Figure 1 - Data-Sets by Organisation
Figure 2 - Distribution Policies
Figure 3 - Data-Set Storage Media
Figure 4 - Data-Set Distribution Media
Figure 5 - Language of Data-Sets
Figure 6 - Language by Agencies
Figure 7 - Data-Set Frequency

Introduction

Important policy decisions facing Malaysia require most careful analysis in areas of natural resource management, policies on sustainable development which impact environmental sustainability and future growth patterns that minimize environmental impacts. The capacity to conduct independent research and analysis on these areas depends largely on the availability of adequate data. However, the fact remains that although data exists it is widely scattered and not accessible in an integrated manner.

With this in mind, under the aegis of the UNDP project MAL/93/016 entitled "Support to Environment and Natural Resource Research and Development Capacity", the Institute of Strategic and International Studies (ISIS) Malaysia ventured, early 1996, to develop an information system on environment and natural resources, for the purposes of policy analysis and planning. The undertaking involved (i) a comprehensive assessment of existing information on environment and natural resources and (ii) design and implementation of a national ‘one-stop’ centre for on-line access to that information.

At the outset a nation-wide survey was carried out, targetting potential data repositories and users of information on environment and natural resources. For the purposes of the survey the definition of environment was based on that found in the Environmental Impact Assessment (EIA) Guidelines of Department of Environment, ie ‘the surrounding zone (the specific zone to be affected by the project), all natural resources (physical and biological and human resources [people, economic development and quality of life values] )’. The definition interprets ‘environment’ in the broadest possible sense incorporating natural resources as an integral part of it. Subsequently the three major divisions used for data classification and organization were the physico-chemical, biological and human aspects of the environment.

Having completed the assessment of existing information, the next step was to develop a management information system (MIS) on environment and natural resources. This resulted in a hard copy and digital output of the Environment and Natural Resource Database Catalogue (ENRDC) which, together with the supplementary modification of the ISIS server and updating of the ISIS Library information system, was designed to be ready for incorporation into the Internet to coincide with the pending launching of the ISIS homepage. A workshop to launch the ENRDC was held mid 1997 for the purposes of disseminating awareness of the MIS, proceedings of which are summarized as Attachment to the paper.

Within the scope and limitations of the study it was not possible to collect the actual data from the data sources contacted, given the time constraint and propriety rights of each data repository. Information compiled in the ENRDC is necessarily confined to metadata or data which discusses content, quality, conditions and other characteristics of data, from which the feeding and updating of actual data would naturally follow in phases consequent to the completion of this project.

The data sources contacted in the survey were classified into several organizational types to facilitate analysis, as follows:

(i)Government bodies, ie ministries concerned with day-to-day governance of Malaysia.,

(ii) government agencies or research agencies and government organizations established with specific intention of conducting research or collecting, monitoring and collecting data, which have the potential to become privatized, ie join the private sector.

(iii) education group ie universities, colleges, polytechnics, technical institutes

(iv) special interest groups ie NGOs and non-profit organizations, professional associations active in the environmental field.

(v) private sector

Metadata attributes from the survey used for analysis included collection authority, distribution policy, frequency of updating, purpose, time coverage, content etc. Although analysis was carried out right down to the data discipline groups ie physico-chemical, biological and human aspects, for the purposes of this presentation, the findings that follow reflect the overall trends without differentiating between the three groups.

As per Fig 1 it was found that the collection and reporting is largely concentrated in government agencies and bodies. These institutions accounted for almost two-thirds of the data-sets identified and catalogued. This appears consistent with their mandate as regulator and administrator of environment and natural resource (E & NR) policy and planning.

The general public has access to a large majority of the E & NR information collected by both public and private institutions (Fig 2). Less than a quarter of the data-sets have dissemination policies which limit or prohibit access to the public. In addition it was found that overlap or duplication of effort was minimal in terms of purpose, geographic coverage and time period coverage; which reflects, perhaps the unique mandates of each organization surveyed.

A total of 39% of the reported data-sets were found to be stored in hardcopy format, with only a small portion (6%) in strictly digital format. As Fig 3 shows, the largest portion of data-sets, just over 55%, were stored in both hard copy and digital media. The distribution format, as per Fig.4, highlighted the preference for report and publication formats at 58%, with only a third as many, ie 18% reportedly distributed in hardcopy file format and half that again, ie 9% in database format. Maps and spreadsheets account for another 8% of data-set distribution formats with the remaining 7% falling in miscellaneous categories. It appears that on-line services, Internet connections and other purely digital means of sharing data are still in very limited use as environment and natural resource data-set distribution channels.

Fig 5 illustrates the distribution of data-sets by language in which it is stored and distributed; showing an even distribution of Malay (49%) and English (48%), with 3% attributed to other languages. A breakdown of language used by data source reveals, not surprisingly, that government bodies produce over 60% of their data-sets in Malay and only 36% in English, as in Fig 6. Government agencies, however, together with NGOs, produced close to 70% of their data in English, reflecting current market pressures to use this language.

Frequency of updating is shown in Fig 7, where only 5% of data-sets are updated either daily or weekly; and about one quarter are updated monthly or yearly. Over half are updated every year or less, while just under 20% are not updated at all. Slightly under a fifth of the data-sets were compiled as one-time data collection activities, indicating static information that will eventually be outdated.

The great majority of data-sets related geographical coverage to a jurisdictional measure, ie by state, district, local authority area, etc., as opposed to geomorphic, socio-cultural or map reference measures. Data covering the whole nation represent about one-third of these data-sets although a significant number of these do not include Sabah and Sarawak. Also, similar data being collected by different agencies for different areas point to a potential for future data incompatibility problems.

Historical coverage

Further analysis indicated that historical trend data for each particular data-set is available. Examination of the bulk of the data-sets highlighted two patterns. The first is that three- quarters of these data-sets were initiated in the 1980s (25%) and 1990s (54%) and therefore reflect fairly recent information. A further 14% were commenced in the 1970s and 5% in the 60s. The second is that a large proportion of the data-sets are subject to on-going collection activities and therefore can be expected to be available into the future.

Action proposals put forward as a result of the comprehensive assessment above were largely intended to reveal important areas where improvements can be made. These are grouped as follows:

1. Implementation of the ENRDB Catalogue

(i) The Catalogue should be continually updated preferably every four to six months to reflect the addition of new E & NR data-sets pertinent to policy analysis and planning; and continual contact with each of the identified organizations maintained over the life of the Catalogue.

(ii) To increase accessibility to the data and ensure widespread use, on-line access should be offered.

(iii) Promotion of the Catalogue should be conducted through a variety of ways, eg newsletters, press releases, workshops etc.

2. Establishment of Data Repository

3. Increasing Role of NGO Information Producers

(i) Greater emphasis is to be placed on encouraging NGOs to become involved in the collection and reporting of E & NR information

(ii)Special Interest Groups, in particular should be focussed on, given their diverse range of activities.

(iii) Incentives must be offered to such groups to ensure their long-term participation and involvement.

4. Providing Wider Basis of Information

(i) More attention should be given to development of data-sets outside of the realm of Human Environment, which currently account for two-thirds of the data-sets in the Catalogue

(ii) Given that this initiative will be a long-term effort, widening the scope of organizations included the Catalogue may provide short-term gains

5. Encouraging Use of Digital Media

(i) Since almost 80% of data-sets identified are available only in hard copy, greater attention should be given to promoting the use of digital storage and distribution.

(ii) The benefits of digital media distribution should be illustrated to the public and private sectors.

6. Encouraging Standards for Storage and Distribution Formats

There was a consensus on designating a focal point to establish rules and procedures for data collection, and help get data into a more formalized state.

The ENRDB which is supposed to be dynamic can be upgraded by (i) distributing through Internet, or (ii) publishing in CD Rom and/or (iii) establishing a common DB format.

A Quality Control mechanism and regular maintenance of data, by computer-trained personnel, is essential to address the contentious issues of data verification and accuracy, and time lapse.

Given that it was immaterial where the data is located if it is in digital form. an alternative would be to establish linkages to data centers, ie make the necessary linkages to websites.

The concept of subscribing to a system was suggested, whereby, rather than plugging on to one server which may be overloaded, requests can be sent out to the particular repository who then distributes the data.

Commercial aspect

The commercial aspect to data was brought up, whereby users are willing to pay for data, and the revenue collected can further develop the database concerned. The impending privatization of local universities may make the commercial aspect of database handling attractive to these academic institutions, thereby necessitating greater inter-varsity links.

From the above findings it is clear that the tasks to be given priority are identifying the data repository, procuring actual data and addressing the commercial aspects of data management. In the interim, as a follow-up to this project, ISIS’ on its part has initiated a collaborative effort with ENSEARCH, an NGO involved in environmental research, to pool the resources of the two parties in developing the Malaysian Environment Network (MEI-NET).

ATTACHMENT

The objective of the workshop which took place pursuant to the publication of the E & NR Database Catalogue and installation of the relevant Management Information System (MIS), was to generate discussion and get feedback on feasible post-project work. In the process, it would help identify the next step in order to chart the future direction of the ENRDB developed at ISIS. Findings of the survey were presented, followed by demonstration of the MIS, preview of the GIS as a potential ‘next step’, and similar databases developed by other organizations.

In the ensuing discussions , the question of data security was raised, as well as the difficulty of private sector developing capability in this subject. Cultural differences were cited, with the local populace being relatively tardy compared to their western counterparts. On top of that data is scattered and not systematically collected; the common observation being: our problem is that data, which is absolutely essential to make better decisions, has simply just not been ‘brought together’. At this, it was suggested that a focal point be designated to establish rules and procedures for data collection, and help get data into a more formalized state. Once a workable system is worked out, a snowball effect is predicted.

Given that the ENRDB is supposed to be dynamic, options put forward on how to upgrade it include (i) distribute through Internet, (ii) publish in CD Rom and (iii) establish a common DB format. A contentious issue was data verification and accuracy, with data quality being of utmost importance. The time lapse between collecting data and when it is finally ready for use results in outdated data which is misleading. In addition, experience in data handling has invariably revealed the existence of ‘gray’ literature, which may account for 10 times the volume of properly published literature. The problem of distinguishing ‘rubbish’ from value, or ‘pulling out gold from the dust’ would require a Quality Control mechanism and regular maintenance of data.

As regards location, a point raised was that it was immaterial where the data is located if it is in digital form. Also, transforming the data into standard format would be a costly exercise. Recognizing that data is currently located in many places, the question arose as to how effective one center would be. An alternative would be to establish linkages to data centers, ie make the necessary linkages to websites. The concept of subscribing to a system was suggested, whereby, rather than plugging on to one server which may be overloaded, requests can be sent out to the particular repository who then distributes the data.

Reservations on the part of the private sector were expressed, viz. many agencies may not be willing to part with their hard-earned data, and may only release it at a price. This brought up the commercial aspect to data, ie within a subscriber system, automatic remuneration to the data provider is triggered as soon as a piece of data is ‘hit’. The actual mechanism would depend on how the database is set up. Experience outside the country has shown that with government endorsement, users are willing to pay for data, and the revenue collected can further develop the database concerned. The impending privatization of local universities may make the commercial aspect of database handling attractive to these academic institutions, thereby necessitating greater inter-varsity links.

The session on existing databases provided useful management pointers to ISIS which by comparison is a new player in the game. From the demonstration given by the Standards and Industrial Research Institute of Malaysia (SIRIM) where the SIRIM Link database is maintained, it was stressed that the data provider must know the needs of the user and have a far-sighted plan. Quality, reliability and good service are of utmost importance, and tasks of personnel, whether monitoring or keying in must be made clear. It is important to establish infrastructure support, and employ computer-trained people for routine and scheduled maintenance check-up.