A concept model for knowledge production GIS
This explains how the definition of a knowledge development GIS (KDGIS) works as a model. This piece is the set-up for a discussion of the elements of the model and their relation to systems and practice that will be followed by the critique of DIKW models you saw earlier. Being explicit with the KDGIS definition and model will make that critique more simple and specific.
FINISH THE PHILOSOPHY OF KNOWLEDGE
CONNECT THIS WITH THE CONTENT FROM THE PHILOSOPHY OF KNOWLEDGE.
REVISIT BENYON AND THE PROPERTIES OF MODELS
SKETCH THE STRUCTURED FUNCTIONS
There is no conceptual model for GIS as systems and not representations of system elements or applications. Concept models for GIScience have been proposed and applied as the discipline has developed and are relevant to system definitions, cartography, analysis and application. These apply to isolated or incremental system activity but they do not treat the composition and function of GIS as an integrated system. In addition they are not sensitive to systems’ extraorganizational situatedness. Need for a model that integrates these is especially apparent for GIS with frequent system state change. Knowledge development GIS (KDGIS) commonly experience system state change due to turnover in personnel or worldviews, short term funding agreements, and exposure to paradigm or priority shifts.
A KDGIS model will support GIS practitioners as they engage with GIS situated in ecologies where problems arise, methodologies are developed, outcomes are disseminated, and uses are defined. This model will represent the structure, operation and constraints in these complex situations for the purpose of communication, analysis, and explanation (Kangassalo 1983). It will constrain and focus discourse on implementation by limiting the range of concepts that can be expressed (Kangassalo 1983; Benyon 1997). Systems’ operational levels are the KDGIS principal concern that addresses information flows and the process of knowledge construction and representation (Benyon and Imaz 2009). This follows Codd’s (Codd 1970; Codd 1982) development of the relational data model in information systems development. The organizational and workplace components of a KDGIS can be understood in their relationships to the structure and function at the operational level.
David Benyon (Benyon 2002) argues there is no universally agreed conception or theory in Human-Computer Interaction (HCI) and presents a comparison of competing methods. He argues that there are varying degrees of utility of abstractions and models offered by the different views. The discussion here considers elements of these views and describes how they are formative to understanding, communicating, testing or predicting aspects of the KDGIS model. Some principles of models discussed are aggregation, classification, structure, function, abstraction, concepts, and physical form.
The purpose of models is to suppress unnecessary detail and two ways this can be done is with aggregation and classification. Aggregation groups related things together represented as a single object while classification groups objects as a class. These are forms of abstraction that reduce detail. Benyon (Benyon 2002) calls attention to the importance of social or world views with examples of Wittgenstein’s (Wittgenstein 1953) consideration of ‘games’ as a class and Lakoff’s (Lakoff 1987) discussion of cultural differences in classifications. The geographic information abstractions used to structure database models of worldviews are classification, generalization, aggregation and association (Nyerges 1991). These work to suppress detail in data and database development and are an important element contributing to individual systems. In a systems model these choices are further generalized and represented as contributions to articulation and development of information needs, worldview abstraction, schema, database model, measurement and analysis. Emphasis on aggregation at this granularity takes place during comparison of systems before they are unpacked to consider specific details or differences. Aggregation of system processes is part of the ‘levels of abstraction’ principle introduced later.
Structure and function are closely related where they are concerned with components of the system and how their interaction is altered with system change. A structural view places emphasis on the main entities or artifacts in the system and how they interact. This is extended to the operations interactions between specific sets of entities. A functional view is concerned with the movement and interactions of a substance as it moves through the system. Structure and function come together during observation of system dynamics as it moves from one state to another.
These aspects of structure and function tempered by abstraction and classification play a part in GIS and are of particular concern for the KDGIS model. System structure in GIS can be coarsely grouped into constitutive sociotechnical entities from development of information needs through to contributions of results as evidence and knowledge. In addition to the examples of aggregation, main entities include measurement, integration, analysis, and production. These are part of the KDGIS model and interactions between entities are characteristics of specific systems. Of greater importance in KDGIS is flow through the system and ways that activity in creation and integration of data resources as they contribute to the system as a whole. This is echoed in system dynamics that are shaped by intra and extra organizational forces.
‘Levels of abstraction’ support structure and function where one may be part of the other. A function might be well suited as a main entity in a structural view and visa versa. This clarifies relationships in KDGIS where the model takes physical form as a structure of functions. The schema (Nyerges 1989) function is made up of main entities and interactions that comprise organizational practice and negotiation with the abstraction (Nyerges 1991) and database model functions. Another level of abstraction can be seen where the schema function becomes a main element part of a structural viewpoint.
Concept is part of models and is the relationship between the object of a model and the concepts from which it is constructed. The affordances of a model rely directly on the choice of main elements and relationships and their effective treatment of the model’s functional needs. This pertains to alignment between the model purpose and appropriateness of the model’s components to achieve that purpose in a suitable and useful manner consistent with the model’s concept. Concept represented in the KDGIS model rests on the arrangement of relationships between the model’s structures of functions and the model as a whole. There must be consistency in ways these relationships are arranged to approach all instances of KDGIS in a valid manner. They must present a suitable framework to support understanding, communicating, testing or predicting aspects of KDGIS.
From email to Tim:
Categories in the diagram:
The diagram is an information flow model through as series of sociotechnical
functions. Information needs initiate the flow and it generally moves 'up'
through the successive functions toward 'knowledge'. This is arranged after an
idealized or commonly recognized associations between functions that move
information and activity through the model. While the model has physical form
in the diagram it isn't literal, some have feedback and others may appear out of
order; 'transformation' and 'integration' are two of these. There is a cyclical
pattern where knowledge or wisdom connect back to information needs.
Emmbeddedness:
Embedded in each of the model's functions are the social and technical
interactions or iterations required for the system to be successful.
The disembeddness of the model is a deliberate reaction to the numerous models
that are not sufficiently specific about how their elements or functions are
connected. To collapse this model to an embedded state might be:
real world-> measurement-> analysis-> cartography/report-> user-> real world.
Throw in a little feedback and this is essentially the cartographic process. My
model is meant to expand the cartographic process to account for the
interactions present in situated systems.
Experience:
The position of experience between 'product' and 'knowledge' is not exclusive
to the model's process. It has an embedded quality of accumulation from within
the model and drawing from outside. Experience is operable when the model
iterates or parallel processes which accumulate at 'experience'. After that...
"The only source of knowledge is experience." Albert Einstein
Development:
Development is meant to be read as the information flow among the model's
structural functions. "Knowledge development" and "Knowledge generation" are
common in the GIScience and business lit. It is more common that knowledge
development pertains to individuals' learning activities, but is also associated
with systems applications. 'Generation' is covered in Ch. 3 of Davenport and
Prusak's 'Working Knowledge'.
One of the diagrams I collected depicts knowledge development as a cycle:
It comes from
Nonaka, I. and H. Takeuchi (1995). The knowledge-creating company : how Japanese
companies create the dynamics of innovation.
My moves with a definition and model for KDGIS is deliberate to make my work
easier and easier to understand. I am proposing the ideas not as final products
but as flexible initiatives that suit my current needs. When applied the model
needs to orient and frame the approach and investigation of a GIS and then 'map'
its relationships with others.
The piece you just saw is the very beginning of my treatment of the many states
data take within and between organizations. In earlier conversations I used
'data stages' to describe the collection of element/functions represented in the
model. Trying to reconcile these, their interactions, and knowledge development
forced me to organize them as a model.
It has always been my intent to describe each of the model's elements in a
manner suitable to grasp the nature of their interaction. The model gives me
structure to organize and apply these thoughts.
Connecting KDGIS with 'Data Worlds' is easy.... they are the 'atomic element'
and subjects of my work. Individually they represent worlds and have
granularity when connected through practice, mandate and hierarchies of power.
Stages
This needs a better definition of the complex nature of Mode 2 knowledge development organizations – consult Perrow and the origins of Mode 2.
Punch line is to set up the manner of interaction between the KDGIS model’s function elements.
Data demand for scientific endeavors arise from the questions that researchers propose (Hesse, Sproull et al. 1993; Michner and Brunt 2000). Uncertainty in social setting of KDGIS is another source of uncertainty that can bring abrupt changes in priorities.
All are science related practices that contribute to uncertainty
Cooperation with other organizations
Standards ISO 2000
Compatibility with other works
Need to account for other works
How much of the science driven issues cross over into other flavors of KDGIS like R&D, legal confrontation, resource management, response to government mandates, protection of rights, or debunking the opposition? Note: all of these are volatile – wildlife is volatile and vulnerable to change. Urban growth and planning, global warming, natural disaster management – 2004 tsunami in Sri Lanka, economic collapse/growth. Response to the opposition. Development of new practices (R&D). Competition. Fast moving confrontation – epidemics and war. Resource scarcity
SITUATEDNESS
Situatedness: the occasioned properties of interactional sequences
- the contingence of action on a complex world … [is not] an extraneous problem … but ... an essential resource that makes knowledge possible and gives action its sense.
- the coherence of action is not adequately explained by either preconceived cognitive schema or institutionalised social norms.
- Rather the organisation of situated action is an emergent property of moment-by-moment interactions
- “the coherence of action is not adequately explained by either preconceived cognitive schema or institutionalized social norms”.
- The organization of situated action is an emergent property of moment-by-moment interactions between actors and the environment that they are interacting with.
- A situation is a user’s perception of a specific moment based on his/her evaluation of the interaction with an information retrieval system and his/her plan.
Context, the representational view
• context is a form of information
• context is stable
• context is uniform
• context is delineable
• context is separate from “content”
“Ongoing production of new forms of working practice”
mutually influential interactions through engagement with internal and external forces interactions that shape the system
goes beyond system design and organizational integration to addresses the situatedness of GIS design and implementation that goes beyond technical practicalities to include organizations using the system and
as these unfold in specific situations. Too often attention to system design retains focus on system design without appropriate consideration for the
USER NEEDS
Look at Huxhold ch 16 in the MPLIS: The Guidebook
Multi-purpose geographic database guidelines for local governments
Erik De Man (De Man 1988)
P324
Much attention has been placed on the technical concerns involved in design and implementation of GIS. There needs to be more attention paid to “…the environment of an envisioned GIS”. This entails examination of the localized function of the system and the “wider object system.”
P325
Sequence of functions graphic is very similar to Nyerges project flow.
See De Man 1984 – Conceptual framework and guidelines for establishing geographic information systems.
“The value of an information system arises out of the usefulness of its resultant information products. Information is an answer to a question. These questions emerge in the context of a problem-solving and often in connection with managerial activities and functions.”
P326
Information accompanies matters of importance combined with uncertainty.
P332
GENERAL OBSERVATIONS: Need recognition and system planning
P333
PLANNING: DECISION AND CHOICES: Understanding context of mission, users, needs and questions of why and what followed by how. Requires “development of sound, comprehensive and creative concepts rather than short-sighted mastery of techniques and tricks.”
PROBLEM FINDING SOLVING (DIAGNOSIS) OR SOLUTION FINDING: Dutton and Kramer (Dutton and Kramer 1978). Simon (Simon 1960) distinguishes between different techniques that are used for handling the so-called programmed and non-programmed aspects of decisionmaking. Programmed: repetitive, routine so definite procedures can be established to resolve them. Non-programmed: novel, unstructured, ‘one shot’.
P334
METHODS OF DESIGN: top-down or feedback iterative (pilot project)
UNCERTAINTIES: A dynamic and unstable environment will influence the system just as the system may effect the uncertainty and contribute stability.
STRATEGIC CHOICES: Interconnected decision areas where choices must be made and several areas may be connected with varying effect.
.
After this he returns to the systems approach by with specific attention to the analytical requirements then addressing planning steps that recognize the importance of org, tech, and design interaction.
He does not get into operation and real-time choices in operation and productivity.
Tim and Piotr
P21 Explanatory relationships rather than just description.
Testing to find if these enhance our understanding about the world.
Human-Computer-Human interactions.
How do people make use of GIS in a problem context?
Group decision making as human-computer-human interaction.
WHAT IF THIS IS REINTERPRETED AS HUMAN-COMPUTER-ACTION?
- “Convening structures” – do these apply to org-org interaction
- Power and control – These are between org and can be applied to an org in the sense that certain collective conscious coalesces orgs from inside, outside or both.
- Subject domain – this is definitely the internal shared interest, but can also bring orgs together.
- Task, purpose, complexity – this aligns with ‘matters of importance together with uncertainty (de Man 1990):326. This can also be equated with ‘Messy GIS” where many stakeholders or interested parties require or seek or order to focus collective action to achieve a future state.
- P28
- A requirement is knowledge about the natural environment – I interpret this as interpretations of reality and certainty in cause effect. There’s a difference between stakeholder interpretations and values. This poses a science vs. politics standoff. The science solution requires buy-into the same interpretations of reality. They explain this further This can fuel KPGIS
- Convenor of participants – The PSP as an example of multi-org situation. Champions, leaders, board of directors and prophets at the org. level. Discourse.
- Choice of participants – Being hired into an org or contracted to provide a service in the case of individual organizations. Multiorg might be criteria to participate or favoritism. Choice for opposition and conflict.
- Rules and norms as social structures among participants – Who is ‘allowed’ to engage with whom in a manner that maintains legitimacy and respect.
FOLLOWING THE BUILD-UP ON KNOWLEDGE
AND RECOGNITION OF OTHER MODELS IN GIS
THE POINT ON THE OTHER MODELS IS THEY ARE PROCEDURAL AND DO NOT ADDRESS THE SITUATEDNESS OF THE GIS OR THE ORGANIZATION
YET, THEY DO TRY TO ADDRESS HIERARCHY AND DEVELOPMENT OF KNOWLEDGE/WISDOM.
REMEMBER TO RETURN TO THE DIKW MODEL FOR A DEFINITION OF WISDOM
The KPGIS model is composed of interrelated functions that represent the behavioral and organizational operations of GIS used for problem solving and decision making. It adheres to the facts that GIS are embedded in social and cultural goals (Chrisman 1987) and GIS need to be designed relative to their actual use (de Man 1990). KPGIS are situated activity (Suchman 1987; Suchman 1995) made of moment-by-moment interactions between actors and their environment. In these ways, the model is an approach that considers GIS as a localized enterprise. For this reason the model is not hierarchical because it represents a complex system with loose couplings (Perrow 1999). The hierarchical and process steps of GIS design and implementation are flattened in KPGIS. Despite this, adjacency of functions in the graphic model and sequential order of the model’s description represents common associations experienced in practice and not exclusivity. See graphic ??
The focus of the model is on the in-situ and lived-in use:
Descriptions of these are here:
The attention to systems’ details and organizational preparedness ignores expected lived-in use or current and future development of practices, especially ones brought on by interaction with GIS (Martin 2000). These sides of organizations or enterprises are part of situated learning (Lave 1993; Wenger 1998), articulation work (Suchman 1995) and computerization movements (Kling and Iacono 1995) and middle spaces (Fauconnier and Turner 1994).