Anusaaraka: An Information-Based Approach to Machine Translation

Title: Anusaaraka: An Information-Based Approach to Machine Translation

Abstract: Anusaaraka is a machine translation system that converts text from one Indian language to another Indian language. It presents an image of the source text in a language similar to the target language. Some constructions from the source language, which do not have equivalents in the target language, spill over to the output. The system also uses special notation. Anusaaraka has been built for five pairs of languages: Telugu, Kannada, Marathi, Bengali, and Punjabi to Hindi. It is available for use through Email servers. Anusaaraka follows the principle of substitutibility and reversibility of strings produced. This implies preservation of information while going from a source language to a target language. For narrow subject areas, specialized modules can be built by putting subject domain knowledge into the system, which produce good quality grammatical output. However, the system may sometimes go wrong, but the output will still be useful.

Research Question: How can an information-based approach be used to improve the quality of machine translation between Indian languages?

Methodology: The researchers developed anusaaraka, a machine translation system that converts text from one Indian language to another Indian language. They used five pairs of languages: Telugu, Kannada, Marathi, Bengali, and Punjabi to Hindi. The system presents an image of the source text in a language similar to the target language, with some constructions spilling over to the output. Special notation is also used.

Results: The researchers found that anusaaraka follows the principle of substitutibility and reversibility of strings produced, implying preservation of information while going from a source language to a target language. For narrow subject areas, specialized modules can be built by putting subject domain knowledge into the system, which produce good quality grammatical output. However, the system may sometimes go wrong, but the output will still be useful.

Implications: The research suggests that an information-based approach can be used to improve the quality of machine translation between Indian languages. The use of special notation and the preservation of information while going from a source language to a target language are key factors in the success of the anusaaraka system. However, the system may sometimes go wrong, but the output will still be useful. This implies that further research and development are needed to improve the accuracy and naturalness of the output.

Link to Article: https://arxiv.org/abs/0308019v1 Authors: arXiv ID: 0308019v1