Abstract/Summary

This thesis set out to understand why there are context-deficiencies in stored data in information systems, and then developed a framework for building more context into the interface of information systems. Data stored in an information system is context lacking although it occurs in the environment often with associated context details such as “where”, “when”, “situation”, “how”, “who”, “why” and “what”. However, due to the limitations of current information systems, only limited context details mainly the “what” are usually captured as the “sign” and stored in the information system (as data fields with associated meta-data). The resultant context deficient data affects the understanding and usability of information when the data is subsequently retrieved from the information systems. The study draws on semiotics theory, human computer interaction and context-aware literature to define human information interface, context-based data and information, and developed a context-based interface framework to enhance human understanding and usability of data or information stored in an information system. The lack of context in the ontology of data, i.e. as a key ingredient in information systems, is highlighted whilst introducing models that demonstrate the interrelationships between context-based data and the quality of data, information and knowledge with individual culture and interface factors as mediators. Using design science research design, and a mixed methods approach, the problem of lack of context of data and information stored in information system and its impact on the quality of information and knowledge activities is demonstrated through three preliminary studies. A conceptual framework of human information interface, for information and knowledge activities, is first developed. Surveys and interviews are then used to demonstrate the mediating role of the human user and interface factors in human information interaction situations, leading to the refined context-based human information interface framework for knowledge activities. A series of models are established and validated through structural modelling and path analysis, interviews, and together with case studies from expert reviews, the potential utility, validity and applicability of the framework is assessed. The results of the study did not only confirm the significance of the semiotics inspired interface factors and individual culture, when humans interface with information, but their mediating roles as well in human interaction with computer-based information systems. Consequently, the quality of data, information, and knowledge were affected by human factors and interface factors; and the design of context-based data and information interface for information systems were found to have a significant impact on the quality of knowledge activities. However, over 80% of the mediation effect was accounted for by the human factor, which emphasised the significant role of the human user in achieving optimum understanding and usability of stored data or information in information systems. The thesis makes several contributions including the need for a paradigm shift from human computer interaction studies to human information interface studies in the face of the “disappearing computer”. The concepts of human information interface (HII), and human-knowledge interface (HKI), which hitherto has not been defined in literature were defined. Another unique contribution of this study is the proposed multi-dimensional approach to the storage of data in information systems, specifically in 3 levels to represent the “what”, “how” and “why” details to provide more context to enhance understanding and usability when data or information is retrieved from information systems. Other significant contributions are the introduction of the concepts of “context-store”, “meta-why” and “meta-how”, which together with the existing “meta-data” or “meta-what” can potentially help improve the human interface with information. The main limitation of this study was the inability to design and implement the context-based interface for information systems in a real situation during the life of this thesis. This does not invalidate the outcome of the study, and considerable practical implications and relevance of the HII framework, for design of information systems and corresponding databases (i.e. information modelling and architecture, machine learning and deep learning, intelligence of context-aware and ubiquitous computing systems) are highlighted. In addition, future research directions are outlined for consideration.