An intelligent interface for online interaction

Abstract

In this paper, the author discusses the ways of improving the performance of online retrieval systems by introducing an automated interface between the enquirer and the system. In the first part of the paper, the main features of such human/machine interaction and the characteristics that the user would like to see incorporated in an interface, are described. Then, studies in artificial intelligence that are particularly relevant to the problems of implementing an intelligent interface, are discussed. The author concludes with a summary of automated mechanisms that will be needed to improve the quality of interaction between the user and the search system.Interactive computer information search normally consists of an enquirer who, often through the agency of an intermediary, interacts with a computer information system. This system usually contains files of information records, word indexes to those files, relationships between words (thesauri) and statistical data about the words (numbers of postings). Search is normally carried out by an iterated series of Boolean word combinations, and by user evaluation of selected intermediate outputs. One way to improve the performance of retrieval systems will be to change the content of the information records and/or the methods of search (eg., to use weighting, clustering, or probabilistic retrieval). An alternative way is to introduce an automated interface between the enquirer and the system, so as to improve the quality of human/system interaction. Of course, both methods of improvement can be introduced together.In this paper, I want to discuss the second approach, setting out what I consider to be the main features of this interaction and the characteristics that the user would like to see incorporated in an interface. Then will draw upon a review conducted by Belkin and myself [73] for clues as to how interface problems could be tackled. I will briefly summarise recent work in information retrieval on implementing an interface, and discuss studies in artificial intelligence that seem to me to be particularly relevant to the problem. I will conclude with a summary statement of the automated mechanisms that will be needed to reach the envisaged goal. It will be assumed that the interface will be designed to cover only a limited subject domain.The development of online information services and the proliferation of databases created a new professional role to be played by the librarian or information scientist -the role of an intermediary who knows where and how to search information systems in order to find answers to requests of end-users. The intermediary is an integral part of an interactive system between the user, the computer (including the query language) and the knowledge base (databases). He/she is responsible for a search that is cost-effective and satisfies the user's need. The processors of information, such as Dialog, SDC, BRS or Data-Star are interested in multiplying the number of searches (for better profits) and envisage the realisation of their goal...

Human machine interactions: from past to future- a systematic literature review

“…The years between 1984 and 2022 saw significant advancements in the field of HMIs as a result of the rapid development of AI and computing technologies. The advent of graphical user interfaces at the beginning of this time period greatly enhanced the overall quality of the user experience and made computing more accessible to everyone (Vickery, 1984; Babel, 2022).…”

“…To our knowledge, no comprehensive research on the relevance of HMIs that uncover history from the past to the present while simultaneously covering future possibilities has Systematic literature review been conducted. There have been conceptual research (Vickery, 1984;Bi and Salvendy, 1994;Hoc, 2000;Seifert et al, 2022), case studies (Kattel et al, 2019;Rezaev and Tregubova, 2018), empirical testing (Lin and Cai, 2009;Lupu et al, 2015;Peruzzini et al, 2020) and exploratory research (Adriaensen et al, 2021;Dombi et al, 2022). SLR and bibliometric analysis techniques provide rigour by reducing researcher bias and enabling a more precise scope of review.…”

“…Organizations should endeavour to maximize output. In this regard, machines have yielded the requisite result by automating routine tasks untiringly and relentlessly (Vickery, 1984). There should be separation of planning and operational work, and managers must train humans and equip machines to perform operational work.…”

“…In the current scenario, AI and ML can prove to be game changers because AI has the ability to emulate human thinking and because robots can self-learn with ML technologies (Vickery, 1984). Apart from mechanistic and clerical jobs, AI and ML are foraying into many “human” jobs, such as disease diagnosis, language translation and customer support.…”