Cross-Lingual Information Processing in Indigenous Language Archives

Abstract

Cross-lingual information processing is a higher state of research, but it can enable the Indigenous Language Archive to become more accessible, particularly digitally and in scholarly terms. This paper intends to work on the ways of bridging the gaps in language to enable multilingual indexing, retrieval, and analysis of archiving resources in Indigenous languages. We propose a cross-linguistic retrieval of resources, which we use on the principles of machine translation, language modelling, and semantic alignment, whereby queries can be asked in more dominant languages to retrieve resources. There is a lot of focus on the cultural and linguistic faithfulness of any translation and interpretation undertaken in Indigenous knowledge systems. Another issue involved in this research is the agrarian resources of Indigenous languages, where annotated corpora are exceedingly uncommon, and digitized resources are scarce. Our interaction relies on the incorporation of the language data and contextual metadata provided by the community, which increases the searchable and interoperable nature of archives across languages, opening up the digital archives. Assessment of the case studies in the collections of Indigenous languages chosen proved that there was greater accuracy in retrieving and accessing by the user in the store. This brings out the effectiveness of the model in the context proposed. This study in particular shows the effect of using previously marginalized technologies on the protection of linguistic diversity and equality in the preservation of culture and heritage in a multilingual digital context.