementation is language-specific
Original languageEnglish
Place of PublicationHelsinki
PublisherUniversity of Helsinki, Institute for Asian and African Studies
Number of pages22
Publication statusPublished - 2020
MoE publication typeD4 Published development or research report or study

Fields of Science

  • 6121 Languages

Cite this