See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/254478326

CIMAT: a Knowledge-based System for Developing Criteria & Indicators for

Sustainable Forest Management

Article · January 1999

CITATIONS READS

5 147

7 authors, including:

Mandy Haggith Ravi Prabhu

33 PUBLICATIONS 392 CITATIONS

Consultative Group on International Agricultural Research
93 PUBLICATIONS 3,969 CITATIONS
SEE PROFILE
SEE PROFILE

Herry Purnomo Jasper Taylor

CIFOR and IPB 12 PUBLICATIONS 232 CITATIONS
125 PUBLICATIONS 2,251 CITATIONS
SEE PROFILE
SEE PROFILE

All content following this page was uploaded by Mandy Haggith on 26 April 2014.

The user has requested enhancement of the downloaded file.

 CIMAT: a Knowledge-based System
for Developing Criteria & Indicators for Sustainable Forest Management

Mandy Haggith*, Ravi Prabhu, Herry Purnomo, Anwar Rizal,

Jasper Taylor**, Yurdi Yasmi and Doddy Sukadri

Center for International Forestry Research (CIFOR), PO Box 6596 JKPWB, Jakarta 10065, Indonesia
* Worldforests, 3 Inchmore, Struy, Beauly, Inverness-shire, IV4 7JX, Scotland, UK
** The University of Edinburgh, Institute of Ecology and Resource Management, Edinburg EH9 3JU, Scotland, UK

Email: [email protected], [email protected], [email protected], [email protected],

[email protected], [email protected], [email protected]

From: AAAI Technical Report WS-99-07. Compilation copyright © 1999, AAAI (www.aaai.org). All rights reserved.

Abstract 1. Introduction
The Center for International Forestry Research (CIFOR) Criteria and indicators for sustainable forest management
has carried out research to develop Criteria and Indicators (C&I) are tools which can be used to collect and organise
for sustainable forest management (C&I) in eight countries information in a manner that is useful in conceptualising,
since 1994. In the CIFOR framework C&I are organised evaluating, communicating and implementing sustainable
hierarchically as principles, criteria, indicators, and forest management. The Center for International Forestry
verifiers. Based on an analysis and synthesis of the C&I Research (CIFOR) has been carrying out research on C&I
research a set of C&I has been identified that can form the for the Forest Management Unit (FMU) level from 1994.
starting point for developing localised sets of criteria and This research has taken place in Indonesia, India, Côte
indicators. This set is called the ‘Generic Template’. d’Ivoire, Cameroon, Brazil, Austria, Germany and USA
A knowledge based system called CIMAT (Criteria and (Prabhu et al. 1996, 1998a). C&I are already widely used
Indicators Modification and Adaptation Tool) has now been in certification of forest management, and the development
developed to support the process of developing locally of standards, and are being increasingly adopted by forest
adapted C&I using the Generic Template as a starting point. managers to structure monitoring programs and to inform
In CIMAT all the principles, criteria, indicators, and management decisions.
verifiers of the Generic Template are represented as items In CIFOR's approach, C&I are organised into a four
that can be modified by the user. The modification can only level hierarchy of principles, criteria, indicators, and
take place when users explain their reason for the change. verifiers (PCI&V). The top two levels define the goals or
All the reasons in the system can be made dependent upon values that the concept of sustainable forest management
one another, they can be revised during modification, and can be broken down into, the bottom two levels deal with
they can be counter-argued by other users. The need to variables used to actually measure or monitor progress
represent and maintain these reasons and their towards achievement of these goals in any forest area.
interdependencies is resolved through the use of a reason These four hierarchical levels have been linked
maintenance system (RMS). Another meta-level component conceptually in (Prabhu et al. 1999) to the four basic
of CIMAT is its ability to compare and analyze multiple entities (wisdom, knowledge, information, and data) in
knowledge bases. information theory of (Liang 1994). An example of an
ecological principle, criterion, indicator and verifier is
Keywords: sustainable forest management, criteria and given in Annex 3.
indicators, knowledge-based system, automated reasoning, Based on an analysis and synthesis of its research on
reason maintenance system, argumentation. C&I, CIFOR has proposed a Generic Template that can be
used as an effective starting point for developing locally
adapted sets of C&I (CIFOR C&I Team 1999b, Prabhu et • Forest assessors/certifiers.
al. 1998a). The Generic Template recognises that a single
C&I set is not likely to be universally applicable. Instead it Before we built CIMAT we needed to know what kind
foresees a process of adaptation and customisation to local of a tool our users need, so that we could design a system
conditions, expectations and management objectives. The to be genuinely useful, making their job quicker, easier,
purpose of the Generic Template, is to inject the results of more manageable, cheaper or better quality. Some of the
international comparative research into this development ways a computer could help may be clerical (producing
process, cut down development time and ensure that useful reports, checklists etc). Some may be organisational
structural and conceptual compatibility and rigour is (keeping track of version changes). Some may be human
maintained with other sets of C&I. (helping a team of experts to co-ordinate their work more
The top level of the Generic Template hierarchy consists effectively). In order for us to match the tool to the task,
of six principles of sustainable forest management, shown we needed first to gain a better understanding of the needs
in Annex 3. Each principle is elaborated in terms of and constraints of our users. To achieve this understanding
criteria, indicators and verifiers. A total of 238 statements a three-month-long process of user requirements capture
are included within the hierarchy. has been carried out using an international standard
If the process of adaptation of this hierarchy to local methodology (Mazza et al. 1996) for the work. The
conditions is not to be a daunting challenge it will need to resulting CIMAT User Requirements Document forms a
be simple, understandable and designed to meet user full record of the users’ requirements. A brief summary is
needs, i.e. it must be user-friendly. In order to fulfil these in Annex 1.
conditions a software tool has been built, designed to help
users to carry out modifications to the generic template Requirements Capture Method
simply yet rigorously. This software is called Criteria and The needs of these users were investigated as follows.
Indicators Modification and Adaptation Tool (CIMAT).
This paper presents an overview of CIMAT. Section 2 1. In-house experts and other C&I stakeholders resident
describes the process of determining the needs of potential in Indonesia were interviewed and their views elicited
users of such a tool, resulting in an understanding of our through informal questioning during the interviews.
user requirements. Section 3 describes the software tool Notes of the interviews were taken and a summary
itself, its main components and functionality. Section 4 written.
takes a closer look at the most novel aspect of the tool, 2. An email discussion group was set up, to involve
which is its mechanism for encouraging users to clearly international stakeholders and others, and their views
state their reasons for changes, and allowing subsequent on an initial discussion document about CIMAT were
debate and counter-argument of these reasons. The result elicited. In addition they were asked to respond to a
of using CIMAT is a new set of C&I, and so CIMAT also set of informal questions, similar to those used in the
supports comparison between different C&I sets, making interviews.
use of the history of argumentation. This feature is also
3. Two fictional ‘scenarios’ of use of CIMAT were
briefly outlined in section 4. Section 5 outlines our
written, and feedback sought on these by in-house
evaluation plan and section 6 concludes and points out
experts and on the email discussion group.
future directions for research.
4. As part of the C&I assessment process, Smartwood
and SGS-Forestry, two forest certification companies,
2. User Requirements carried out a simulated forest certification exercise in
Central Kalimantan (Indonesia) using the generic
template. The first part of this exercise was to modify
Who are the users?
the generic template to the conditions of the exercise.
CIMAT is a tool for people who are developing C&I, Protocols were observed of early modifications at a
using CIFOR’s generic template as a starting point. Three preparatory meeting in Oxford, UK, and further
groups of C&I users were identified. localisation during fieldwork planning meetings at
CIFOR immediately prior to the exercise. These
• Experts within CIFOR’s network of C&I researchers
protocols provided valuable input about how C&I are
who have been involved in the evolution of the current
used in a certification process and what modification
C&I sets.
processes are involved. They also provided an
• International C&I stakeholders, including developers opportunity for acquisition of knowledge about how to
of national forest stewardship standards and forest ‘localise’ C&I.
managers.
5. A synthesis of the requirements from the previous users. The result is a network of indicator objects that
steps was generated, in the form of a ‘hitlist’ of 50 change and evolve over time in response to the users’
requirements statements. modifications and adaptations. As the user gives reasons
6. Confirmation and prioritisation of these requirements for modifications, these are also included in knowledge
was carried out by a user survey (involving the email base, as justification objects and assumption objects. The
discussion group and the certification team). The body of knowledge about modifications thus also grows
survey used a formal questionnaire in which the users and evolves over time.
were asked to categorise the statements as ‘Essential’, In addition to the generic C&I knowledge base, and
‘Useful’ or ‘Not helpful’. A simple scoring method knowledge about modifications, there are a host of useful
was used to analyse the results, to produce a ranking general documents providing guidelines for developing
of requirements as high, medium or low priority. The and using C&I (eg. Prabhu et al. 1996, Stork et al. 1997,
high priority requirements are listed in Annex 1. Colfer et al. 1998), carrying out sustainability assessments
using C&I and providing examples or illustrations of C&I
in particular contexts. CIMAT includes a repository of
3. CIMAT overview : the object level such general information. To facilitate this, the existing
information resources and their relevance and usefulness
were investigated, resulting in a C&I information map
CIMAT’s C&I knowledge base (Davenport 1997, Prabhu et al 1998b) which forms the
At the heart of CIMAT is a knowledge base of C&I for basis of the general knowledge base in CIMAT.
sustainable forest management. It is thus a knowledge
system (Stefik 1995). The core knowledge base includes Functionality
CIFOR’s generic template, plus knowledge about how to
modify it and general knowledge about C&I. The CIMAT has two main functions - navigation and
knowledge base is essentially incomplete, and contains lots modification.
of ‘hooks’ upon which users can hang knowledge which is
1. The first function is to support the user in navigating
relevant to sustainability of forest management in the
the C&I information resources, and the hierarchy of
particular context they are interested in. CIMAT invites the
C&I. CIMAT provides navigation support using
user to bring their knowledge to the system, in order to
nested lists and menus. Navigation breaks down into
enhance and build upon the knowledge within it. CIMAT
navigation of the C&I hierarchy; and exploration of
is a knowledge-based system, but it is not a conventional
the general knowledge and information resources
expert system because it does not propose a solution to the
contained in CIMAT.
C&I modification problem, nor will it act as an expert
guiding a user through an assessment of sustainability. 2. CIMAT’s second function is to support the user in
Instead it records the user’s reason for particular modifying their own set of C&I to meet their local
modifications and structures this information in such a way conditions. In addition to the necessary editing
that all these reasons can be used as knowledge for facilities the modification guide includes simple
subsequent modifications. Its role is to manage knowledge, dialogues to prompt users for their reasoning.
rather than to advise, and the 'locus of responsibility'
(Whitby 1988) remains firmly with the user. In designing Other user interface features include a comprehensive
CIMAT we were informed by the work on argumentation help facility and access to utilities such as printing sets of
in (Haggith 1996). C&I in various formats, saving files etc. The user interface
Each principle, criterion, indicator and verifier in the is implemented in the Tcl/Tk scripting language which is
hierarchy is an ‘object’, which can be changed, deleted, platform independent and therefore offers a good
added or moved. In the current prototype knowledge base complement to the knowledge core which is implemented
the C&I are structured as a hierarchy of objects in SICStus prolog (SICS 1995).
represented using the logic programming language, prolog The user can modify any object in the C&I hierarchy,
(Clocksin & Mellish 1981). This enables modular except for the six principles. The six modification options:
knowledge base development and reasoning with the C&I Add, Change wording, Substantive change, Move, Delete,
knowledge in different ways in different parts of the Restore.
hierarchy. The user may add additional remarks, comments For example, to reflect the local conditions in a
and sources of relevant knowledge to all objects. Each commercial forest area in Kalimantan, and to reduce the
object remembers the sequence of modifications that it effort involved in using the C&I set as the basis of
undergoes, so it ends up with its history of how it has been monitoring, the user would probably choose to delete
modified. This provides an ‘institutional memory aid’ for many of the C&I which are most appropriate for
management of sensitive areas of high conservation status, problem solver, which is usually an inference engine, or
such as the ecological C&I concerning genetic variability. theorem prover, which draws conclusions from and makes
They would also be likely to make substantive changes to changes to an underlying knowledge base. In CIMAT this
some of the production-oriented C&I to reflect the kind of process of making changes to the knowledge base is
information they have available as a result of legal carried out by the user. The RMS module maintains
requirements for particular documentation, such as maps appropriate data structures to keep track of the reasoning
and management plans. They would make many of the carried out by the problem solver (i.e. the user) and keeps
indicators and verifiers specific by naming particular these data structures consistent. For more on RMS see
coupes, water courses, company personnel etc. In the (Doyle 1979, Doyle 1992, Forbus & de Kleer 1993).
social C&I, references to ‘other stakeholders’ would be RMSs are often used for solving problems related to
changed by identifying particular local Dayak communities context determination, consequence determination, and
and other commercial interests in the area. A user may belief revision as explained in Kraetzschmar et al. (1997).
choose to group together under a social criterion several Context determination involves determining all logical
indicators concerning conflict by moving them from their consequences of the clauses and the propositions (called
original positions (for example, under the policy principle). context). Determining whether a particular proposition
Every time a user modifies an item, they are asked to should be believed in a certain environment is the
submit a reason explaining why such a modification is consequence determination problem. Any change to the
needed. Such reasons may be new or may already exist in reasons database may result in changes to the truth-value
CIMAT as a previously submitted reason. Reasons may of a proposition. The problem of updating contexts
take the form of facts or assumptions that can be retracted because of changes to the clause database is called belief
or reasserted depending on the context of a modification. revision.
For example, if there is no logging (fact), this can be used For a formal specification of the CIMAT reason
as a reason for removing an indicator about a particular maintenance system and its solution of the context and
logging technique. However this reason would obviously consequence determination and belief revision problems in
be no longer valid if an annual logging plan had been C&I modification see (Purnomo et al. 1999). A brief
approved, i.e. it was now possible to log (new fact). explanation of its use to keep track of the reasoning about
Therefore if ‘there is no logging’ is the reason for modifications is as follows.
removing some items from the hierarchy, CIMAT must be A modification is accomplished by submitting a valid
able to restore all these item if another user counters this reason. The submitted reason can be a new reason or a
argument by pointing out that the ‘annual logging plan has valid reason that already exists in the system. It can be a
been approved’. This gives rise to the problem that there is reason for a C&I object's existence in the hierarchy or it
need for a rigorous way to represent and maintain all can be a reason which counters any other reason in the
reasons and dependencies between them. CIMAT solves this system.
problem by incorporating a Reason Maintenance System
• To add an item, the user must submit a valid reason
(RMS), for more details of which see the next section.
for its existence in the hierarchy - this can either be a
new reason or one already in the system.
4. Argumentation and comparison: • To delete an item, CIMAT establishes the valid
the meta-level reasons in the system justifying its presence, and the
user must counter these reasons by submitting
Some of the most interesting aspects of CIMAT result counter-argument reasons. Again these may already
from it being not only an 'object-level' system, in which exist or be new. A consistency check determines all
C&I knowledge can be represented and manipulated, but other items whose truth-value may change as a result
also containing 'meta-level' elements for reasoning about of the counter-argument. For example, if a number of
this knowledge. One aspect of this is the need to keep track C&I have been included for the reason
of users' reasoning about modifications, which is achieved R1:‘Requirement to ensure high quality construction
using a reason (truth) maintenance mechanism. The other of logging tracks’, and the user is deleting one of them
aspect is CIMAT's ability to analyse the commonalities for the reason R2:‘No additional logging tracks will be
and differences in content and reasoning in two knowledge constructed’, then the other items which were justified
bases developed using the system. by R1, which has now been countered, can also be
deleted. If the user accepts deletion of these items,
Reason maintenance then the net of reasons is updated accordingly.
A reason maintenance system (RMS) is usually one of two • To restore an item, the process is similar to that of
parts of a knowledge-based system. The other part is the deletion, except that the user can either restore an
 object by assigning it a valid justification for being in 5. Evaluation
the hierarchy, or they can counter-argue the reasons
given for its deletion. In the latter case, their counter- An important role of the user requirements is to provide a
argument may affect the validity of other objects in framework for evaluating the first working prototype of
the hierarchy. For example, if a user has deleted a CIMAT. The primary evaluation question is thus ‘Does
number of social C&I on the basis of R3:‘There are no CIMAT do what the users require?’ Our evaluation work is
local community stakeholders’, a second user may aimed at providing useful feedback for future
wish to counter-argue this with R4:‘The local Dayak developments of the system, as opposed to attempting to
people have traditional rights to the forest land’. Once measure improvements in users’ performance of known
this counter-argument (R4) is added, all the items tasks (see Doukidis et al. 94). The user requirements have
deleted for reason R3 can be restored. been used to produce structured evaluation tools and
question sets. Trials of the prototype began early in 1999.
• Moving is the combination of deleting an item and
Early tests have been useful in revealing bugs and interface
adding a new one in any other place in the C&I
problems.
hierarchy.
Preliminary results suggest that CIMAT is being well
• Substantive changing is also treated as the received by users. The argumentation aspect has been a
combination of deleting the old item and adding a new success though it is too early to determine how it will
one. The only difference is that during changing there perform, and how usable it will be, once complex bodies
is no shift in locus of the modified item. of arguments have developed. However, it is already
• Minor wording changes do not require reasons. provoking interesting feedback on methods for handling
alternative, conflicting, opinions amongst teams of users.
Knowledge base comparison Many users wish that the system were available for use
Once a user has modified their C&I they may wish to over the Internet. Another useful piece of feedback is that
compare the result with the original. Two developers or users who may already use sets of C&I from other sources
teams may wish to compare their results. One of the high (such as the Forest Stewardship Council, whose C&I are
priority user requirements was to be able to compare a widely used for forest certification) are interested in being
C&I set with a new Generic Template as CIFOR's research able to incorporate these sets into CIMAT. If this can be
leads to new developments. To facilitate this, a comparison achieved the potential audience for CIMAT becomes even
submodule of CIMAT has been developed. The broader than we anticipated.
comparison involves:
1. Broad comparison of the two sets of C&I summarised 6. Conclusions and Future Work
in a numerical manner. This gives a picture of where
The construction of CIMAT has successfully shown the
gross differences in detail and elaboration occur
applicability of knowledge based systems techniques to
between the sets.
assist in delivering knowledge about Criteria and
2. Comparison of indicator objects in the two sets carried Indicators of sustainable forest management. A useful tool
out by matching of the texts of the objects. This is now available for use by forest managers to inform their
provides a complete description of all the indicator decisions and help them to address the question of
objects that the two sets have in common, plus all the sustainable management in a structured way. Crucially
differences. This is carried out principle by principle, CIMAT has been built in the expectation that forest
and involves exhaustive matching of object texts. It managers, with considerable local knowledge, can modify
lines up all those indicator objects that occur in both and adapt the C&I to suit their local circumstances, thus
sets and all those that are different in the two sets. integrating local and remote sources of knowledge. The
argumentation component of CIMAT helps to structure the
3. Analysis of the argumentation involved in generating
user's reasoning about these changes.
the two sets. This exploits CIMAT’s recording of the
Our future research will involve two main thrusts. The
user’s reasoning when making changes, and the
first new direction is the development of computer support
histories of the objects, to provide a comparison of
for applications of C&I to assess sustainability, to inform
how different users reason in different ways from each
monitoring decisions, and to structure large bodies of
other in modifying the generic template. It provides a
detailed monitoring information. This research will be
picture of which indicators have been controversial in
focussed on exploring how to support C&I use in adaptive
one instance but not in another.
forest management involving communities and other forest
stakeholders. The second new direction is the promotion of
debate and argumentation around C&I using telematics –
the WWW and email – to develop further tools for ACCESS TO RESOURCES AND ECONOMIC
structuring the work of multi-disciplinary teams of C&I BENEFITS
developers with different viewpoints, and to explore ways
P.4 CONCERNED STAKEHOLDERS HAVE
of building consensus.
ACKNOWLEDGED RIGHTS AND MEANS TO
MANAGE FORESTS COOPERATIVELY AND
Annex 1 EQUITABLY
P.5 THE HEALTH OF THE FOREST ACTORS,
High priority user requirements for CIMAT CULTURES AND THE FOREST IS ACCEPTABLE
TO ALL STAKEHOLDERS
CIMAT should:
P.6 YIELD AND QUALITY OF FOREST GOODS
• Provide users with CIFOR’s generic set (template) of AND SERVICES ARE SUSTAINABLE
PCI&V.
• Support easy navigation around its information and
around the C&I set.
Annex 3
• Support customisation (localisation) of the generic
C&I template to a local area. Examples of Criteria, Indicators, and Verifiers under P.2
• Use links between C&I to help users to reduce the
number of indicators they will use and to think about
how single indicators can address multiple issues. P.2 MAINTENANCE OF ECOSYSTEM INTEGRITY
• Enable users to undo changes they have made to the C.2.1 The processes that maintain biodiversity in managed
generic template and start again. forests (FMUs) are conserved
• Provide help screens. I.2.1.1 Landscape pattern is maintained
• Guide the user through modification of C&I using a V.2.1.1.1 FMU compiles information on areal extent
series of screens which prompt them at each step. of each vegetation type in the intervention
• Allow the user to save their changes to the C&I set. area compared to area of the vegetation
• Allow the user to output a modified C&I set to a word type in the total FMU
processor. V.2.1.1.2 Number of patches of each vegetation type
at the FMU is maintained within natural
• Allow the core generic C&I template to be updated as
variation
CIFOR does more research.
V.2.1.1.3 Largest patch size of each vegetation type
• Display the C&I as a multi-level list.
is maintained within critical limits
• Retain a history of local modifications thereby acting V.2.1.1.4 Area weighted patch size is maintained
as an 'institutional memory aid'. within critical limits
• Allow the user to print their modified C&I set as a V.2.1.1.5 Contagion index of the degree to which
table. vegetation types are aggregated, is
• Support customisation (localisation) of the generic maintained within critical limits
V.2.1.1.6 Dominance of patch structure does not
Annex 2 show significant change as compared to
unlogged site
V.2.1.1.7 Fractal dimension of patch shape is
List of Principles maintained within critical limits
P.1 POLICY, PLANNING AND INSTITUTIONAL V.2.1.1.8 Average, minimum, and maximum
FRAMEWORK ARE CONDUCIVE TO distance between two patches of the same
SUSTAINABLE FOREST MANAGEMENT cover type are maintained within natural
variation
P.2 MAINTENANCE OF ECOSYSTEM INTEGRITY
V.2.1.1.9 Percolation index, specifying landscape
P.3 FOREST MANAGEMENT MAINTAINS OR connectedness, is maintained within
ENHANCES FAIR INTERGENERATIONAL critical limits
 V.2.1.1.10 Linear measures of the total amount of Doyle, J. 1979. A Truth Maintenance System. Journal of
edge of each vegetation type exist Artificial Intelligence 12.
V.2.1.1.11 Amount of edge around the largest patch
does not show significant change as Doyle, J. 1992. Reason Maintenance and Belief Revision:
compared to undisturbed forest Foundations versus Coherence Theories. In Gardenfors, P.
(ed). Belief Revision. Cambridge University Press.
I.2.1.2 Change in diversity of habitat as a result of
human interventions are maintained within Forbus, K.D. and de Kleer, J. 1993. Building Problem
critical limits Solvers. MIT Press, London, England.
V.2.1.2.1 Vertical structure of the forest is
maintained within natural variation Haggith, M. 1996. A meta-level argumentation framework
for reasoning about disagreement. PhD Thesis. Department
V.2.1.2.2 Size class distribution does not show
of Artificial Intelligence, University of Edinburgh, Scotland.
significant change over natural variation
V.2.1.2.3 Frequency distributions of leaf size and Kraetzschmar,G.K. 1997. Distributed Reason Maintenance
shape are maintained within natural Systems. Springer Verlag, Berlin Heidelberg, Germany.
variation
V.2.1.2.4 Frequency distribution of phases of the Liang, T-Y. 1994. The basic entity model: a fundamental
forest regeneration cycle is maintained theoretical model of information and information
within critical limits processing. Information and Management 30:647-661.
V.2.1.2.5 Canopy openness in the forest understorey
is minimized Mazza, C.; Fairclough, J.; Melton, B.; De Pablo, D.;
Scheffer, A.; Stevens, R.; Jones, M. and Alvisi, G. 1996.
V.2.1.2.6 Other structural elements do not show
European Space Agency Software Engineering Guides.
significant change
Prentice Hall.
V.2.1.2.7 The distribution of above ground biomass
does not show significant change as Prabhu, R.; Colfer, C.J.P. and Dudley, R.G. 1999.
compared to undisturbed forest Guidelines for developing, testing & selecting criteria and
indicators for sustainable forest management. ). Criteria &
Indicators Toolbox Series No. 1. CIFOR, Bogor,
References
Indonesia.
CIFOR C&I Team. 1999. The CIFOR Criteria and
Indicators Generic Template. 1999. Criteria & Indicators Prabhu, R.; Colfer, C.J.P. and Shepherd, G. 1998. Criteria
Toolbox Series No. 2. CIFOR, Bogor, Indonesia. and Indicators for Sustainable Forest Management: New
Findings from CIFOR’s Forest Management Unit Level
Clocksin, W.F, and Mellish, C.S. 1981. Programming in Research in Rural Development Forestry Network.
Prolog. Springer-Verlag. Network Paper 23a. Overseas Development Institute,
London, England.
Colfer, C.J.P.; Brocklesby, M.A.; Diaw, C.; Etuge, P.;
Günter, M.; Harwell, E.; McDougall, C.; Porro, N.M.; Prabhu, R.; Yasmi, Y.; Haggith, M.; Sukadri, D. and
Porro, R.; Prabhu, R.; Salim, A.; Sardjono, M.A.; Purnomo, H. 1998. Information Mapping of Criteria and
Tchikangwa, B.; Tiani, A.M.; Wadley, R.L.; Woelfel, J. Indicators. Unpublished report, CIFOR.
and Wollenberg, E. 1999. The BAG (Basic Assessment
Guide for Human Well-Being). Criteria & Indicators Prabhu, R.; Colfer, C.J.P.; Venkateswarlu, P.; Tan, L-C.;
Toolbox Series No. 5. CIFOR, Bogor, Indonesia. Soekmadi, R. and Wollenberg, E. 1996. Testing Criteria
and Indicators for Sustainable Management of Forests:
Davenport, T.H. 1997. Information Ecology: Mastering Final Report of Phase I. CIFOR Special Publication,
the Information and Knowledge Environment. Oxford Bogor, Indonesia.
University Press.
Purnomo, H.; Rizal, A.; Prabhu, R.; Yasmi, Y.; Sukadri, D.
Doukidis, G.I.; Cornford, T. and Forster, D. 1994. and Haggith, M. 1999. CIMAT: A Knowledge-Based
Medical Expert Systems for Developing Countries: System for Developing Localized Criteria and Indicators
Evaluation in practice. Expert Systems with Applications for Sustainable Forest Management. The Third Workshop
7(2):221-233. on Electro-communication and Information”, March 1999.
 Dept. of Electrical Engineering, Institute Technology
Bandung.

SICS. 1995. SICStus Prolog User’s Manual. Kista,

Sweden: Swedish Institute of Computer Science.

Stefik, M. 1995. Introduction to Knowledge Systems. San

Francisco: Morgan Kaufmann Publishers Inc.

Stork, N.E.; Boyle, T.J.B.; Dale, V.; Eeley, H.; Finegan,

B.; Lawes, M.; Manokaran, N.; Prabhu, R. and Soberon, J.
1997. Criteria and Indicators for Assessing the
Sustainability of Forest Management: Conservation of
Biodiversity. CIFOR Working Paper no. 17. CIFOR,
Bogor, Indonesia.

Whitby, B. 1988. Artificial Intelligence: a handbook of

professionalism. Ellis Horwood.

View publication stats