Abstract :
Regional language is one of the cultures that
exist in Indonesia with a variety of regions throughout the
archipelago. The Komering language is a local language
spoken by the Komering tribe in South Sumatra province. The
Komering language is used to speak in conveying daily words
to the permanent residents of the Komering area. In fact, the
Komering language in various villages has different types of
speech delivery variations with different sounds and
vocabulary, but the Komering language remains one family,
namely the Komering language. The pronunciation of the
Komering language is Ogan Komering Ilir, which is a
permanent resident along the Komering river, covering
Tanjung Lubuk sub-district and Kayu Agung city. To improve
the preservation of this Komering regional language, this
research needs to use a Machine Learing technology-based
translator system media in order to help translate Komering
language into Indonesian and vice versa. the existence of a
Komering regional language translation machine that follows
the development of the world of technology will make it easier
to find out the translation of Komering language into
Indonesian and vice versa. Therefore, the researchers used the
Encoder-Decoder algorithm. The encoder-decoder algorithm is
one of the machine learning architectures commonly used for
machine translation. This algorithm is divided into two parts,
the encoder serves to learn the input document and then
process it into a translation for each Komering language word.
Decoder serves to take the input data representation generated
by the encoder and process it into translation results to
Indonesian and vice versa.
Keywords: Komering L