Artificial intelligence at the service of digital archives: reconstituting archival aggregations and enriching metadata schemas

Abstract

This paper aims to present the results of a study conducted as part of the international InterPARES Trust AI project and entitled «The role of AI in identifying or reconstituting archival aggregations of digital records and enriching metadata schemas». The overall objective of this study is to investigate the ability of artificial intelligence to support the creation (or re-creation) of archival aggregations to solve the problem of the presence of unaggregated, unordered or de-contextualised documents (both in the current and semi-current phase of the archive) that arises in many situations. Often, in both public administrations and companies, documents are neither classified nor aggregated; or document aggregations are formed but incorrectly. Furthermore, metadata – which are necessary to guarantee authenticity, reliability and searchability – are not often correctly identified and associated with documents. As a result, the organisation’s archive is not properly formed, which is a severe problem because it leads to an uncontrolled number of documents that are unordered, misplaced and therefore difficult to find. Unfortunately, in spite of the progress made by information technology to provide help in document management activities, it must be recognised that current software products are able to provide very limited support for this type of need. However, artificial intelligence techniques seem to promise great progress in this field. The study aims to address these issues by providing answers to research questions such as the following: can artificial intelligence tools help to create document aggregations when these have never been formed or to re-create them when they had been formed but have been lost? Can they provide valuable help in identifying metadata and associating it with related documents?