Content area
Abstract
Archival descriptions play multiple roles in understanding and managing records in archives. However, archival descriptions are not efficiently organized and presented on the Web. In particular, archival context is often lacking in current descriptions, which makes it difficult for users to search and access records. To address the challenges of current archival descriptions in web environments, the overarching question is asked: How can archival context be improved for searching, using and understanding in web environments? In addition, it is necessary to build a conceptual/operational model for representing and restructuring the archival context and provenance information of archival descriptions. Two research questions are formulated to address the two objectives of the study, which are 1) to identify and examine the essential components for archival context and their relationships in archival descriptions, and 2) to explore whether archival context can be explicitly represented and modeled using ontology technologies.
To answer the research questions, a multi-method is taken in three studies: 1) Study 1 is a content analysis of archival descriptions; 2) Study 2 is a methontology to develop an event-based ontology; and 3) Study 3 is a focus group study to evaluate the usefulness of the built ontology. In Study 1, sample archival descriptions are drawn from 40 archival institution sites. By conducting the content analysis, the ten essential entities from archival descriptions have been identified as Agent, Artifact, Event, Feature, Function, Place, Relation, Role, Situation, and Time. In Study 2, the results of Study 1 were used to build the ontology model. By taking a methontology approach, five stages are performed: to identify ontology requirements at the specification stage, to identify relationships and concepts at the conceptualization stage, to develop the ontology at the implementation stage, to align with other ontologies based on the Basic Formal Ontology at the formalization stage, and to validate the ontology at the evaluation stage. As a result, an Event Ontology for Archival Context (EOAC) is developed with 233 classes and 87 properties, which were explicitly formalized in OWL (Web Ontology Language) for implementation. In Study 3, two focus group sessions with users and professional archivists were conducted to elicit both users’ and professional opinions on the current status of archival descriptions and the ontology model. Participants explained the advantages and disadvantages of the current descriptions and the ontology model and ten out of eleven participants were willing to adopt the event ontology model if supporting tools were available.
The findings of the three studies led to the development of a comprehensive conceptual model for archival context. The model consists of three levels: (1) the micro-level with static and fluid views; (2) the meso-level with three separate contexts (i.e., creation, description, and usage contexts); and (3) the macro-level with a series of activity systems. By integrating all three levels, the EOAC provides an explicit and comprehensive way to transform archival context into more structured, connected and interactive forms of entities.
This research adopts an innovative event-centred approach and emphasizes that events are as important as objects. The three-level conceptual model for archival context provides a framework to deepen the understanding of archival context. The EOAC ontology provides a new way to transform archival contexts into Linked Open Data, which can be utilized to improve searches and retrievals of archival descriptions and meet the expectations of users.





