By: Lila Rao and Maurice McNaughton (Mona School of Business and Management – The University of the West Indies) of the Disaster Preparedness through Open Ontologies Project

Summary:

•    Many Small Island Developing States (SIDS) do not have the resources necessary for the development of comprehensive Disaster Recovery Plans (DRPs) and would benefit from collaborating to develop them.

•    This OCSDNet project uses a collaborative approach for the development of a common language/vocabulary, which will be used to represent the domain of DRP in the form of an ontology.

•    This project is focused on islands in the Caribbean region and a critical first stage of the development process required identifying and understanding the function of the existing stakeholders that have some level or responsibility or interest in DRP whether at the local or regional level.

•    So far we have developed a partnership with two key stakeholders and they have been extremely willing to discuss their initiatives and share the documents that they have developed/are developing. We have found that stakeholders are very willing to share their information and documentation once they understand how our work aligns with and can support their strategic objectives. Thus, the best way to establish this collaboration is to show how there will be mutual benefits from sharing resources.

—-

The range, incidence and scale of recent natural disasters in the Caribbean region highlight the need for a holistic approach to disaster planning, preparation and recovery. The primary natural disasters facing the islands of the Caribbean are earthquakes and hurricanes although some are also prone to volcanic activity. The flooding and torrential rains that can come with these hurricanes or even tropical storms can also be considered natural disasters themselves [1].

1
Figure 1 – Hurricane Ivan (2004) [2]
Figure 2 - Seismic and plate tectonic setting (Caribbean Region) [3]
Figure 2 – Seismic and plate tectonic setting (Caribbean Region) [3]
As recently as August 2015, the island of Dominica was hit by Tropical Storm Erika which resulted in the death of 20 persons and the need for the country’s prime minister to seek international help after the cost of the storm’s damage was estimated to be half the country’s annual GDP [4]. The major damage was not from the wind, but rather the flooding and mudslides caused by the rainfall that the storm brought with it.

Figure 3a - Destruction from Tropical Storm Erika (Dominica) [4]
Figure 3 – Destruction from Tropical Storm Erika (Dominica) [4]

In January 2010, a devastating earthquake with a magnitude of 7.0 struck Haiti, killing more than 160,000 and displacing close to 1.5 million people [5]. The scale of the disaster made the coordination of the relief effort extremely complex and difficult.

Figure 4a - Destruction from Earthquake in Haiti [6]
Figure 4 – Destruction from Earthquake in Haiti [6]
Other natural disasters to affect the Caribbean region in recent times include:

  • In 2007: Hurricane Dean caused significant damage to the islands of Jamaica and St. Lucia
  • In 2004: Hurricane Ivan caused significant damage to the islands of Grenada, Jamaica and the Cayman Islands
  • In 1995: the Soufriere Hills volcano in Monserrat became active resulting in a significant portion of the population to evacuate the country

Disaster recovery plans (DRPs) seek to ensure that entities function effectively before, during and following a disaster in order to save lives and reduce economic losses [7]. However the development of comprehensive DRPs require significant resources, the coordination and collaboration of a number of stakeholders and access to a number of disparate data sources. Given the vulnerability (which seems likely to increase with climate change) of the Caribbean region, and indeed many Small Island Developing States (SIDS), to natural disasters and their lack of the resources required to develop, test and maintain comprehensive DRPs, it would be beneficial for countries and public and private sector agencies in the region to collaborate and share data, information and knowledge in order to make DRP development more cost effective and improve DRP quality and effectiveness.

A major obstacle to this collaborative approach is the ambiguity, inconsistency and communication problems that often arises when autonomous entities must work together and disparate data sources are to be integrated and shared. We believe that the solution to this problem is the collaborative development of a common representation of the domain including a common language/vocabulary. This representation will be in the form of an ontology, which is defined as a formal description of a domain that can be shared among different applications and expressed in a language that can be used for reasoning [8]. Our project seeks to use a collaborative approach to the development of a DRP ontology which will be developed as an open resource for others to reuse.

An important objective of the project is to demonstrate to key stakeholders and DRP entities throughout the Caribbean region that there are considerable benefits to be gained from this collaborative approach to solving common regional problems and more specifically Disaster Recovery Planning. Given that the ontology will be a representation of the domain and will provide a consistent/common vocabulary through which all stakeholders can communicate it is imperative that these stakeholders are involved in its development. Additionally, the existing documentation of these entities is a valuable input into understanding the domain as it exists. Therefore we recognized that the first stage of this project would not be about the development of the ontology, but rather about understanding who the existing stakeholders and entities are, what their objectives are and what documents they have developed or are in the process of developing. We want to ensure buy-in from these regional entities who will contribute to and use the open resource. Therefore, over the past months, we have been focused on building such partnerships and examining their existing documentation.

Thus far, we have been fortunate to secure buy-in from two important existing regional stakeholder groups, The Enhancing Knowledge & Application of Comprehensive Disaster Management (EKACDM) group and The Caribbean Disaster Emergency Management Agency (CDEMA).

Partnership 1:

The Enhancing Knowledge & Application of Comprehensive Disaster Management (EKACDM) group (http://www.uwi.edu/EKACDM/index.aspx) is working on a five-year research project to implement the CARICOM (Caribbean Community and Common Market) Enhanced Comprehensive Disaster Management (CDM) Framework and has as one of its key outcomes, “the creation of a regional network which generates, manages, and disseminates knowledge on disaster management.” The importance of this relationship was not just that our work aligned with some of their key outcomes, but also that they had already established relationships with regional stakeholders and were very willing to introduce us to relevant people.

This led to the establishment of a relationship with the Caribbean Disaster Emergency Management Agency (CDEMA) a stakeholder that we had recognized from the outset of this project was going to critical to the regional collaboration we were seeking. To get interest from the EKACDM group we demonstrated to them that our project is aligned with a number of the intended outcomes of the EKACDM project, thus we have been able to build a collaboration of mutual benefit. One shared intended outcomes defined by EKACDM is improved knowledge generation, and management and dissemination for Caribbean Disaster Management. We demonstrated that the development of an open DRP ontology would help achieve these shared intended outcomes. EKACDM therefore agreed to provide support to our project through key stakeholder consultation, access to domain expertise, access to their documents and assistance in the validation of the ontology when developed. The documents are already proving to be useful as many of them have glossaries which are valuable in understanding and developing the common vocabulary for the domain. In turn, EKACDM will have access to the final ontology and it will become an important component of the group’s Caribbean Disaster Portal which they are developing.

Partnership 2:

Through our relationship with EKACDM, we were able to link into a second important regional stakeholder group, the Caribbean Disaster Emergency Management Agency (CDEMA), which has also committed to lending support to our project. CDEMA is a regional inter-governmental agency for disaster management in the Caribbean Community (CARICOM). The Agency will be hosting one of the largest gathering of professionals and academics in the field of disaster management in the Caribbean region in December 2015 and we have been invited to present. This provides us with an ideal opportunity to build awareness of our project throughout the region. We are also in discussion with CDEMA about hosting a stakeholder consultation workshop at their event.

From the work carried out so far, we have leveraged important relationships with groups who are already working within the domain of interest and building partnerships with these groups. Given that these are already established groups with a strong regional footprint, we found the best strategy for getting this buy-in was to demonstrate how our project supports and creates synergy with their already existing objectives.

Given our affiliation with Mona School of Business & Management (MSBM), one may wonder how a business school becomes associated with a project of this nature. One of the research units at MSBM, the Centre of Excellence for IT-enabled Innovation, led by Maurice McNaughton, has been actively engaged in action research on the emergence and application of open ICT Ecosystems in the Caribbean, in particular open source software, open data and mobile innovation ecosystems. All of these ‘open systems’ share some common characteristics with the Open Science narrative, and generally appear to offer new development opportunities and mechanisms for the SIDS of the Caribbean that have traditionally been disadvantaged by scale inefficiencies and small disconnected markets.

On the technology side, we are currently exploring the use of semantic web and linked open data concepts, standards and technologies, which will be a key enabler of the principles of open data/open knowledge that underpin this project. We currently have one of our graduate students evaluating Apache Jena as a free and open source framework for building semantic web and linked data applications. We are working towards the development of the initial prototype of an online interactive vocabulary based on the Disaster Recovery Planning ontology that will be demonstrated at the upcoming CDEMA conference.

The lesson we have learnt so far is the importance of the establishment and development of relationships with the key entities and stakeholders of the domain, without which the project would have limited relevance and impact. We have been fortunate that those we have contacted so far have been very willing to work with us and share information, largely because we have been able to demonstrate the relevance and mutual benefits that our work provides to them. However, we do recognize that, thus far, these relationships have evolved in a rather informal way and are based largely on individuals and therefore if these individuals are no longer around new relationships may need to be established. The idea of how best to develop sustainable institutional relationships to maintain the collaborative nature of projects of this kind will require further thought and exploration. Any ideas or best practice examples of how to achieve this in a sustained way are welcome in the comments section below!

 

References

[1] http://www.oas.org/pgdm/document/BITC/papers/gibbs/gibbs_02.htm

[2] http://www.trinidadandtobagonews.com/forum/webbbs_config.pl?md=read;id=2447

[3] http://repeatingislands.com/2015/01/14/univ-of-west-indies-caribbean-countries-should-brace-for-eventual-massive-earthquake/

[4] http://www.theguardian.com/world/2015/sep/04/dominica-asks-for-aid-tropical-storm-erika-damage

[5] http://time.com/3662225/haiti-earthquake-five-year-after/

[6] http://www.dailymail.co.uk/news/article-1243540/Haiti-earthquake-pictures-Shell-shocked-survivors-roam-streets-scavenging-scraps.html

[7] http://www.odpem.org.jm/DisastersDoHappen/NationalDisasterPlans/tabid/57/Default.aspx (National Damage Assessment Plan)

[8] Gruber, T. R. (1995). Toward Principles for the Design of Ontologies Used for Knowledge Sharing. International Journal of Human-Computer Studies, 43(5-6), 907-928