Int. J. Technology Management, Vol. 47, Nos. 1/2/3, 2009
Using ontology-based knowledge networks for user training in managing healthcare processes A. Macris* Department of Business Administration, University of Piraeus, Piraeus, Greece Fax: +30 210 4142339 E-mail:
[email protected] *Corresponding author
D. Papakonstantinou, F. Malamateniou and G. Vassilacopoulos Department of Informatics, University of Piraeus, Piraeus, Greece E-mail:
[email protected] E-mail:
[email protected] E-mail:
[email protected] Abstract: The cooperative and collaborative nature of healthcare delivery requires active user participation in healthcare process design/redesign. Hence, there is a need to provide users with reusable, flexible, agile and adaptable training material in order to enable them instil their knowledge and expertise in healthcare process modelling and automation activities. This paper presents a prototype approach for designing user training material which is based on externalising domain knowledge in the form of ontology-based knowledge networks. The approach provides significant advantages to both designers (knowledge–content reusability and Semantic Web enabling) and users (semantic search, knowledge navigation and knowledge dissemination). Keywords: healthcare process modelling; user training; ontology-based knowledge networks. Reference to this paper should be made as follows: Macris, A., Papakonstantinou, D., Malamateniou, F. and Vassilacopoulos, G. (2009) ‘Using ontology-based knowledge networks for user training in managing healthcare processes’, Int. J. Technology Management, Vol. 47, Nos. 1/2/3, pp.5–21. Biographical notes: A. Macris is an Assistant Professor in Management Information Systems at the Department of Business Administration, University of Piraeus, Greece. His current research interests are in the areas of knowledge management, e-learning and business intelligence. D. Papakonstantinou received her BSc Degree in Informatics from the University of Piraeus, Greece. She is currently pursuing her Doctoral Degree at the University of Piraeus, working in the area of healthcare information systems. F. Malamateniou received her PhD Degree in Health Informatics from the University of Athens, Greece. She is a research associate at the Health Informatics Laboratory at the University of Piraeus. Her current research Copyright © 2009 Inderscience Enterprises Ltd.
5
6
A. Macris et al. interests include electronic patient record, healthcare process management and healthcare systems security. G. Vassilacopoulos is a Professor in Information Systems in the Department of Informatics, University of Piraeus, Greece. His current research interests include healthcare information systems, information security, business process management and workflow systems.
1
Introduction
In recent years, business process management has emerged as an important discipline which is concerned with the understanding of, communication about and evolution of business processes in a variety of application domains, such as banking, healthcare, technology and logistics (Currie and Willcocks, 1996; Lenz and Kuhn, 2004; McAdam et al., 2005; Rutner et al., 2003). The topics addressed in this field cover areas like business process modelling, process mining and workflow management. Business process modelling is used for designing new processes or analysing existing processes with the objective of redesigning them to improve current work practices (Reijers and Mansar, 2005). Process mining facilitates the discovery of process specifications from process logs that are readily available in many organisations and, therefore, may support process design and redesign (Dustdar et al., 2005). Workflow management is used to automate business processes by coordinating the available resources with a more global perspective, taking into account the characteristics and workload of process participants (Haux et al, 2003). Thus, business process modelling constitutes a basic premise for both process mining and workflow management.
1.1 Healthcare process management: key objectives The drive in healthcare to simultaneously reduce costs and improve quality has created a shift from organisation-centred care to patient-centred care, requiring increased integration among functional units as well as enhanced cooperation and collaboration among disparate healthcare organisations. Hence, considerable attention has been paid to designing process models of healthcare delivery and to developing healthcare information systems that support intra- and inter-organisational healthcare processes (Andersson et al., 2003; Berg and Toussaint, 2003; Malamateniou et al., 2003; Poulymenopoulou et al., 2003; Vassilacopoulos and Paraskevopoulou, 1997). Healthcare process models convert the tacit knowledge contained in various healthcare processes into explicit knowledge through externalisation, thus allowing healthcare organisations to evaluate and improve their procedures (Nonaka and Takeuchi, 1995; Rinkus et al., 2005). Moreover, healthcare process modelling is an important activity of healthcare process management that is commonly considered to be the development of computerised applications that directly follow the execution logic of the underlying healthcare process, thus placing the resulting information system in its organisational context. In fact, the requirements posed by the problem domain may stimulate constructive interaction between users and analysts that often leads to the definition of a variety of healthcare process models with different goals in mind and
Using ontology-based knowledge networks for user training
7
different levels of resolution in their representation. Hence, one important consideration in healthcare process management is to enable and foster active user participation, since users are required to think of their activities, not in isolation, but as constituents of healthcare processes and, hence, to instil their knowledge and expertise in the definition and automation of healthcare processes (Vassilacopoulos et al., 1996). Enabling active user participation in managing healthcare processes requires changes in the users’ perception of their work (from function-oriented to process-oriented) and an in-depth understanding of each healthcare process (Barrow and Mayhew, 2000; Lee, 1986). Indeed, user participation has been strongly advocated by several studies as a way of ensuring that genuine user needs are satisfied, and for gaining user acceptance of a new system (Anderson, 1985; Cavaye, 1995; Dvir, 2005). The basic contention of these studies is that users who are actively involved in the system development process will likely influence design and implementation decisions in accordance with their needs and desires, resulting in a system that they perceive as useful and usable, and an implementation strategy that they perceive as suitable for the particular situation. Thus, users are expected to develop positive attitudes toward the new system, increasing the likelihood of system success.
1.2 Research problem Along with the adoption of a participative approach to healthcare process management there is some concern about designing and conducting appropriate user training courses with the objective to develop relevant user competencies so that users are enabled to become full and active members of the process management team. Hence, user participation will be concerned not only with making user knowledge explicit but also with developing user knowledge and producing new conceptions, (Cartwright et al., 2002; Hartman and Crow, 2002; Rinderle et al., 2005). In particular, properly designed user training courses should enable users to understand process-oriented healthcare delivery, visualise intra- and inter-organisational healthcare processes, assimilate the inherent logic underlying existing processes and identify areas where redesigning or streamlining existing processes is required to adapt to today’s complex and dynamic healthcare environment (Wald et al., 2001; Wieringa et al., 2003). The development of a process-oriented healthcare information system essentially involves redesigning existing healthcare processes with the objective of shifting the focus from traditional transactional processing to cooperation and collaboration (Poulymenopoulou et al, 2003; Reijers and Mansar, 2005). In this context, process participants are required to exert a constructive influence by instilling their knowledge and expertise into the process management activity. As an example, consider a healthcare process concerning radiological requests issued by health centre physicians to the radiological department of a hospital. This process involves issuing a radiological request by a health centre physician, receiving and scheduling the radiological procedure(s) requested by a radiologist, preparing and transferring the patient to the radiology department according to schedule, performing the radiological procedure(s) requested by the assigned radiologist and interpreting results and sending a radiological report to the requesting physician. Such inter-organisational healthcare processes (i.e., processes that cross organisational boundaries) form the basis for a collaborative environment by bringing together healthcare practitioners who are geographically dispersed. Hence, what gives value to such a healthcare process is the ability to create virtual healthcare
8
A. Macris et al.
workgroups organised around patient care. In turn, this improves decision making by sharing knowledge and, ultimately, the quality of care (Andersson et al., 2003; Malamateniou et al., 2003). The knowledge contained in such healthcare processes (e.g., flow of activities, resources involved, physical location, coordination requirements) and the data content (e.g., documents in paper and electronic form, data flows) must be made explicit through training, so that users understand the requirements of the new collaborative environment and participate actively in its development. Figure 1 depicts a transition from existing to redesigned healthcare processes (i.e., from the ‘as-is’ to the ‘to-be’ situation) where user training is shown as an important activity for enabling active user participation throughout the process management activity (i.e., in analysing existing processes and in designing new processes). Hence, user training is not intended to imply a once-through process, since the control action necessary at any training stage may involve significant iteration. Figure 1
Enabling user participation through training
From these considerations, the research problem addressed in this paper is concerned with developing a training aid that enables users to understand healthcare process modelling concepts, with the objective of enhancing and empowering user-to-analyst interaction in the context of a process-oriented healthcare information system development. To this end, an ontology-based knowledge network is developed that can be used as a tool for the semantic representation of healthcare process concepts and, hence, as a means for the development of an appropriate training aid. An illustrative example of an ontology-based knowledge network regarding a healthcare process is also presented.
2
Design of an ontology-based training aid
The proposed model is developed on the basis of the principles set by various initiatives concerning the Semantic Web, all having in common the focus of extending current web technology with machine-understandable metadata (Berners-Lee et al., 2001). Those metadata are stored in ontologies (Gruber, 1993) and play an essential role in Semantic Web, since they provide the shared conceptualisations expressed in a logical form. Web services are layered services, capable of exploiting the semantics provided by these metadata descriptions in order to expand the current capabilities of web technology (Sycara, 2004).
Using ontology-based knowledge networks for user training
9
The Semantic Web vision has been combined with the principles of knowledge transformation in order to provide a theoretical model of e-learning processes (Naeve et al., 2005; Nonaka and Takeuchi, 1995; Yli-Luoma and Naeve, 2006), thus enhancing the Knowledge-creating company towards the vision of the Semantic Learning Organisation (SLO) (Sicilia and Lytras, 2005). In essence, active user participation which is enabled through user training is considered as a knowledge-creation spiral that emerges when the interaction between tacit and explicit knowledge is elevated dynamically from lower to higher ontological levels (Nonaka and Takeuchi, 1995). Figure 2 shows, graphically, how an existing individual’s tacit knowledge (as-is) is converted by healthcare process analysts into explicit knowledge. In turn, this knowledge is enriched in order to be converted into group’s knowledge through group training, thus enabling user groups to participate in the development of new systems (to-be) in co-operation with healthcare analysts. New healthcare processes designed represent organisational knowledge in intra-organisational process activities, and inter-organisational knowledge in inter-organisational process activities. Figure 2
Developing organisational knowledge through users’ participation
2.1 Ontology design Ontologies are collections of concepts (universals), instances of concepts (particulars) and relations among them (Fielding et al., 2004). Attributes are assigned to concepts, instances and relations in order to specify the content of the knowledge network. In addition, ontology constructs (e.g., concepts, relations and instances) could be enriched with terms, definitions, axioms and constraints that are expressed at the desired level of formality and that are deemed to be important in characterising the knowledge domain under consideration at the desired level of detail (Grenon, 2003; Sowa, 2000; Gruber, 1993). These are used in asking and answering questions, making assertions, offering insights, describing practices and discussing investigations. The development of an ontology is usually a top-down process which starts at the highest level of abstraction considered and finishes at the lowest level of abstraction which is considered appropriate for the purpose of the ontology building process
10
A. Macris et al.
(Colomb and Dampney, 2005; Cristani and Cuel, 2005; Masuwa-Morgan and Burrell, 2004). As an example, in an ontology development process there could be considered three levels of resolution: the upper-ontology (which includes the basic concepts and relations), the mid-ontology (which includes more detailed concepts and relations) and the lower-ontology (which includes all the concepts, instances and relations necessary for the specific purpose of the ontology).
2.2 Knowledge networks: using Semantic Web principles in training Most of the existing automated training aids are essentially collections of multimedia objects (content). These multimedia objects are usually grouped hierarchically (e.g., in units and sub-units), indexed and combined, through hyperlinks, in order to support various training needs. However, these training aids only provide for manipulating and restructuring multimedia objects in order to create training material, serving specific needs, for the knowledge domain under consideration. Hence, this knowledge must be externalised and made explicit by the user in order to become diffused and reusable. The approach proposed in this paper enables experts in the domain of healthcare delivery to externalise the domain knowledge in the form of ontology-based knowledge networks (training scenarios serving specific training needs) and, hence, better communicate it and make it reusable. The basic structure of the proposed approach is a domain specific ontology which captures the relevant knowledge. Thus, training scenarios combine ontology constructs with supportive multimedia objects helping trainees acquire an in depth understanding of the knowledge domain. The approach is based on Sowa’s definition of user perception as the process of building a working model that represents and interprets sensory inputs (mosaic of percepts) into a more abstract part (conceptual graph) (Novak and Gowin, 1984; Sowa, 1984). Hence, understanding of a training material by a trainee can be modelled as a two stage process: •
the analysis sub-process, where the material is broken down into concepts
•
the synthesis sub-process where concepts are linked to other concepts which are found either in the training material at hand or in other related material that the trainee has already analysed before in order to form more complex structures (conceptual graphs).
Thus, meaning is not discovered but constructed, and training material has meaning only in relation to other material, being interconnected to each other as codes and systems in the minds of the trainees according to their cultural and social traits. In designing an ontology-based training aid, the main objective is to capture and represent the knowledge which is implicit in the application domain so that it can be made reusable. Thus, domain experts record their knowledge on the particular field under consideration in terms of an ontology which is recorded in the ontology repository. Hence, each ontology construct is recorded only once and can be made available to every training scenario using it. In addition, relevant supportive material (either existing or created), in the form of multimedia objects (e.g., text, image, video and animation), is used in order to develop a collection of reusable multimedia objects that are related to the knowledge domain under consideration (Chebotko et al., 2005; Steinmetz and Seeberg, 2003). This collection of multimedia objects comprises the content repository. The ontology and content repositories are then used to create knowledge networks, each
Using ontology-based knowledge networks for user training
11
corresponding to a training scenario, which are recorded in the knowledge repository. Figure 3 shows a schematic representation of the three repositories used in the proposed approach. Figure 3
Knowledge networks: linking ontology constructs to multimedia objects
Contrary to traditionally designed training scenarios which are based on mere user navigation to multimedia objects, training scenarios that are based on the proposed approach are enhanced and empowered in that they allow users to navigate into the domain knowledge which has been represented in the form of a knowledge network. Thus, the user of the training scenarios is guided either through a semantic search followed by a navigation to the knowledge network, or directly through navigation to the knowledge network. To enhance his or her understanding of each ontology construct included in a knowledge network, the user can access relevant supportive material in the form of multimedia objects and identify the relation of the particular construct with other relevant constructs. Most research work presented in the literature deals with the manipulation of multimedia objects or unstructured content, using ontologies, mainly for clinical data retrieval (Eichelberg et al., 2005; Bicer et al., 2005; Chu and Cesnik, 2001) and not for representing user training material in the form of knowledge networks that link ontology constructs with supportive multimedia objects. However, there are some similarities with •
Geueke and Stausberg who have developed a solution called Learning Resource Server Medicine (LRSMed), which provides a service for the learners to retrieve WWW resources for medical education at the time of learning, based on meta-data collection of third-party electronic learning modules (Geueke and Stausberg, 2003)
•
Chu and Cesnik approach who use conceptual graphs in order to structure free-text medical documents for information retrieval and knowledge discovery (Chu and Cesnik, 2001).
3
A case study in a healthcare environment
In recent years, healthcare organisations have been facing the challenges of the rapidly changing environment with healthcare processes displaying more complexity due to interactions between their internal components, and interactions of healthcare processes with the external environment. Moreover, the need for managing healthcare processes and improving the quality of care provided has created a shift from function-centred to
12
A. Macris et al.
patient-centred services, requiring increased cooperation and collaboration among functional units. Hence, considerable attention has been paid on designing new healthcare processes, or redesigning existing ones, and on developing process-oriented healthcare information systems to support them (Lenz and Kuhn, 2004; Andersson et al., 2003). Moreover, protocols for patient care are explicit organisational models, since they describe the steps (work processes) necessary to care for patients with particular illnesses (Field and Lohr, 1990); they are used in order to reduce practice variability and to improve the quality of patient care while reducing the costs incurred (Loback and Hammond, 1994). Although there has been significant effort at both the national and the institutional levels to create standard protocols, more effort is required to make them adaptable to a variety of clinical settings and be accepted by healthcare professionals. Hence, many institutions, recognising the importance of creating medical process descriptions that are specific to their organisation, have appointed a medical director or committees of health care providers in order to transform the generic protocols into site-specific protocols, more acceptable to the practitioners in the institution and more capable of effecting change in medical practice (Gensensway, 1995). However, protocols of care are dynamically changing, as healthcare organisations undergo a constant flux of employees and restructuring of responsibilities. Thus, each time a change occurs, the relevant protocols must also be updated through a collaborative work between healthcare professionals and process experts. For healthcare process modelling to be successful it is necessary to get hold of the necessary requirements that are, mainly, clustered around understanding of the problem domain, management of complexity and clear and concise communication. Such requirements are then used to design a healthcare process model that offers clear and concise views, consistent with each other, that demonstrably fulfil the requirements placed on process capture. Hence, on the common assumption that requirements elicitation and verification for healthcare process modelling is best performed through active user participation, empirical evidence has shown that there is a need for providing users with suitable and, in some cases, adaptable training material so as to enable them instil their knowledge and expertise in the healthcare process modelling activity.
3.1 Motivating scenario The basic motivation for this research stems from our involvement in a project concerned with the design and incremental implementation of a process-oriented healthcare system architecture spanning a health district. The district organisations had implemented a number of heterogeneous, function-oriented, autonomous transaction processing systems that had been adapted to organisational or individually accepted practices and, hence, they had to be retained in order to preserve existing investments on information processing resources. The need to involve users in the healthcare process definition (both intra- and inter-organisational) and automation activities was the key factor for developing a suitable user training aid that allows incorporating the underlying knowledge in these healthcare processes. For illustrative purposes, a simple healthcare process concerned with certain patient flow tasks, as well as with the functional units and healthcare professionals involved in each task, was considered. Thus, each patient is assumed to undergo a number of medical activities such as clinical examination, laboratory tests and diagnosis. In addition, laboratory tests involve requesting a test, scheduling a test, preparing the patient,
Using ontology-based knowledge networks for user training
13
collecting and delivering the specimen, performing the test and issuing a test result report. In order to enable users to understand the healthcare process considered and participate actively in the healthcare process modelling activity, a training scenario (knowledge network) incorporating the underlying logic was created using the tools developed by the CULTOS (Cultural Units of Learning – Tools and Services; http://www.cultos.org) project. These tools are: •
K-infinity (http://www.i-views.de) tool (for creating and populating the ontology repository)
•
CULTOS media import tool (for creating and populating the content repository)
•
CULTOS authoring tool (for creating and populating the knowledge repository).
The training scenario (knowledge network) created for healthcare process modelling is saved as a structured multimedia meta-object containing expert knowledge which is called Enhanced Multimedia Meta-Object (EMMO). Thus, each scenario is a self-contained entity that includes relevant entries of the three repositories defined above.
3.2 The healthcare process training ontology For the purpose of this research, which is mainly to show the advantages of the knowledge network approach, the ontology model proposed by Sowa was used, due to its simplicity (Sowa, 2000). However, the dynamic spatial ontology model SNAP and SPAN, proposed by the Institute for Formal Ontology and Medical Information Science (IFOMIS, http://www.ifomis.uni-saarland.de) can also be considered as a more elaborate model. Figure 4 shows the ontology concepts of the healthcare process considered, in the form of a generalisation-specialisation hierarchy, linked with relations. The upper ontology consists of the six categorisations of concepts proposed by Sowa (2000): Physical (concerning matter or energy) and Abstract (concerning pure information structures) which are further broken down into Physical and Abstract Continuants are further broken down into Physical and Abstract Continuants (having stable attributes that enable their various appearances at different times to be recognised as the same object) and Physical and Abstract Occurrents (processes or events that are in a state of flux and that can only be identified by their locations in some region of time-space). These concepts are then further specialised into lower level sub-concepts found in the literature (Kumar et al., 2006). For example, in Figure 4, the Physical Continuant concept is specialised into the concepts healthcare professional and medical resource while the medical resource concept is specialised into the concepts premise and medical equipment and the healthcare professional concept is specialised into the concepts medical staff, nursing staff and laboratory staff. Also, the Physical Occurrent concept is specialised into the concepts patient and patient tasks. Moreover, relations are defined between ontology concepts of upper and lower level of resolution. Any relation defined between two concepts also holds for lower level concepts and for concept instances. For example, the relation R between concepts X1 and Y1 holds also for concepts X2 and Y2 that are sub-concepts of X1 and Y1, respectively. Also, the relation R between concepts X1 and Y1 also holds for x11 and y11 that are concept instances of concepts X1 and Y1, respectively. Table 1 shows the relations defined for the needs of the training scenario.
14
A. Macris et al.
Figure 4
A training ontology for the healthcare process
Table 1
Ontology-defined relations
Source concept Target concept
Relation
Abstract occurrent Abstract occurrent Abstract occurrent
Abstract occurrent Patient
‘Is followed by’ ‘Follows’
Healthcare professional
Abstract occurrent
Healthcare professional
Hospital
Hospital
Inverse relation
Use To define flows of tasks
‘For’
‘Is the recipient of’ To show the recipient of tasks ‘Takes place at’ ‘Provides the To show departments infrastructure for’ where medical tasks take place ‘Is involved in’ ‘Involves’ To show which professionals are involved in each task ‘Belongs to’ ‘Occupies’ To show professionals’ occupation in departments
3.3 A training scenario The training scenario (knowledge network) was designed on the business process considered, where concepts, and instances of concepts, are represented as rounded rectangled nodes and relations are represented as oval edges. In what follows, a description of the above business process is provided using ontology concepts (shown in italics), ontology global and local relations (shown in single quote enclosures)
Using ontology-based knowledge networks for user training
15
and local instances associated with multimedia objects (shown in plain text words separated with underscores). Figure 5 shows the training scenario on the above healthcare process. Figure 5
A training scenario for the healthcare process
Table 2 shows how medical agents relate to various tasks in the flow. For example Clinical Examination ‘Takes place at’ a Medical Department and Laboratory Test ‘takes place at’ a Laboratory, in order to show the organisational structure that provides the necessary infrastructure for each examination. Table 3 shows some of the instances of concepts defined in the training scenario. For example, the instance Lab_Exam_Request_Form is associated with multimedia objects that may include sample forms and guidelines for the completion of each form. Also, the instance Receive_Specimen ‘describes procedure’ Receive specimen (i.e., includes instructions in various forms, such as text, pictures, videos etc) in order to assist Nursing Staff to Receive Specimen. In practice, using the above approach for designing training material would involve both a high level view of a healthcare process model, such as the one described above and a low level view of the model that takes into account the characteristics of a particular healthcare unit. In addition, multimedia objects will be associated with concepts, concept instances and relations in order to demonstrate both the static and the dynamic features of the healthcare process under investigation. Thus, each training scenario designed can be easily modified to represent another version of the healthcare process model considered so that to enable users to assess the pros and cons of a process redesign exercise. Redesign of a healthcare process model can be performed by simply manipulating already defined objects; hence, providing flexibility, agility and reusability of the training material designed.
16
A. Macris et al.
Table 2
Locally-named relations (defined in the scenario)
Source concept
Target concept
Ontology relation
Local name
Medical staff
Clinical examination
‘Is involved in’
‘Performs’
Medical staff
Lab test request
‘Is involved in’
‘Issues’
Medical staff
Lab test report
‘Is involved in’
‘Evaluates’
Medical staff
Diagnosis
‘Is involved in’
‘Issues’
Nursing staff
Preparation for examination
‘Is involved in’
‘Prepares’
Nursing staff
Receive specimen
‘Is involved in’
‘Receives’
Nursing staff
Send specimen for test
‘Is involved in’
‘Sends
Laboratory staff
Lab test request
‘Is involved in’
‘Receives’
Laboratory staff
Laboratory test
‘Is involved in’
‘Performs’
Laboratory staff
Lab test report
‘Is involved in’
‘Issues’
Table 3
Instances of concepts
Instance
Leading concept
Ontology relation
Lab_Test_Request_Form
Lab Test Request
‘Is instance of’
‘Gives details about’
Receive_Specimen
Receive specimen
‘Is instance of’
‘Describes procedure’
Lab_Test_Report
Lab Test Report
‘Is instance of’
‘Gives details about’
4
Local name
Discussion
The approach proposed in this paper is mainly concerned with capturing and representing the knowledge found in the logic, the structure and the ways of use of healthcare processes as an ontology-based knowledge network (a training scenario serving a specific training need). The ontology contains all the relative concepts and instances of concepts and the relations between them. The knowledge network relates the basic entities defined in the ontology with the various multimedia (text, image, video, animation etc.), which are supportive for better understanding the ontology constructs. Thus, the user of the resulting training material is enabled to search for an ontology construct (for example a healthcare task) and understand its meaning and usage with the help of the supportive multimedia. Furthermore, the user can navigate to associated ontology constructs (for example, other tasks, healthcare units and staff) in order to acquire in depth knowledge about the healthcare process tasks, the data and control flows between process tasks, organisational structures and healthcare professionals and the needs for designing new or redesigning old healthcare processes. The proposed model does not disregard existing methodologies for structuring training material, but enhances and empowers them by allowing the semantic representation of knowledge so as to enable users to navigate into a knowledge network based on the characteristics of the application domain under consideration. Thus, the model can combine the existing multimedia material with ontology constructs, using knowledge-based multimedia authoring tools, in order to build user training scenarios and satisfy specific training needs. Hence, in addition to the existing
Using ontology-based knowledge networks for user training
17
multimedia objects, the knowledge built into both the ontology and the training scenarios is fully reusable. With regard to the creator of the training material, the main advantages of the proposed model can be grouped around its three main components (repositories): •
Ontology repository. There are many benefits when using ontologies, in order to make explicit the underlying business process logic in an application domain so that it can be made reusable, that have already been recognised in the learning technology community (Sampson et al., 2004; Lytras et al., 2003).
•
Content repository. Content reusability of the educational material created, that is achieved with the proposed model and is related to the knowledge domain under consideration, is a key issue in the literature (Chebotko et al, 2005; Steinmetz and Seeberg, 2003).
•
Knowledge repository. A knowledge network is a self contained entity that serves a specific training need in a specific knowledge domain, in a specific business context and has a specific structure (Stojanovic et al., 2001). Reusability of knowledge recorded into training scenarios is also achieved as knowledge constructs instilled into older scenarios and can be used in new scenarios in order to meet new training needs.
With regard to the trainee, the main advantages of the proposed model are the following: •
•
•
Semantic search. This allows to search ontology or knowledge constructs semantically instead of textually (i.e., the search is based on language-agnostic semantic matching instead of keyword matching, that involves: •
the knowledge domain under consideration
•
the business context
•
the structure of the educational scenario putting emphasis on matching the content and the real meaning of each relevant concept searched (Stojanovic et al., 2001).
Knowledge or conceptual navigation. This allows the use of browsing and navigation capabilities in order to identify the ontology or knowledge constructs, as they are recorded, into the knowledge repository and used in the training scenarios and involves: •
navigation into the knowledge domain under consideration
•
navigation into the business context
•
navigation into the structure of the educational scenario (Stojanovic et al., 2001).
Knowledge dissemination. This is an important function of any kind of training activity that can only be achieved if the trainee is provided with the ability to extract the knowledge implicit in the problem domain, as opposed to the mere presentation of facts and disconnected information which, in most cases, is not adequate. With the proposed model, knowledge is made explicit in order to assist the trainees’ combination (from explicit to explicit) and internalisation (from explicit to tacit) knowledge transformation processes (Nonaka and Takeuchi, 1995).
18
A. Macris et al.
With regard to healthcare process modelling, there is a need for in-depth understanding of each process, including the process task structures and dependencies, to enable a change in the users’ perception of their work (from function-oriented to process-oriented). Considering a healthcare process from another viewpoint (that of the healthcare analyst) besides your own (user) leads to deeper understanding of the healthcare processes under consideration and this understanding is essential when trying to develop critically thinking users. And the ability to think critically is a necessary criterion in order to say that knowledge has been disseminated. Obviously, this ability is much more important when studying complex healthcare processes that consist of many interrelated tasks that combine various resources available. The proposed approach to user training may have a significant impact on enabling users to participate actively in healthcare process management activities since they are equipped with an appropriate tool for acquiring a clear and an in depth understanding of a healthcare process. Based on this understanding, increased cooperation and collaboration between users and analysts can be achieved towards a common objective of improving healthcare process performance in an ever changing environment, since process changes that are deemed appropriate can be incorporated easily by simply manipulating objects already defined in the ontology and/or in the knowledge network. Besides, the training tool provides users with healthcare process objects, such as mock-ups and animations, to enable them to get a grip of the situation under study. Hence, user acceptance of designed, or redesigned, healthcare processes can be considerably improved due to an increased likelihood of being perceived useful and usable by users. Due to the encouraging results of the approach described, it is intended to evaluate it extensively, using more elaborate implementation tools and more complex healthcare processes in real-world situations.
5
Concluding remarks
This paper presents an ontology-based prototype approach for the construction of user training scenarios on healthcare process modelling concepts, whereby both the multimedia objects used and the knowledge built into the training scenarios are fully reusable. The approach consists of the following steps: •
define and implement a general ontology
•
design the training scenario that best fits training objectives and define the multimedia objects required
•
refine the general ontology by adding all ontology constructs required by the various training scenarios
•
develop or select the multimedia objects which are deemed necessary to support the various ontology constructs to be used in the training scenarios
•
construct each training scenario (knowledge network) by combining ontology constructs with multimedia objects.
The objective of the approach presented in this paper is to enable users to get familiar with and, hence, participate actively, in healthcare process modelling activities. In this context, a sample training scenario in the healthcare domain was built, using the
Using ontology-based knowledge networks for user training
19
CULTOS tools, that is based on a self-contained reusable knowledge repository which combines the ontology constructs (stored in the ontology repository) with supportive multimedia objects (stored in the content repository). Hence, this training scenario can also be used by third party hypermedia production tools in order to present both the knowledge and the content of the training material for various purposes.
References Anderson, E.E. (1985) ‘Managerial considerations in participative design of MIS/DSS’, Information and Management, Vol. 9, No. 4, pp.201–207. Andersson, A., Hallberg, N. and Timpka, T. (2003) ‘A model for interpreting work and information management in process-oriented healthcare organisations’, International Journal of Medical Informatics, Vol. 72, pp.47–56. Barrow, P.D.M. and Mayhew, P.J. (2000) ‘Investigating principles of stakeholder evaluation in a modern IS development approach’, Journal of Systems and Software, Vol. 52, Nos. 2–3, pp.95–103. Berg, M. and Toussaint, P. (2003) ‘The mantra of modelling and the forgotten powers of paper: a sociotechnical view on the development of process-oriented ICT in health care’, International Journal of Medical Informatics, Vol. 69, pp.223–234. Berners-Lee, T., Hendler, J. and Lassila, O. (2001) ‘The semantic web’, Scientific American, May, pp.28–37. Bicer, V., Kilic, O., Dogac, A. and Laleci, G.B. (2005) ‘Archetype-based semantic interoperability of web service messages in the health care domain’, International Journal of Semantic Web and Information Systems, Vol. 1, No. 4, pp.1–23. Cartwright, J., de Sylva, S., Glasgow, M., Rivard, R. and Whiting, J. (2002) ‘Inaccessible information is useless information: addressing the knowledge gap’, Journal of Medical Practice Management, Vol. 18, No. 1, pp.36–41. Cavaye, A.L.M. (1995) ‘User participation in system development revisited’, Information and Management, Vol. 28, No. 5, pp.311–323. Chebotko, A., Deng, Y., Lu, S., Fotouhi, F. and Aristar, A. (2005) ‘An ontology-based multimedia annotator for the semantic web of language engineering’, International Journal of Semantic Web and Information Systems, Vol. 1, No. 1, pp.50–67. Chu, S. and Cesnik, B. (2001) ‘Knowledge representation and retrieval using conceptual graphs and free text document self-organisation techniques’, International Journal of Medical Informatics, Vol. 62, pp.121–133. Colomb, R.M. and Dampney, C.N.G. (2005) ‘An approach to ontology for institutional facts in the semantic web’, Information and Software Technology, Vol. 7, No. 12, pp.775–783. Cristani, M. and Cuel, R. (2005) ‘A survey on ontology creation methodologies’, International Journal of Semantic Web and Information Systems (IJSWIS), Vol. 1, No. 2, pp.49–69. Currie, W.L. and Willcocks, L. (1996) ‘The new branch Columbus project at royal bank of Scotland: the implementation of large-scale business process re-engineering’, The Journal of Strategic Information Systems, Vol. 5, No. 3, pp.213–236. Dustdar, S., Hoffmann, T. and van der Aalst, W. (2005) ‚Mining of ad-hoc business processes with TeamLog’, Data and Knowledge Engineering, Vol. 55, No. 2, pp.129–158. Dvir, D. (2005) ‘Transferring projects to their final users: the effect of planning and preparations for commissioning on project success’, International Journal of Project Management, Vol. 23, No. 4, pp.257–265. Eichelberg, M., Aden, T. and Thoben, W. (2005) ‘A distributed patient identification protocol based on control numbers with semantic annotation’, International Journal of Semantic Web and Information Systems, Vol. 1, No. 4, pp.24–43.
20
A. Macris et al.
Field, M.J. and Lohr, K.N. (Eds.) (1990) Clinical Practice Guidelines: Directions for a New Program, IOM report, National Academy Press, Washington DC. Fielding, J., Simon, J., Ceusters, W. and Smith, B. (2004) ‘Ontological theory for ontological engineering: biomedical systems information integration’, Proceedings of the Ninth International Conference on the Principles of Knowledge Representation and Reasoning (KR2004), Whistler, BC, 2–5 June, pp.114–120. Gensensway, D. (1995) ‘Putting guidelines to work-lessons from the real world’, ACP Observer, Vol. 15, No. 1. Geueke, M. and Stausberg, J. (2003) ‘A meta-data-based learning resource server for medicine’, Computer Methods and Programs in Biomedicine, Vol. 72, pp.197–208. Grenon, P. (Ed.) (2003) Reference Ontologies and Application Ontologies, Institute for Formal Ontology and Medical Information Science (IFOMIS) Reports, ISSN 1611-4019. URL http://www.ifomis.uni-saarland.de/Downloads/Reports/IR-0304_Grenon.pdf Gruber, T.R. (1993) ‘A translation approach to portable ontology specifications’, Knowledge Acquisition, Vol. 5, No. 2, pp.199–220. Hartman, S.J. and Crow, S.M. (2002) ‘Executive development in healthcare during times of turbulence: top management perceptions and recommendations’, Journal of Management in Medicine, Vol. 16, No. 5, pp.359–370. Haux, R., Seggewies, C., Baldauf-Sobez, W., Kullmann, P., Reichert, H., Luedecke, L. and Seibold, H. (2003) ‘Soarian – workflow management applied for health care’, Methods of Information in Medicine, Vol. 42, No. 1, pp.25–36. Kumar, A., Smith, B., Pisanelli, D., Gangemi, M. and Stefanelli, M. (2006) ‘Clinical guidelines as plans: an ontological theory’, Methods of Information in Medicine, Vol. 45, No. 2, pp.204–210. Lee, E.K. (1986) ‘An exploratory contingency model of user participation and MIS use’, Information and Management, Vol. 11, No. 2, pp.87–97. Lenz, R. and Kuhn, K.A. (2004) ‘Towards a continuous evolution and adaptation of information systems in healthcare’, International Journal of Medical Informatics, Vol. 73, No. 1, pp.75–89. Loback, D.F. and Hammond, W.E. (1994) ‘Development and evaluation of a computer-assisted management protocol’, Proceedings of the Symposium on Computer Applications in Medical Care, American Medical Informatics Association, Washington DC, pp.787–791. Lytras, M.D., Tsilira, A. and Themistocleous, M. (2003) ‘Towards the semantic e-learning: an ontological orientated discussion of the new research agenda in e-learning’, Proceedings of the 9th Americas Conference on Information Systems, Tampa, FL, USA, pp.2985–2997. Malamateniou, F., Poulymenopoulou, M. and Vassilacopoulos, G. (2003) ‘Developing a process-oriented virtual healthcare enterprise using web services’, Proceedings of the ICICTH, Samos Island, Greece, pp.230–234. Masuwa-Morgan, K.R. and Burrell, P. (2004) ‘Justification of the need for an ontology for accessibility requirements (Theoretic framework)’, Interacting with Computers, Vol. 16, No. 3, pp.523–555. McAdam, R., Keogh, W., Galbraith, B. and Laurie, D. (2005) ‘Defining and improving technology transfer business and management processes in university innovation centres’, Technovation, Vol. 25, No. 12, pp.1418–1429. Naeve, A., Yli-Luoma, P., Kravcik, M., Lytras, M.D., Lindegren, M., Nilsson, M., Korfiatis, N., Wild, F., Wessblad, R., Kamtsiou, V., Pappa, D. and Kieslinger, B. (2005) ‘A conceptual modelling approach to studying the learning process with a special focus on knowledge creation’, Deliverable 5.3 of the Prolearn EU/FP6 Network of Excellence, IST 507310, June. Nonaka, I. and Takeuchi, H. (1995) The Knowledge-Creating Company, Oxford University Press, New York. Novak, J.D. and Gowin, D.B. (1984) Learning How to Learn, Cambridge University Press, UK, NewYork and Cambridge, pp.15–25.
Using ontology-based knowledge networks for user training
21
Poulymenopoulou, M., Malamateniou, F. and Vassilacopoulos, G. (2003) ‘Emergency healthcare automation using workflow technology and web services’, Medical Informatics and the Internet in Medicine, Vol. 28, No. 3, pp.195–207. Reijers, H.A. and Mansar, S.L. (2005) ‘Best practices in business process redesign: an overview and qualitative evaluation of successful redesign heuristics’, Omega, Vol. 33, No. 4, pp.283–306. Rinderle, S., Weber, B., Reichert, M. and Wild, W. (2005) ‘Integrating process learning and process evolution – a semantics based approach’, Lecture Notes in Computer Science, Springer-Verlag, Vol. 3649, p.252. Rinkus, S., Walji, M., Johnson-Throop, K.A., Malin, J.T., Turley, J.P., Smith, J.W. and Zhang, J. (2005) ‘Human-centered design of a distributed knowledge management system’, Journal of Biomedical Informatics, Vol. 38, No. 1, pp.4–17. Rutner, S.M., Gibson, B.J. and Williams, S.R. (2003) ‘The impacts of the integrated logistics systems on electronic commerce and enterprise resource planning systems’, Transportation Research Part E: Logistics and Transportation Review, Vol. 39, No. 2, pp.83–93. Sampson, D.G., Lytras, M.D., Wagner, G. and Diaz, P. (2004) ‘Guest editorial: ontologies and the semantic web for e-learning’, Educational Technology and Society, Vol. 7, No. 4, pp.26–28. Sicilia, M-A. and Lytras, M.D. (2005) ‘The semantic learning organization’, The Learning Organization, Vol. 12, No. 5, pp.402–410. Sowa, J. (1984) Conceptual Structures, Information Processing in Mind and Machine, Addison-Wesley Publishing Company Inc., New York. Sowa, J. (2000) Knowledge Representation – Logical, Philosophical and Computational Foundations, Brooks/Cole, USA. Steinmetz, R. and Seeberg, C. (2003) ‘Meta-information for multimedia e-learning’, Lecture Notes in Computer Science, Springer-Verlag, Vol. 2598, pp.293–303. Stojanovic, L., Saab, S. and Studer, R. (2001) ‘Knowledge technologies for the semantic web’, Proceedings of the World Conference on the WWW and the Internet (WebNet 2001), Orlando, FL, USA, pp.1174–1183. Sycara, K. (2004) ‘Dynamic discovery, invocation and composition of semantic web services’, in Vouros, G.A. and Panayiotopoulos, T. (Eds.): Methods and Applications of Artificial Intelligence, Springer-Verlag Berlin Heidelberg, SETN, LNAI 3025, pp.3–12. Vassilacopoulos, G. and Paraskevopoulou, E. (1997) ‘A process model basis for evolving hospital information systems’, Journal of Medical Systems, Vol. 21, pp.141–153. Vassilacopoulos, G., Chrissikopoulos, V., Alexandris, N. and Tsouroplis, A. (1996) ‘Partcipative HIS development: an approach and a case study’, Journal of Medical Systems, Vol. 20, No. 3, pp.157–165. Wald, J.S., Pedraza, L.A., Reilly, C.A., Murphy, M.E. and Kuperman, G.J. (2001) ‘Requirements development for a patient computing system’, Proceedings of the AMIA Annual Symposium, Washington, DC, pp.731–735. Wieringa, R.J., Blanken, H.M., Fokkinga, M.M. and Grefen, P.W.P.J. (2003) ‘Aligning application architecture to the business context’, Lecture Notes in Computer Science, Springer-Verlag, Vol. 2681, pp.209–225. Yli-Luoma, P.V.J. and Naeve, A. (2006) ‘Towards a semantic e-learning theory by using a modelling approach’, British Journal of Educational Technology, Vol. 37, No. 3, pp.445–459.
