Research and Advances
Computing Applications Virtual extension

iCare Home Portal: An Extended Model of Quality Aging E-Services

  1. Introduction
  2. The iCare Home Portal Framework
  3. Theory Underlying the iCare Decision Dimension
  4. iCare e-Services
  5. iCare e-Service Ontology
  6. Implementation of the iCare Home Portal
  7. Definition of Metrics
  8. Performance Benchmarks
  9. Conclusion
  10. References
  11. Authors
  12. Footnotes
  13. Figures
  14. Tables
  15. Sidebar: Aging Service Technologies
  16. Table

The quality of life of senior citizens is a critical issue around the world today. According to the World Health Organization (, there were 600 million people aged 60 or above in 2000; there will be 1.2 billion by 2025 and 2 billion by 2050. In addition to the thousands of unnecessary deaths that occur every year, missed health care opportunities cost U.S. businesses more than US$1 billion in avoidable hospital bills and nearly 41 million work days, resulting in the loss of US$11.5 billion per year, according to statistics from the U.S. Department of Health and Human Services ( These figures mean that assisted living and home care has become a fast growing health care sector.

Scrutiny of the extant technologies for aging services (see Sidebar 1) reveals that they are invariably aimed at or electronic care (eCare), but overlook social and behavioral aspects. eCare is an emerging health care field that utilizes Web technologies. Angood2 identified three eCare trends that utilize the Internet: medical informatics (focused on information), telemedicine (focused on communication), and cybermedicine (focused on global networking technologies). These eCare trends share the belief that maintaining an independent lifestyle is socially important to the quality of life of seniors and caregivers and that it helps to reduce the potential health care costs that are associated with hospitalization or placement in a full-time care facility. Unfortunately, the existing health care or eCare services for seniors are mostly oriented toward clinical gerontology or neuropsychology. They overlook certain dimensions of quality, such as community involvement, consumer participation, and continuous quality improvement.

In this article, we describe an intelligent care (iCare) system that results in an eCare system that is ambient enough to address the social and behavioral aspects of aging services. This iCare system is characterized by iCare ontology that features ambient service accessibility, unbound information reachability, attentive personalized service provision, innovative lifestyle creation, precious digital memory, and seamless social connection. The objective of the system is to provide quality e-Services to the elderly anywhere and at any time via an iCare home portal.

Back to Top

The iCare Home Portal Framework

The issues associated with existing eCare technologies can be organized into three dimensions: environmental, physical, and relationship. The environmental dimension refers to the software and hardware (wired and wireless) required to deploy the iCare infrastructure. The physical dimension concerns the issues of independence, safety, and the quality of life of the aging. The relationship dimension focuses on social networks that feature meaningful human interactions. The iCare home portal framework expands these dimensions by adding decision as the fourth dimension (See Figure 1(a)). The decision dimension represents the collective decisions that are automated to determine the appropriate personalized e-Services, thus enabling the effective delivery of quality e-Services to the elderly. These decisions can be automatically and intelligently rendered using a set of delegated agents that represents the myriad participants who are involved in caring for the elderly.

Both the physical and relationship dimensions are fundamental to iCare, as they provide the contextual data required for reasoning during the collective decisions in the decision dimension. The physical dimension is concerned with formal caregiving in which professional caregivers (such as nurses or doctors) are involved. The aims of this dimension are to promote healthy behavior for problem prevention, to achieve early disease detection to establish disease signatures before symptoms become apparent, and to improve treatment compliance to help people recover smoothly. The relationship dimension, conversely, is concerned only with informal caregivers, such as friends, neighbors, coworkers, family members, content providers, etc.

Back to Top

Theory Underlying the iCare Decision Dimension

The novel value of iCare rests in its decision dimension, which represents the virtual collective decisions (involving numerous participants) that determine which e-Services (in terms of the service concepts defined below) are appropriate and can be tailored to the context of the elderly. The underlying theory of the iCare decision dimension is the Stacey Model.

The Stacey Model (also called the Certainty-Agreement Model) was proposed by Ralph Stacey in 1996.11 Its objective is to select appropriate management actions within a complex adaptive system that involves multiple participants based on the degrees of certainty and the levels of agreement (Table 1). Many different disciplines in collective decision-making correspond to different zones of decision problems. In accordance with the different degrees of certainty and agreement, the disciplines employed by iCare include garbage-can reasoning, brainstorming, and case-based reasoning.

In the simple zone of collective decision-making, decision problems are straightforward to plan and control (such as, close to certainty and agreement), and the case-based reasoning principle is adopted to resolve them. In the chaos zone, decision problems are difficult to handle without conformity (such as, far from certainty and agreement). The zone between the simple and chaos zones is called the “zone of complexity,” and it embraces the principles of garbage-can reasoning and brainstorming.

Collective decision-making is automatically triggered by the iCare home portal based on the needs information supplied by the elderly. Each zone of the collective decision is equipped with a self-learning mechanism (as indicated in Table 1) that is capable of continuous improvements in decision-making. In the iCare environment, a real person with a well-defined role, such as a family member, friend, doctor, or neighbor, is represented by an agent that assists and acts on his or her behalf to perform collective decision-making and some-how has the ability to adapt and learn. This agent is connected to the person in terms of the provision/modification of the knowledge bases it embodies. However, the elderly are also involved in the service provision by accepting or rejecting the e-Services that are collectively recommended.

In iCare, the decision dimension (such as, virtual collective decision-making by consumers) is a significant feature, as it captures collective user participation in the care goal (such as, the appropriate contextual care service) and the subsequent status of goal-fitting (such as, continuous care quality improvement). The Stacey Model is employed in this study as the theoretical foundation to resolve collective decision problems under the dimensions of certainty and agreement.

Back to Top

iCare e-Services

Based on care-taking and care-giving functions, and assuming that the elderly in question are not seriously ill, iCare e-services can be categorized into taking iCare e-Services and giving iCare e-Services, as follows.

  • Giving iCare e-Services: e-Services that engage the elderly to contribute as individuals or communities, thus unfolding two further sub-categories of e-Services (individual-centric and community-centric).
  • Taking iCare e-Services: e-Services that engage the elderly to attain four kinds of resources (physical, mental, combined, or informative), as follows.
  • Physical: for example, telemedicine e-Services that provide remote medical assistance.
  • Mental: for example, home-movie e-Services and connection-oriented e-Services that aim to aid lonely aging persons based on social connections.
  • Combined (mental plus physical): for example, therapy entertainment that is presented to elderly persons (individual-centric) or groups (community-centric).
  • Informative: for example, the provision of information (specialized, generic, or only for personal information management) to the elderly.

Back to Top

iCare e-Service Ontology

Figure 1(b) displays the ontology of the iCare e-Service, which consists of four primary concepts (home-portal, service, device, and participant). The home-portal is a smart interface that connects the elderly with service providers and other participants. The other three concepts are as follows.

  • The service concept is characterized by three high-level attributes (scope, source, and type).
  • Scope refers to the degree of heterogeneity and reachability that a service embodies (such as, how specialized and wide-ranging the e-Service is).
  • Source indicates the resource origins that the e-Service draws upon (for example, human beings or content providers).
  • Type specifies the form in which an e-Service is rendered. The form could be data-oriented (such as, text or multimedia), community-oriented (such as, a community channel for the elderly), personal information-oriented (such as, information management that pertains to the daily life of the aging), or information category-oriented (such as, specialized information such as medical or financial information).
  • The device concept refers to a set of devices (involved in the iCare e-Services) that can be divided into two categories (electronic and non-electronic).
    • Electronic refers to the electronic hardware involved in the iCare environment (for example, sensors, and handheld devices).
    • Non-electronic indicates the non-electronic equipment in the person’s surroundings (such as, myriad objects).
    • The participant concept is characterized by four attributes (role, relationship, profile, and preference).
    • Role indicates the character played by the participant involved in the iCare environment (such as, the elderly person, the caregiver, the family member, or the remote caregiver).
    • Relationship refers to the relationship between the participant and the elderly person.
    • Profile refers to the demographic profile of the participant.
    • Preference refers to the preferences of the participant (in the example of an elderly person, these include the set of e-Services that he or she consumes in the course of his or her habits and hobbies).

    In Figure 1(b), node refers to a concept described above, a dotted arrow line indicates an “is-a” relationship from a subclass concept to a class concept, and a solid arrow line denotes a “part-of” relationship from a partial concept to an integral concept.

    A home-movie e-Service example is utilized to explain how an e-Service can be described using the concepts defined in the iCare service ontology. This service provides a customized video that comprises photos, pictures, and videos (most likely provided by remote participants or the media). The objective of this e-Service is to entertain elderly people who are lonely by combining media provided by family members or friends. The characteristics of the home-movie service are as follows.

    • Source: an appropriate social circle based on the “relationship proximity” (high, medium, and low) of the elderly and their given recognized “mood status” (for example, happy or homesick).
    • Type: a data-oriented form that uses a multimedia format.
    • Scope: “profile-matched” heterogeneity and designated areas of reachability with respect to an elderly person’s preferences.

    Using the iCare service ontology, care e-Services can be well delimited according to aspects of the expanded eCare model. Furthermore, this study can construct agents (who are committed to this ontology) who engage in virtual collective decision-making using the iCare decision dimension.

    Back to Top

    Implementation of the iCare Home Portal

    To demonstrate the features of the iCare home portal, we have implemented a prototype with the Service Oriented Architectural (SOA) strategy, as depicted in Figure 2. There are three possible types of collective decision blocks for service recommendation: the case-based reasoning model, which provides regular decisions within a single need; the brainstorming model, which offers innovative decisions within a single need; and the garbagecan model, which renders appropriate decisions within multiple needs and resource constraints. Here, we demonstrate the brainstorming model.

    Brainstorming is a collective decision-making application that involves numerous participant types (with moderate degrees of certainty and agreement) to determine appropriate e-Services that are tailored to the context of the elderly. This type of collective decision-making is good for elderly people who are extroverted and favor innovative and interesting lifestyles. For instance, a collective decision could be derived via brainstorming in a situation in which the agents for a son, a daughter, and the family doctor aim to discover collectively an innovative e-Service for a rainy day (e.g., an indoor yoga exercise) that would have the same effect as an outdoor exercise the elderly person needs to perform routinely.

    Figure 3 demonstrates a snapshot of the iCare system running under the brainstorming model. The user value vector on the left-hand side represents the elderly person’s view of the relative importance of the aspects of family, health, friends, time, and money (from 7 to −7). The available resources (for example, time and money) of the elderly and their caregivers are exhibited at the bottom of the left-hand side. The iCare system also shows the current collective decision model (such as e-Brainstorming). The bottom of the right-hand side shows the elderly person’s dynamic contextual information (such as, time, place, temperature, and the members of the household). Finally, the detailed XML input/output information of the services in the current decision model is listed on the right-hand side.

    In the iCare Home Portal, a specific mechanism – Semantic Ideation Learning for Agent-Based e-Brainstorming – has been developed and integrated to represent the participants who engage in an idea generation session for care service recommendations. This mechanism provides an ideation protocol that resolves the problems associated with the three boundaries (understanding, cognitive, and endurance boundaries), and thus idea quantity could be considered to be the dominant measure of e-Brainstorming effectiveness.

    Assessing this agent-based e-Brainstorming mechanism within the service scope of iCare involves justifying the improvement in the number of ideas generated (such as group creativity) and the diversity of the ideas created in recommending innovative care services. An idea in the context of iCare represents the recommendation of an e-Service. Accordingly, the set of valued ideas generated is equivalent to the set of recommended e-Services.

    Without a loss of generality, an exemplar of the given experiment settings is given as follows: (1) a universe of eight e-Services (such as, possible ideas for services to be delivered) for the mental needs under consideration; (2) up to 10 possible roles (such as, agents representing son, or daughter) involved in e-Brainstorming, each of which embodies its own knowledge of the domain of mental needs (simply represented by the correspondent marks made with respect to the e-Services); (3) a bench-mark greedy mechanism that randomly selects an e-Service from the existing knowledge of relevant domains for service recommendations; and (4) two metrics, Average Service Types and Service Diversity Rate (see definition here), which are used to inspect the mechanism’s performance. Figure 3 shows the portal with the e-Brainstorming mechanism, and Table 2 presents the evaluation results.

    Back to Top

    Definition of Metrics

    Average Service Types = cacm5211_aa.gif and

    Service Diversity Rate = (Average Service Types/T) * 100%, where

    • N = total number of experiment iterations,
    • ti = number of Service Types generated from the i th experiment, and
    • T = Number of available e-Services.

    Back to Top

    Performance Benchmarks

    This study benchmarks the greedy model against the e-Brainstorming model by comparing the number of types and the diversity of the generated e-Services with the different numbers of roles (three, five, or seven) that are involved in the decision process to justify the capability of group creativity in e-Brainstorming. It can be seen from Table 2(a) that when the e-Brainstorming mechanism does not account for agent knowledge learning capabilities (for example, agents are without the function to forward valued ideas to the idea knowledge base of every session participant to expand domain knowledge as the ideation rounds go by), a relatively good performance of the e-Brainstorming mechanism is attained (the different numbers of participants involved) in both the Average Service Types (such as, 3.7 > 2.77 when roles = 3) and Service Diversity Rate (46.25% > 34.58% when roles = 3) in comparison with the greedy approach. However, the results also reveal that the performance gap becomes small when the number of roles is increased.

    When agents have the capability to gain knowledge in e-Brainstorming, the performance improvements increase (such as, Table 2(b) shows the increased improvements of a case with three participants). The number of Average Service Types increases according to the extended VIQ number (from 4.8 to 6.07 when the VIQ is 1 and 3, respectively). A VIQ is a specified number of new idea instances added to the idea knowledge base. The Service Diversity Rate percentage also increases. Finally, in Table 2(c) it can be seen that the mechanism’s performance improvement in the Service Diversity Rate is proportional to the number of participants involved (such as roles) and the number of ideation rounds (such as ideation rounds). For example, the Service Diversity Rates are 46.25%, 70%, and 83.33% when the number of roles is three, five, and 10, respectively, for the four ideation rounds in e-Brainstorming.

    In short, this application demonstrates how the iCare model (augmented with the decision dimension) engenders consumer participation (in the form of virtual collective decision-making by agents working on behalf of the son, daughter, or family doctor of an elderly person, who connects to the agents in terms of their knowledge provision) in the recommendation of appropriate contextual e-Services for the elderly (in the form of the generation of a set of valued ideas, with the recommended e-Service being the most valued idea). Furthermore, a greater number of participants and ideation rounds in the decision process results in a higher Service Diversity Rate and more Average Service Types; that is, the results demonstrate that the iCare model ensures not only greater consumer participation/community involvement, but also continuous quality improvements in e-Service recommendations.

    Back to Top


    As the worldwide elderly population is expanding much faster than that of the younger generation, electronic care for the elderly is increasingly popular. This study proposes an electronic iCare model that utilizes collective decision-making to underscore the desired care quality elements of consumer participation and continuous quality improvement. This model goes beyond environmental, physical, and relationship aspects to envision possible forms of iCare e-Services and the ontology required to empower agents to fulfill the collective decision process.

    The impacts of the iCare model are threefold. First, the iCare home portal provides the elderly with multiple types of care, a personalized service, and real-time monitoring features. Second, the iCare platform can discover and respond to the needs of the elderly by utilizing cultural and technological resources, thereby enriching their lives and earning their trust in the service. This helps to alleviate the burden on their families and society. Third, the iCare platform is capable of combining various application services (such as, portal sites, Web services, and home management) to form an eCare service industry chain that can provide more innovative, multi-oriented, and caring services to the elderly and to society as a whole.

    We hereby urge existing service providers to re-package their eCare content or services. We also call for action by governments around the world to improve the efficiency of their eCare infrastructures by developing and deploying the iCare home portal to ensure that society becomes a more welcoming place for the elderly once eCare infrastructures and services are in place.

    Back to Top

    Back to Top

    Back to Top

    Back to Top


    F1A Figure 1a. A Framework of iCare Home Portal

    F1B Figure 1b. iCare Ontology

    F2 Figure 2.

    F3 Figure 3.

    Back to Top


    T1 Table 1. Stacy Model equipped with three different principles employed for iCare virtual collective decision making

    T2 Table 2. Performance of the e-Brainstorming home portal

    Back to Top

    UT1-1 Table A. Comparisons among Environment-Oriented Technology, Care-Oriented Technology, and iCare Technology

      1. Adlam, T.D. and Orpwood, R.D. Taking the Gloucester Smart House from the laboratory to the living room. Proceedings of UbiHealth04 Conference, (2004).

      2. Angood, P.B. Telemedicine, the Internet, and World Wide Web: Overview, current status and relevance to surgeons. World Journal of Surgery 25, (2001), 1449–1457.

      3. Helal, S., Giraldo, C., Kaddoura, Y. and Lee, C. Smart phone based cognitive assistant. Proceedings of UbiHealth 2003, (Oct. 2003).

      4. Intille, S.S. Designing a home of the future. IEEE Pervasive Computing, (Apr.-June 2002), 80–86.

      5. Kautz, H., Arnstein, L., Borriello, G., Etzioni, O. and Fox, D. Towards personal service robots for the elderly. Proceedings of AAAI-2002 Workshop on Automation as Caregiver, (2002), 60–65.

      6. Kidd, C.D., Orr, R.J., Abowd, G.D., Atkeson, C.G., Essa, I.A., MacIntyre, B., Mynatt, E., Starner, T.E., and Newstetter, W. The aware home: A living laboratory for ubiquitous computing research. Proceedings of the Second International Workshop on Cooperative Buildings, Position paper, (Oct. 1999), 191–198.

      7. Laerhoven, K.V., Lo, B.P.L., Ng, J.W.P., Thiemjarus, S., King, R., Kwan, S., Gellersen H.W., Sloman, M., Wells, O., Needham, P., Peter, N., Darzi, A., Toumazou, C. and Yang, G.Z. Medical health care monitoring with wearable and implantable sensors. Proceedings of UbiHealth04 Conference, (2004).

      8. Pollack, M.E., Brown, L., Colbry, D., McCarthy, C., Peintner, B., Ramakrishnan, S. and Tsamardinos, I. Autominder: An intelligent cognitive orthotic system for people with memory impairment. Robotics and Autonomous Systems 44, (2003), 273–282.

      9. Roy, N., Baltus, G., Fox, D., Gemperle, F., Goetz, J., Hirsch, T., Margaritis, D., Montemerlo, M., Pineau, J., Schulte J. and Thrun, S. Towards personal service robots for the elderly. Proceeding of Workshop on Interactive Robots and Entertainment (WIRE 2000), Pittsburgh, PA, (2000).

      10. Seo, J.W. and Park, K. The development of a ubiquitous health house in South Korea. Proceeding of the 6th International Conference on Ubiquitous Computing (UbiComp 2004), Nottingham, England, Sept. (2004).

      11. Stacey, R.D. Complexity and Creativity in Organizations. Berrett Koehler, San Francisco, 1996.

      12. Tsai, T.M., Liu, J.T., and Hsu, Y.J. MiCare: Context-aware authorization for integrated health care service. Proceedings of the 6th International Conference on Ubiquitous Computing (Nottingham, England, Sept. 2004).


    Join the Discussion (0)

    Become a Member or Sign In to Post a Comment

    The Latest from CACM

    Shape the Future of Computing

    ACM encourages its members to take a direct hand in shaping the future of the association. There are more ways than ever to get involved.

    Get Involved

    Communications of the ACM (CACM) is now a fully Open Access publication.

    By opening CACM to the world, we hope to increase engagement among the broader computer science community and encourage non-members to discover the rich resources ACM has to offer.

    Learn More