Knowledge representation for diagnosis of care problems through an expert system: Model of the auto-care deficit situations M. Lourdes Jimenez a, * , Jose ´ M. Santamarı ´a b , Roberto Barchino a , Laura Laita c , Luis M. Laita d, * , Leo ´n A. Gonza ´lez a , Angel Asenjo e a Computer Science Department, Alcala University, Escuela Te ´cnica Superior de Ingenierı ´a Informa ´ tica, Ctra. Barcelona km. 31.7, 28805 Alcala ´ de Henares, Madrid, Spain b Gerencia A.P. A ´ rea11, Madrid, Spain c Hospital San Rafael, Madrid, Spain d A.I. Department, Universidad Polite ´ cnica de Madrid, Boadilla del Monte, Madrid, Spain e Nursing Department, University of Alcala ´ , Alcala ´ de Henares, Madrid, Spain Abstract The caregiver role strain is today an increasing problem because of the population aging; moreover, its diagnosis is highly difficult. In this article, we summarize the design of an Expert System for the diagnosis of this health problem. The Expert System Knowledge Base is composed by a set of production rules written in classic bi-valued logic and by a set of potential facts. In order to build this Knowledge Base it has been necessary to design previously a Model of the problem treatment. The Expert System Inference Engine uses Gro ¨ bner Bases and Normal Form to obtain the diagnosis from the information stored in the Knowledge Base. Furthermore, a Graphic User’s Interface has been implemented to make easier the access to the System by all kinds of users. Ó 2007 Elsevier Ltd. All rights reserved. Keywords: Knowledge-based systems; Computer algebra; Diagnosis model; Caregiver’s overload diagnosis 1. Introduction In order to improve theories as well as to obtain useful practical applications, it has become necessary in our society to establish interdisciplinary relations among different knowl- edge branches. It was under these premises that the collabo- ration between the Computer Science and the Nursing Departments of the University of Alcala ´ was settled down. The specific topic about people’s care which is the nurs- ing diagnosis known as ‘‘judgement about the response of a person, family or community, to vital processes of real or potential health problems, that provides the therapy basis for successful objectives of the tasks for which the nurse is the person in charge’’ was chosen for study (Herna ´ndez & Esteban, 2003). Due to its social importance, a fact supported by differ- ent arguments, like the aging of the population, depen- dency policies, the social change produced by women’s (who were the habitual caregivers in their family environ- ments) incorporation to the labour world, etc., the ‘‘care- giver role strain’’ (Nanda, 2001) was, among a great amount of possible aspects related to the above mentioned topic, the one chosen as the candidate for being the expert system’s subject. This care problem presents a high difficulty both in its assessment and its treatment. There are few approaches to its valuation; the most used is Zarit’s scale, Zarit (1983) that, although it has been validated, the arrange- ments of its elements (factors and symptoms) can be improved. Its (partial) shortcomings regarding diagnosis are basically based on the following facts: 0957-4174/$ - see front matter Ó 2007 Elsevier Ltd. All rights reserved. doi:10.1016/j.eswa.2007.05.039 * Corresponding authors. Tel.: +34 91 885 69 57; fax: +34 91 885 66 46 (M.L. Jimenez). E-mail addresses: lou.jimenez@uah.es (M.L. Jimenez), laita@fi.upm.es (L.M. Laita). www.elsevier.com/locate/eswa Available online at www.sciencedirect.com Expert Systems with Applications 34 (2008) 2847–2857 Expert Systems with Applications