Analysis of the International EMBOK Model as a Classification System

May 23, 2007

Paper presented at the 2007 Las Vegas International Hospitality & Convention Summit

© 2007. Julia Rutherford Silvers, CSEP. All Rights Reserved

ABSTRACT

Hundreds of thousands of people are involved in creating and staging events of all types and sizes throughout the world and hundreds of millions of people are attending these events. By any measure this is a huge industry. However, event management has yet to achieve status as a true profession because it has not taken the steps necessary for legitimacy, one of which is the establishment of an acknowledged industry-wide body of knowledge that has captured the sum of knowledge pertinent to this emerging profession. By any measure this is a huge task. Establishing a classification system will be fundamental in managing this overwhelming amount of information, experience and heuristics — breaking it down into manageable and meaningful parts. This paper will explore the nature of classification systems and the strengths, weaknesses, challenges and opportunities of the International EMBOK Model as a classification system for events and event management.

INTRODUCTION

Event management is an emerging profession, one that is growing in recognition and scrutiny in terms of occupational popularity and occupational responsibilities. It will not reach the status of an actual profession until standards are developed based upon a practitioner-codified body of knowledge and these standards are included, imparted, and improved upon in the academic setting and curricula. Event management is also a broad spectrum profession, encompassing a wide range of event genre from meetings and conventions to trade shows and festivals to sports events and special events. It will not reach its full potential as an economic and political force until there is industry-wide recognition of the commonality of skill sets. The movement of the profession into legitimacy and economic validity may be facilitated by the development of a structure for collecting, analyzing, and retrieving the knowledge that will guide development of standards criteria, curricula, and competency assessments, and that will acknowledge the holistic nature of the industry.

This paper begins with an overview of the International EMBOK (Event Management Body of Knowledge) Model, a conceptual framework including processes, phases, core values, and function or knowledge domains that may be customized to meet the needs of various users including governments, academe, and industry practitioners. The nature and features of classification systems will be examined and the International EMBOK Model will be assessed as a classification system for knowledge management in the events industry in its entirety, recognizing the potential for foreign and mutual credential recognition that will facilitate job creation and labor mobility.  Finally, the missing link — an event classification system — will be explored to suggest the attributes necessary to provide design guidance based on magnitude metrics that reflect the “scale, nature, spatial and temporal aspects of an event” (BSI, 2006, p. 3).

OVERVIEW OF THE INTERNATIONAL EMBOK MODEL

The International EMBOK Model was instigated by and included the classification work of Silvers (2004) to create a framework for the recognition and collection of the body of knowledge pertinent to the events industry, which was based on an affinity analysis of the knowledge systems in the event management industry including numerous occupational standards, certification competency domains, educational curriculums, and industry literature from Australia, Canada, South Africa, the United Kingdom, and the United States. The initial framework and taxonomy developed by Silvers was expanded upon at the 2004 International EMBOK Imbizo, where academic and practitioner experts formed the International EMBOK Executive and expanded the previous work done by each to develop the conceptual framework for a common international event management body of knowledge that may be customized to meet the needs of various cultures, governments, education programs, and organizations (Silvers, 2005; Silvers, Bowdin, O’Toole, & Nelson, 2006).

The goal of the International EMBOK Model is to serve as a resource for current and future industry practitioners, curriculum developers, and competency credentialing organizations to ensure the profession of events management achieves credibility through continuous improvement. As illustrated in Table 1, the framework presents five Processes, five Phases, five Core Values, and five Knowledge Domains that encompass thirty-five functional areas, referred to as Classes. This framework of categories facilitates study, analysis, and knowledge management, as well as providing a structure for systematically and consistently approaching the scope of responsibilities and competencies necessary to bring legitimacy and gravitas to this profession.

Table 1. Scope of the International EMBOK Model

THE NATURE OF CLASSIFICATION SYSTEMS

Classification, simply put, is the method by which things are identified and organized in categories so that they may be conveniently located and their nature understood. Classification serves to provide a shorthand description of an object within a category and to group it with similar objects or concepts in meaningful clusters (Kwasnik, 1999). The various structures and systems devised to organize this information range from life and library sciences to knowledge management, each system designed with the user in mind — division and clustering based on user-specified criteria (Bruno & Richmond, 2003), but most incorporate one or more of the hierarchical, tree, and faceted analysis structural approaches for aggregation and differentiation.

The hierarchical approach presents a super class (primary defining character) and sub-classes (differentia) that shows the “is/a” relationship and presumes an inheritance from the super class throughout the sub-classes (Kwasnik, 1999). For example, in life science all life forms are classified according to seven components: Kingdom, Phylum, Class, Order, Family, Genus and Species, and a single Species has all the characteristics of the Genus, which shares all the characteristics of the Family, and so on. The Dewey Decimal System is another example of the hierarchical approach.

The tree approach, although hierarchical in character, illustrates a “whole/part” relationship without the inheritance assumption (Kwasnik, 1999; Bruno & Richmond, 2003). The United Nations Standard Products and Services Code, for example, uses a five-level whole/part hierarchical convention of Segment, Family, Class, Commodity and Business Function to classify all products and services (UNSPSC, 2001). This type of hierarchy is often seen in organizational charts wherein personnel are assigned to a supervisor within a department within a division representing a specified chain of command and control.

The faceted analysis approach establishes categories based on fundamental descriptive criteria (variable attributes or “type/instance” relationships) and may be hierarchical in terms of citation order, but may also be flexible according to the descriptors of interest at the time of retrieval (Kwasnik, 1999; Factiva, n. d.; KMconnection, n. d.). For example, the U.S. Dictionary of Occupational Titles (DOL, 1991) uses a nine-digit code to classify industry and worker function characteristics, as illustrated in Figure 1. Faceted analysis systems offer flexibility in that new facets may be added at any time without disturbing the structure, can accommodate concepts that can belong to multiple hierarchies, and may be searched using a single facet or multiple facets without regard to parent categories. For example, one might search the Dictionary of Occupational Titles for all service occupations (assigned the number 3 in the 1st-digit Broad Category within the Industry) that require negotiating skills (assigned the number 1 in the 5th-digit People within the Worker Functions).

Figure 1. Overview of the U.S. Dictionary of Occupational Titles code number

Another example, as illustrated in Figure 2, is the University of New Mexico Fine Arts Slide Library’s three line code used to classify its inventory of slides used in art history classes for its filing and retrieval system, which allows users to search for items by medium, nationality, artist, or the stage within an artist’s body of work depending on the user’s need. It also allows acquisitions to be integrated into the slide collection with a minimum of expertise in the field, needing only identification of the artist, date of the work, and medium in order to properly position it within the classification system. This simple alphanumeric code becomes a shorthand descriptor for users.

Figure 2. Classification code for UNM Fine Arts Slide Library

Each classification approach relies on the elimination of, as well as seeks to eliminate, ambiguity through use of a controlled vocabulary wherein naming conventions provide explicit and unambiguous descriptors. Given that language is variable and terminology is often industry or user specific, a controlled vocabulary with standardized terms must be established for a classification system (Factiva, n. d.). This must be supplemented with a thesaurus that incorporates satellite vocabularies, equivalency variants (spelling, regional and multi-lingual equivalents, organization-specific preferred terminology, acronyms and abbreviations), synonyms or aliases for the one preferred or “hub” term, and disambiguation modifiers for homographs where one word refers to several different things, e.g. using parenthetical modifiers — Mercury (planet) and Mercury (substance) (Factiva, n. d.). This becomes increasingly important in an electronic knowledge management system to ensure navigation tools (e.g. search fields and breadcrumbs) are distinctive and apparent, and facilitates development of a structure capable of reflecting inference by context placement (Bruno & Richmond, 2003).

ASSESSMENT OF THE EMBOK AS A CLASSIFICATION SYSTEM

The International EMBOK Model offers a structural arrangement that allows the division of knowledge and heuristics into recognizable groupings using both the hierarchical tree and faceted analysis approaches. Its hierarchical tree classification system characteristics are most evident in the knowledge domains, showing the relationships that link them together and the whole/part affiliations typically found in management structures of event organizations (e.g. departments and/or committees in an organizational chart). The Model’s faceted analysis classification characteristics are also evident in the knowledge domains, showing the top level categories of an information architecture, but also accommodates the processes, phases, and core values that, although having hierarchical characteristics in a procedural sense, do not typically fit into the structure modeling of an organizational chart or easily fit into the bibliographic nature of a knowledge management system.

Strengths

The International EMBOK Model’s tree nature is rich in its representational power (Kwasnik, 1999) as well as robust in its breadth and depth. The faceted analysis nature of the Model illustrates its flexibility and user-specific strengths as a classification system. The facets may be studied independently, provide budgetary line items for funding (O’Toole, 2006), and modularize the division of labor (including job description development), training, and competency assessment in a job creation or credentialing context, as illustrated in Table 2. This faceted approach also illustrates the comprehensiveness of the Model and provides a structure for continuous improvement using the processes, phases and core values in a planning system and to collect historical data, as shown in Table 3.

Table 2. Job Analysis for Financial Management (source: Silvers & Nelson, 2005)

Table 3. Planning System for Financial Management (source: Silvers & Nelson, 2005)

The Model has also introduced a controlled vocabulary that provides discrete and standardized terminology, which will facilitate electronic knowledge management using its object oriented modeling character with each element or cell representing an object (Silvers, et al, 2006), as well as semantic coherency across all industry sectors. Given that the suggested terminology has its roots in traditional project management, the application of the EMBOK as a classification system will also facilitate the incorporation and adaptation of knowledge from industries outside the events industry. This also illustrates the congruity of the various events industry sector competencies that will facilitate foreign and mutual credential recognition and the commonality of skill sets that will facilitate career transition and labor mobility.

The comprehensive nature of the Model provides a scalable structure that reflects the total picture of industry’s activities. Although different event sectors will have different applications of the Model including differing terminology, techniques and elements, the processes, phases, core values, and domains remain consistent, and whilst the scale of an event project may change, the scope of responsibilities does not; only the applicability of the functions (classes) will be different. The Model does provide an exhaustive, stable, and balanced structure — the qualities necessary for a good classification system that supports effective data collation (United Nations Statistics Division, n. d.), yet still provides flexibility for expansion.

Weaknesses

The comprehensiveness of the International EMBOK Model does pose some difficulty regarding its complexity, ranging from legibility to intimidation. Although the Model represents a comprehensive structure of primary facets developed using competency guidelines from a broad spectrum of industry sectors, additional facets are necessary to provide a deeper classification system, such as the taxonomy suggested by Silvers (2004). In addition, the Model still requires the development of considerable explanatory material, a chicken/egg conundrum because its newness and relative obscurity has limited the resources available to create such materials. The overwhelming nature of the Model could, however, be considered a strength in that it illustrates the scope of responsibilities associated with the occupation.

There is an instance within the controlled vocabulary for the Model that may require clarification or a homographic modifier to ensure unique descriptors: the term “communication” is found in both the processes and the operations domain. Although there is significant commonality in the skill set for each instance, the tools and techniques vary according to its application, e.g. communications in general (process) and on-site communications (operations). In addition, a thesaurus for the controlled vocabulary has yet to be developed. This thesaurus needs to reflect the sector-specific variations within the industry as a whole (e.g. donor cultivation versus sponsorship) as well as definitions and international equivalents.

Challenges

The International EMBOK Model needs validation; testing its efficacy in applied contexts as both a data warehouse and a management system for a variety of events in a variety of industry sectors as issues get more narrow and expanded at the same time in different contexts and at different times. Its adoption as a classification system will require industry-wide “thesis-antithesis-synthesis” review and agreement on the alignment with sector-specific knowledge systems. Awareness of the Model needs to be raised and its applicability throughout the scope of the industry needs to be illustrated.

The biggest challenge for the Model, however, is its acceptance by an at-present fragmented industry that has resisted integration and recognition of its commonality and holistic character. Industry and academic factions exist, each with their own agendas, initiatives and isolated programming, which raises numerous political and economic issues. Despite the many benefits that may be accrued if the full scope of the horizontal and vertical nature of the events industry was embraced, resistance to amalgamation remains persistent across various industry sectors, even though the development of the Model was based on an affinity analysis of all industry sectors.

This separatist attitude may be beginning to change, evidenced by the acknowledgment of measuring the economic importance of the meetings industry as an important tool for gaining political, economic, and professional recognition (World Tourism Organization, 2006). However, there is still a lack of national, international, and industry-wide recognition of commonality. In the North American Industry Classification System (NAICS), for example, conventions and trade shows are included under Administrative and Support and Waste Management and Remediation Services (code 56), festivals and sports events are included under Arts, Entertainment, and Recreation (code 71), and weddings and social events are classified under Caterers under Accommodation and Food Services (code 72) (U.S. Census Bureau, 2002). Since these classifications are used to determine the economic and employment statistics, it prompts one to contemplate the impact and stature the entire events industry could have if it were recognized as a single holistic economic entity.

Opportunities

The opportunities associated with adoption of the International EMBOK Model as a classification system are numerous. As a structure for knowledge management it will serve as framework for the collection and aggregation of data — a superset of categories into which information, statistics, and heuristics may be placed. The faceted nature of the Model as a heuristics system will allow unlimited knowledge capture and classification for flexible and easy retrieval according to the user’s objectives and areas of interest.

As a curriculum framework the Model will be clearly useful for students, helping them understand the theories in event planning, as well as suggesting a research menu that facilitates the creation of new knowledge through applied research, the significant link between academe and industry. This will provide students with learning opportunities and industry practitioners with data that can serve as a foundation for good decision making, event development, and strategies for continuous improvement.

The Model also provides an agenda for professional development programming created by industry associations to meet the lifelong learning needs of their constituencies, as well as competency guidance that may enhance credentialing programs, training initiatives, and hiring and promotions criteria. Industry agencies and practitioners may also adapt the Model for organizing tasks and personnel, as well as developing tools and tactics that may expand their markets and profitability. Industry sectors and the industry as a whole may find the Model useful for establishing standards in collaboration with or in response to regulatory requirements or funding or permission guidelines devised by governmental agencies and non-governmental organizations, which may include an agreed-upon standard of occupational equivalencies that can lead to foreign credentials recognition.

THE MISSING LINK — EVENT CLASSIFICATION

The International EMBOK Model suggests a flexible classification system for knowledge management, organizing functions, and standards development for events management. What has yet to be identified, however, are the scale and scope variables that will allow events to be fully classified. The scale of an event could be small, medium, large, or mammoth (or other designations) and the challenge will be putting agreed-to numbers to these qualifiers. The scope of an event could be minimal to expansive and would encompass a number of spatial, temporal, and constituent variables that suggests a paradigm approach to classification wherein entities are described by the intersection of attributes (Kwasnik, 1999). An event classification system, as opposed to an events management classification system, would be a magnitude rating system capable of providing “if/then” design guidance particularly in the context of expertise urgency.

Rating systems use a set of attributes (dimensions of interest) to develop a meaningful description of patterns of similarity and difference (Kwasnik, 1999) that may serve as a shorthand descriptor of a complex entity. For example, seismologists use the Richter Scale to express the magnitude of seismic energy released by an earthquake and the Mercalli Intensity Scale to express the variable effects of an earthquake (Louie, 1996); the U.S. Department of Homeland Security uses the color-coded Threat Level Advisory System of red/severe, orange/high, yellow/elevated, blue/guarded, and green/low to indicate the potential risk of terrorist attacks and communicate the need for protective measure implementation; and meteorologists use the Saffir-Simpson Hurricane Scale to gauge the likely landfall damage a hurricane will cause based on the velocity of sustained winds to provide predictive guidance for emergency response. Each of these uses the shorthand (i.e. 6.3 on the Richter scale, threat level yellow, level 3 hurricane) to quickly communicate a set of conditions, expectations, and action advisories or procedures based on a set of measurable attributes.

This diagnosis-prognosis capability of a magnitude rating system would be extremely useful in the events industry in the contexts of design guidance, risk management guidance, competency guidance (e.g. experience level), and regulatory guidance. Some governments and agencies have used attendance numbers as the baseline for certain statutory requirements and “if/then” guidelines. South Africa’s new Sports and Safety legislation, for example, requires a “qualified event organizer” (among other things) for any event of 2,000 people or more (Ronan, 2004). The National Fire Protection Association’s Life Safety Code (NFPA, 2000), adopted by most jurisdictions in the U.S. and elsewhere, specifies any assembly occupancy of 6,000 requires a life safety evaluation and occupancy loads exceeding 1,000 require a trained crowd manager per every 250 occupants (unless exemption given by the Authority Having Jurisdiction). Numerous event safety guides offer attendance volume guidance for calculating the provision of sanitary facilities and medical services (see, for example, Emergency Management Australia,1999; FEMA, n. d.; HSE, 1999; and Standards South Africa, 2004).

What must be determined for such an event classification system are the attributes of interest and the metrics (i.e. properties, features, characteristics, or mathematical measurements) for each, which must be established based on quantified and qualified statistical research. Otherwise, such a system would be as subjective (and often self-designated) as the ubiquitous star rating system for hotels and the capacity/capability tier rating for convention cities, for which there seems to be no definition of the measurement criteria (Phillips, 2006). However, the development of such a magnitude rating system would provide the framework for collecting the heuristic intelligence and comparable statistics that could provide an internationally-validated set of descriptors of use to event organizers and the jurisdictions in which the events occur. The beginnings of such a collection system may exist within the IFEA-sponsored eSAFE initiative developed at Temple University (Hu & Racherla, 2005).

As with the controlled vocabulary classification system for event management embodied in the International EMBOK Model, unique and unambiguous terminology (or numerics) must be used for attributes within the event classification system to ensure explicit specificity. Each attribute must have a name and a value, and the descriptors must be semantically neutral to prevent conveyance of a positive or negative opinion or quality level connotation. These identifiers will facilitate the ability to adopt and adapt intelligence from similar events as well as events in different contexts with equivalent attributes.

Certain attributes have been identified in the Life Safety Code (NFPA, 2000) and event safety standards (HSE, 1999; Standards South Africa, 2004), and the eSAFE system identifies four attribute sort fields. A set of suggested attributes and their significance are shown in Table 4 and an example of a paradigm-style classification matrix with suggested metrics (not yet quantified) is shown in Figure 3 in order to provide an illustration of the shorthand-descriptor potential of such an event magnitude rating system. Many of these factors are included in a risk analysis to measure probability and severity, yet other attributes also assessed in a risk analysis may be pertinent, including the nature of the event organization (e.g. sufficiency and competency of human resources, volunteer or paid professional staff, presence or absence of policies and procedures, etc.), the event host (e.g. corporate, non-profit, government; suggesting presence or absence of rituals or restrictions), or the event attractions (e.g. hazardous or high-demand activities, food and alcohol service, special effects, etc.). Given the broad spectrum of possible attributes and the uniqueness of even long-standing recurring events, a single rating matrix cannot anticipate every aspect of an event and would not be a substitute for a proper risk analysis, but could provide a starting point for risk assessment that identifies design and response guidance tipping points.  In addition to the risk management potential, these attributes and tipping points may serve to guide event permission, funding, and sponsorship decisions for major stakeholders (O’Toole, 2006).

Table 4. Suggested event rating attributes

Reprinted with permission from Risk Management for Meetings and Events (Butterworth-Heinemann, 2008)

Figure 3. Suggested event rating system matrix

Please note that these descriptors or descriptions are for illustration purposes only and do not reflect codified data.

CONCLUSION

Classification systems provide a way of describing items and ideas, a way of locating information, and a way of seeing relationships. Due to its hierarchical tree and faceted analysis approach characteristics, the International EMBOK Model has great potential for serving as a classification system that can provide the framework for capturing the ever-increasing wealth of knowledge in the industry for knowledge management, curriculum development, and competency assessment purposes. Achieving this potential will require industry-wide validation, a challenge hampered by the segmented status of the various industry sectors. But recognition of the commonalities in occupational responsibilities across all sectors may be facilitated by using the International EMBOK Model as the framework for sharing knowledge and expertise, recognizing mutual strengths and opportunities, and to contribute to the elevation of the profession. The final component of the EMBOK should be an event classification system that defines and codifies the attributes of events so that practical and predictive design guidance may be established based on common conditions, causes, expectations, and outcomes that acknowledge the variable magnitudes of scope and scale.

REFERENCES

British Standards Institution (BSI) (2006). British Standard BS8901: Sustainable Event Management System, Guidance for use, 2nd Draft. British Standards Institution

Bruno, D. & Richmond, H. (2003). The truth about taxonomies. Information Management Journal, Mar/Apr 2003. Retrieved March 17, 2005 from http://www.findarticles.com/p/articles/mi_qa3937/is_200303/ai_n9184320/pg_1

Emergency Management Australia (1999). Safe and Healthy Mass Gatherings, A Medical, Health and Safety Planning Manual for Public Events. Australian Emergency Manuals Series, Manual 2. Commonwealth of Australia. Retrieved May 28, 2002 from http://www.dhs.sa.gov.au/pehs/publications/ema-mass-gatherings-manual.pdf

Factiva (n. d.). Factiva Synapse White Paper KMS101: Taxonomies, Ontologies, Thesauri and Authority Files: The Key to Better Information Retrieval. Retrieved November 9, 2006 from http://www.factiva.com/collateral/files/synaptica20WP.pdf

Federal Emergency Management Agency (FEMA), (n. d.). Special Events Contingency Planning. FEMA

Heath and Safety Executive (HSE) (1999). The Event Safety Guide: A guide to health, safety and welfare at music and similar events. Heath and Safety Executive, U. K.

Hu, C., & Racherla, P. (2005). eSAFE: An online knowledge based system for safe festivals & events. Perspectives on Tourism & Technology No. 3. National Laboratory for Tourism & eCommerce, School of Tourism and Hospitality Management, Temple University

KMconnection (n. d.). Faceted Classification of Information. Knowledge Management Connection. Retrieved November 10, 2006 from http://www.kmconnection.com/DOC100100.htm#Defining_faceted

Kwasnik, B. H. (1999). The role of classification in knowledge representation and discovery. Library Trends, Summer 1999. Retrieved March 17, 2005 from http://www.findarticles.com/p/articles/mi_m1387/is_1_48/ai_57046525

Louie, J. (1996). What is Richter Magnitude? Retrieved November, 2006 from http://www.seismo.unr.edu/ftp/pub/louie/class/100/magnitude.html

National Fire Protection Association (NFPA) (2000). NFPA 101 Life Safety Code, 2000 edn. National Fire Protection Association

O’Toole, W. J. (2006). Personal correspondence

Phillips, G. (2006). Tier Be or Not Tier Be. Convene Magazine, February 2006. Retrieved November 16, 2006 from http://www.pcma.org/resources/convene/archives/displayArticle.asp?ARTICLE_ID=5250

Ronan, P. (2004). Safety at Sport and Recreational Events Bill. Draft legislation. Department of Sport and Recreation South Africa

Silvers, J. R. (2004). Global knowledge domain structure for event management. In Z. Gu (ed.), Conference Proceedings, Las Vegas International Hospitality and Convention Summit (pp. 228-245). University of Nevada Las Vegas

Silvers, J. R. (2005). The potential of the EMBOK as a risk management framework for events. In Z. Gu (ed.), Conference Proceedings, 2005 Las Vegas International Hospitality and Convention Summit. University of Nevada Las Vegas

Silvers, J. R. & Nelson, K. (2005). Introduction to Financial Management for Meetings & Events. Speaking of Events

Silvers, J. R., Bowdin, G. A. J., O’Toole, W. J., & Nelson, K. B. (2006). Towards an international event management body of knowledge (EMBOK). Event Management, 9(4), 185-198. Cognizant Communications Corporation

Silvers, J. R. (2008). Risk Management for Meetings and Events. Butterworth-Heinemann (in press)

Society of Fire Protection Engineers (SFPE) (2002). Engineering Guide to Human Behavior in Fire, Review Draft. Retrieved March 22, 2003 from http://www.sfpe.org/sfpe/pdfsanddocs/DraftHumanBehaviorGuide.pdf

Standards South Africa (2004). SANS 10366:2004: Health and safety at live events, Edition 1. Standards South Africa

UNSPSC (2001). Using the United Nations Standard Products and Services Code. Granada Research. Retrieved December 2, 2006 from http://www.unspsc.org/AdminFolder/Documents/UNSPSC_White_Paper.doc

U.S. Census Bureau (2002). North American Industry Classification System. Retrieved November 10, 2006 from http://www.census.gov/epcd/www/naics.html

U.S. Department of Labor Office of Administrative Law Judges (DOL) (1991) United States Department of Labor Dictionary of Occupational Titles, 4th edn. Retrieved June 14, 2005 from http://www.oalj.dol.gov/LIBDOT.HTM

United Nations Statistics Division (n. d.). UN Glossary of Classification Terms. Retrieved November 19, 2006 from http://unstats.un.org/unsd/class/family/glossary_short.htm

World Tourism Organization (2006). Measuring the Economic Importance of the Meetings Industry, Developing a Tourism Satellite Account Extension. Retrieved December 6, 2006 from http://www.iccaworld.com/cnt/docs/UNWTO_Complete%20Stats%20Report.pdf