Virtual Environments for Geospatial Applications Figure a) 3D virtual environment depicting geospatial processes (1 picture of a series) such as landscape change over time etc.; b) 3D virtual environment depicting water flow in a reservoir Critical Issues in the Design and Implementation of Geospatial Virtual E nvironments This section concisely discusses some limitations and constraints typically experienced in several virtual world generations as well One noteworthy issue is that in visualizing real-world scenarios, there is an inevitable trade-off amid performance and resolution Exploiting the complete capabilities of virtual environments over the Web continues to pose problems As the number of objects in a virtual environment increases, online hosting becomes an issue as spontaneous rendering of numerous objects is no easy task Scenes with a greater number of polygons decelerate the system and make the interactivity poor Several factors need to be considered during visualization such as the type and volume of data to be visualized, memory constraints, and system performance Table presents a summary of the significant issues concerning geo-virtual environments In their work on information visualization, Robertson et al (1993) have presented a terse compilation of the important issues In his work on dynamic and interactive webbased visualizations, Huang and Lin (1999, 2001, and 2002) discuss in detail some of these concerns and also address some critical issues concerning online hosting of interactive visualizations The Java-3D based hybrid method that Huang and Lin (1999, 2001) propose offers a standard framework Table A summary of critical issues in designing and implementing 3D virtual worlds Photo-realistic scene generations Generating 3D virtual e nvironments adequate photo-realism w ith User Navigation capabilities Users n eed skills to n avigate and situate themselves within immersive virtual worlds Bandwidth Limitations 3D Scenes with n umerous objects, rendering difficulty, and transmission speed Browser and Plug-in Compatibility Compatibility a mong v arious b rowsers as well as plug-ins Data Integrity and online security issues Sensitive data must be p rotected, and the data represented by such 3D worlds should be up-to-date Online hosting Server load must be balanced to handle multiple simultaneous requests 328 Lag in real-time interaction Complex Scenes n ot o nly take t ime to r ender, b ut also cause delays/lags during navigation/interaction Spatio-temporal representations The issue of representing spatio-temporal aspects as dimensions within virtual worlds must be resolved Virtual Environments for Geospatial Applications Figure Online hosting of interactive visualization (From Huang et al.,1999) for visualizing dynamic environmental processes Figure illustrates a 3-tier configuration that Huang and Lin (1999) proposed in GeoVR The visualization server that is interlinked to the spatial database accesses the geospatial information from the data repository and the web server accesses the visualization server for 3D information This framework efficiently handles requests for visualizing dynamic processes and based on the client requests, the web server provides the appropriate information in the conventional HTML or 3D VRML format Di scuss ion and C onc lus ion Over the past several decades, information presentation has inspired the development of several new tools and techniques The information revolution has resulted in vast amounts of data that are far too complex, both in quality and quantity, to be handled by conventional tools and techniques Recent technological advances in the realm of remote sensing have dramatically increased the amount of geospatial data available Virtual environments are an efficient means of visualizing voluminous geospatial data and are efficient in elucidating the intricate patterns as well as hidden and associated information Such virtual environments facilitate understanding of the complex relationships among the various components of a multi-level scenario This paper discussed the design and implementation of virtual worlds that can be used to generate both static representations depicting real-world settings and dynamic representations that can simulate geospatial processes and environmental phenomena The paper discussed the generation of such geo-virtual environments with examples and provided explanations as to how such geo-visualization applications facilitate understanding of various geospatial phenomena and environmental processes The fundamental principles underlying the generation of virtual worlds, both static and dynamic, were elaborated and the common issues involved in the generation of such 3D virtual worlds were discussed Furthermore, the issues related to the online hosting of such virtual environments were tersely delineated and possible solutions to frequently encountered problems were provided 329 Virtual Environments for Geospatial Applications R eferences Ames, L A., Nadeau, D R., & Moreland, J L (1996) VRML 2.0 Sourcebook Bonham-Carter, G F (1994) Geographic Information Systems for Geoscientists: Modeling with GIS Pergemon: Oxford (p 398) Boyd, D S., Lansdown, J., & Huxor, A (1996) The Design of Virtual Environments SIMA Chandramouli, M., Lan-Kun, C., Tien-Yin, C., &vChing-Yi, K (2004) Design and Implementation of Virtual Environments for Visualizing 3D Geospatial Data TGIS Conference, Oct 28-29 2004 Chandramouli, M., Huang, B., Yin Chou, T., Kun Chung, L., & Wu, Q. (2006) Design and Implementation of Virtual Environments for planning and Building Sustainable Railway Transit Systems, COMPRAIL July 2006, Prague Colin, W (2000) Information Visualization: Perception for Design Morgan Kaufmann Series in Interactive Technologies GeoVRML, (www geovrml.org) Huang, B., & Lin, H (1999) GeoVR: A Webbased tool for virtual reality presentation from 2D GIS data Computers & Geosciences, 25(10), 1167-1175 Huang, B., Jiang, B., & Lin, H (2001) An integration of GIS, virtual reality and the Internet for spatial data exploration International Journal of GIS, 15(5): 439-456 Huang, B., & Lin, H (2002) A Java/CGI approach to developing a geographic virtual reality toolkit on the Internet Computers & Geosciences, 28(1), 13-19 Karel, C., & Jiri, Z (n/d) Using VRML for creating interactive demonstrations of physical models Department of computer science and Engineering Czech Technical University 330 Robertson, G., Card, S., & Mackinlay, J D (1993) Information Visualization Using 3D Interactive Animation Communications of the ACM, 36, 57-71 Shiode, N (2001) 3D urban models: recent developments in the digital modeling of urban environments in three-dimensions GeoJournal, 52(3), 263-269 SGILICGF, UW Sea Grant Institute and Land Information and Computer Graphics Facility http://coastal.lic.wisc.edu/visualization/Visualization.htm Sutherland, I E (1965) The ultimate display In the proceedings of the IFIPS Congress, 2, 506508 New York City, NY key T er ms Immersion: A Sense of being present within the virtual world and a ‘sense’ being able to visualize objects by being amidst their surroundings and navigating through the world Node: An entity within the hierarchical scene structure that represents a group of objects OpenSource: Source code or computer software that is freely offered and is available to the public for building software applications Scene-Hierarchy: The organization of the elements of a 3D virtual scene into successive levels, in such a way that the object under which other objects are grouped is called the parent and the grouped objects are called its children When a parent object is transformed, the children are also transformed SCRIPT: Program scripts that are used to perform calculations and return values to the calling programs Transformation: Operations such as translation, rotation, or scaling involving objects in a virtual environment Virtual Environments for Geospatial Applications Virtual Reality: A three-dimensional visual immersive setting that facilitates user to navigate within the scene and perform operations in real time 331 332 Chapter XLI Managing Uncertainty in Geospatial Predictive Models Iftikhar U Sikder Cleveland State University, USA Abstr act Geospatial predictive models often require mapping of predefined concepts or categories with various conditioning factors in a given space This chapter discusses various aspects of uncertainty in predictive modeling by characterizing different typologies of classification uncertainty It argues that understanding uncertainty semantics is a perquisite for efficient handling and management of predictive models S pat ial Pred ict ion and Cla ss if ic at ion Geospatial predictive models entail an array of analytical techniques of data mining, classical statistical and geostatistical models that attempt to predict spatial states and behavior of objects from a fine set of observations The process of prediction presupposes a set of spatial concepts and categories to which objects are to be mapped For example, spatial processes, such as classification of land cover from satellite image, modeling for- est fire, propagation of epidemics, and prediction of urban sprawl require a unifying and common reference of “space” or location where the multiple features of spatial attributes are to be mapped to predefined class labels The prediction of spatial features can be conceived as a process of driving classification schemes in relation to certain spatial properties such as neighborhood, proximity, dependency, as well as similarity of non-spatial attributes (Han & Kamber, 2006; Shekhar & Chawla, 2003) In data mining, a classification function is often defined as a mapping function: Copyright © 2009, IGI Global, distributing in print or electronic forms without written permission of IGI Global is prohibited Managing Uncertainty in Geospatial Predictive Models f : A → C , where A is the domain of function, f represents attribute space and C is the set of class categories U ncert ainty in S pat ial Cla ss if ic at ion Uncertainty may emerge from ontological constraints in classification i.e., from the lack of specification of what kind of spatial objects exist, as well as from epistemic limitations which concern whether such objects are knowable to subjective schemes, and if so, to what extent they can be represented in the subjective framework, given the limited empirical evidences Epistemic uncertainty in spatial classification emerges due to inadequate representation of spatial knowledge which is often incomplete, imprecise, fragmentary, and ambiguous The attributes of spatial objects or evidences suggesting various conceptual or thematic classes may often suggest conflicting categories Moreover, classification labels are dependent on the resolution of observation and the extent of granularity For example, the observation of coarser granularity offers less detail while the clumping of information into pixels in remotely sensed images may prevent sub-pixel entities being distinguished (Fisher, 1997) The classification of land cover from satellite image depends not only on a specific spatial resolution, radiometric resolution and the corresponding spectral signatures limit predictive accuracy Therefore, spatial characteristics of a given observation are indiscernible with respect to attributes associated with it For example, the number of vegetation types that can be identified from an NDVI (Normalized Difference Vegetation Index) image significantly increases when a very high radiometric resolution is used Moreover, in a specific case, a multispectral image may provide more accuracy than a hyperspectral image, but such accuracy is of little value if it is achieved at the cost of less specificity or higher imprecision T ypo log ies of Cla ss if ic at ion U ncert ainty While there is increasing awareness of uncertainty, and its aspects and dimensions in predictive as well as classificatory schemes, little agreement exists among experts on how to characterize them Many typologies of uncertainty have been suggested from risk analysis perspective, which often overlaps and builds on each other (Ferson & R. Ginzburg, 1996; Linkov & Burmistrov 2003; Regan et al., 2002) These typologies make distinctions between variability and lack of knowledge at the parameter and model level However, from the geographic information perspective, the ontological specification of imperfection of geographic data provides some key vocabularies and taxonomies to deal with spatial uncertainties (Duckham et al., 2001; Worboys & Clementini, 2001) Such ontology distinguishes between inaccuracy (i.e., errors or commission or omission) and imprecision, which arises from limitations on the granularity of the schema or levels of detail obtainable for an observation under which the observation is made (Worboys, 1998) The concept “vagueness” refers to indeterminate boundary-line cases or “inexact concepts” Classification of geographic objects with indeterminate boundaries offers many challenges (Burrough & Frank, 1996) which emerge from the boundary of many real entities representing natural, social, or cultural phenomena (for example, forests, mountains, areas ethnic distribution etc.) Since many common geographical concepts are vague (Fisher, 2000), the explicit specification of vagueness is essential to characterize the classification performance As a special type of vagueness, nonspecifity originates due to our inability to discern the true alternatives among several alternatives in a given context It implies 333 Managing Uncertainty in Geospatial Predictive Models cardinality of undiscerned alternatives (Klir & Yuan, 1995) The larger the set of alternatives, the higher is the nonspecifity For example, in a remotely sensed image, a pixel with class type “forest” and the mean annual temperature > 30C has less nonspecifity than the pixel labeled only with “forest” type This is because in the latter case a pixel can have a large number of possible variations of “forest” type Broadly, three major categories of uncertainty can be identified in dealing with predictive and classificatory problems: ontological uncertainty, epistemological uncertainty, and deontological or normative uncertainty The typology illustrated in Figure is relevant to mainly geospatial data and includes many important components and concept provided in Morgan & Henrion (1998), Finkel (1990), and Cullen & Frey (1999) and Haimes (2004) The types presented here is in by no means mutually exclusive, i.e., some concepts may subtly overlap each other in a specific context Ontologically, variability, also known as aleatory or objective uncertainty, occurs when the object that needs to be classified actually exhibits multiplicity across space, time and scale An empirical quantity measured in a single point may objectively manifest multiple aspects in a collective process For example, land cover classes are not only influenced by seasonal and spatial extent, but also the topographic formation due to self-similar features of geological objects requires specificity of fractal dimension of a classification scheme The spurious correlation representing the so called of ecological fallacy resulting from modifiable areal unit problem (MAUP) (Openshaw, 1984) indicates the requirement of adequate disaggregation in spatial data to be analyzed In image processing, uncertainty often arises due to the assignment of more than one class to a pixel This specialized type of pixel, often known as mixel, indicates uncertainty resulting from variability Similarly, variability or the degree of spatial heterogeneity Figure Types of uncertainty in dealing with geospatial predictive and classificatory systems Types of Uncertainty Variability • heterogeneity • Process uncertainty • Self-similarity Deontological/ Epistemological Ontological • Metric error • Sampling error • Commission/omission error • • Monte Carlo • • Choice of models • • Imprecise Probability- Interval valued probability, general upper & lower probability, belief / plausibility, Rough, Fuzzy, hybrid , Evidence reasoning) • 334 Decision / Value Model Parameter -Spohn calculus and kappa-calculus • M • , value measures • Range of risk tolerance • Managing Uncertainty in Geospatial Predictive Models is also reflected in the measures of fragmentation of a landscape The uncertainty stemming from variability can not be handled by a reductionist approach, but needs to be managed by a process of disaggregation of data Measures often used to manage this kind of uncertainty are: estimating space-time frequency distribution, disaggregation by pixel unmixing or decoupling, estimating entropy as indicator of fragmentation, computing self-similarity and fractal dimension (Kallimanis et al., 2002), and multiscale and multiresolution analysis using wavelet (Kolaczyk et al.,2005; Nychka et al., 2001) While the origin of uncertainty due to variability is objective and ontological in nature, parameter uncertainty and model uncertainty reflect the epistemic state or lack of knowledge in a classificatory scheme Parameter is an empirical quantity that is measurable in principle, and is part of the system components or construct of a definition Parameter uncertainty is mainly due to the result of measurement error and sampling error For example, the misclassification rate of land cover classification, measured by the so called error of commission or omission is as good as the choice of sampling scheme, the systematic bias introduced by the selection of space-time boundary conditions, level of precision, and other parameters internal to the system Moreover, the selection of parameters may depend on the degree of variability A high degree of spatial heterogeneity requires an intensive sampling scheme across multiple scales Quantitatively, parameter uncertainty can be modeled by using probability distribution based on statistical variance of observed error e.g., Gaussian distribution can be used to predict the relative abundances of different magnitudes of error or perform Monte Carlo simulation to estimate the effect of error on a digital elevation model (Heuvelink, 1998; Longley et al., 2001) Model uncertainty, or sometimes called informative uncertainty (van Asselt, 1999) is due to limitation in the ability to represent or model real-world processes with the given data Although both parameter uncertainty and model uncertainty represent the epistemic or subjective aspect of the state of our knowledge, the line between these two types of uncertainties can not be sharply divided, because the choices of the model form have implications for a parameter, and the parameter itself can be the output of complex models (Krupnick et al., 2006) Many schemes have been developed to formalize the uncertainty due to limitation of models The probabilistic intolerance to imprecision of classical probability theory has led to many alternative formations of uncertainty models For example, traditional classification models such as multi-source classification (Lee & Swain, 1987) or the so-called maximum likelihood classification (Tso & Mather, 2001) allows no room for expressing modeler’s ignorance in the model construct This has led to new model constructs such as, interval representation in DempsterShafer’s evidence theory (Shafer, 1976) where the numbers of all possible subsets of the frame of discernment are candidate classes of belief function The belief is extracted from the sum of the probability of all the attributes that an object has, and the plausibility is the sum of the probabilities of all the attributes that the object does not have The uncommitted belief is assigned to the frame of discernment, thus allowing representation of modeler’s ignorance The evidential reasoning approach has been adopted for multi-source remotely sensed images (Lee & Swain, 1987; Srinivasan & Richards, 1990; Wilkinson & Megier, 1990) Rough set theory (Pawlak, 1992), a variant from multivalued logic is recently being used to model vagueness and imprecision by using an upper and a lower approximation Ahlqvist et al (2000) used a rough set-based classification and accuracy assessment method for constructing rough confusion matrix In the integration model of rough set theory and evidence theory, the “belief” is extracted from the lower approximation of a set and the “plausibility” from the upper approximation (Skowron & Grzymalla-Busse, 1994) In spatial 335 Managing Uncertainty in Geospatial Predictive Models prediction, this approach was further extended by introducing evidences from spatial neighborhood contexts (Sikder & Gangapadhayay, 2007) Using rough–fuzzy hybridization and cognitive theory of conceptual spaces a parameterized representation of classes are modeled as a collection of roughfuzzy property where an attribute itself can be treated as a special case of a concept In spatial classification, the fuzzy approach is mainly used to provide a flexible way to represent categorical continua (Foody, 1995) In this approach instead of explicitly defining concept hierarchies, different conceptual structures emerge through measures of concept inclusion and similarity, and fuzzy categorical data is presented in terms of fuzzy membership (Cross & Firat, 2000; Robinson, 2003; Yazici & Akkaya, 2000) Deontological or normative uncertainty is associated with consequentiality paradigm of decision or value judgments, e.g., in multicriteria classification, risk perception, preference elicitation There has been extensive research from behavioral decision theoretic perspective to understand human judgment under uncertainty (Tversky et al., 1974) The heuristics that decision makers use (Kahneman et al., 1982) can lead to biases in many spatial decision making scenarios, such as watershed prioritization, location or facility planning, habitat suitability modeling Uncertainty may also spring from conflicting value-laden terms or preference-ordered criteria (Li et al., 2004; 2005) It could be possible that preference order induced from a set of attributes may contradict the assignment of the degree of risk classes, resulting in potential paradoxical inference Pöyhönen & Hämäläinen (2001) showed that the use of weights based on the rank order of attributes can only easily lead to biases when the structure of a value tree is changed While it is difficult to extract complete preferential information, research is going on to work with information-gap uncertainty in preferences by using graph model for conflict resolution (BenHaim & Hipel, 2002) 336 C onc lus ion Uncertainty in spatial predictive and classificatory system is an endemic and multi-faceted aspect Recognition and agreement of appropriate characterization and definition of typologies of uncertainty semantics are prerequisite to efficient handling and management This article characterizes the objective, subjective and normative aspect of uncertainty It specifically differentiates uncertainty resulting from lack of knowledge and objective variability or intrinsic properties of spatial systems Various new directions of uncertainty handling mechanism are discussed While currently there are many promising directions of research in managing different types of uncertainty, a new paradigm is required in spatial analysis that is fundamentally driven by the consideration of uncertainty R eferences Ahlqvist, O., Keukelaar, J., & Oukbir, K (2000) Rough classification and accuracy assessment International Journal of Geographical Information Science, 14(5), 475–496 Ben-Haim, Y., & Hipel, K W (2002) The Graph Model for Conflict Resolution with InformationGap Uncertainty in Preferences Applied Mathematics and Computation, 126, 319-340 Burrough, P., & Frank, A (1996) Geographic Objects with Indeterminate Boundaries London: Taylor and Francis Cross, V., & Firat, A (2000) Fuzzy objects for geographical information systems Fuzzy Sets and Systems, 113, 19–36 Cullen, A C.,& C Frey, H (1999) Probabilistic Techniques in Exposure Assessment: A Handbook for Dealing with Variability and Uncertainty in Models and Inputs New York, NY: Plenum Press Managing Uncertainty in Geospatial Predictive Models Duckham, M., Mason, K., Stell, J., & Worboys, M F (2001) A formal approach to imperfection in geographic information Computers, Environment and Urban Systems, 25, 89-103 Kolaczyk, E D., Ju, J., & Gopal, S (2005) Multiscale, Multigranular Statistical Image Segmentation Journal of the American Statistical Association, 100(472), 1358-1369(1312) Ferson, S., & R Ginzburg , L (1996) Different Methods Are Needed to Propagate Ignorance and Variability Reliability Engineering and Systems Safety, 54, 133-144 Krupnick, A., Morgenstern, R., Batz, M., Nelson, P., Burtraw, D., Shih, J., et al (2006) Not a Sure Thing: Making Regulatory Choices under Uncertainty: U.S EPA Finkel, A M (1990) Confronting Uncertainty in Risk Management: A Guide for Decision Makers Washington, DC: Resources for the Future, Center for Risk Management Lee, R T., & Swain, P H (1987) Probabilistic and evidential approach for multisource data analysis IEEE Transactions on Geoscience and Remote Sensing, 25, 283-293 Fisher, P (1997) The pixel: a snare and a delusion International journal of Remote Sensing, 18(3), 679-685 Li, K W., Hipel, K W., Kilgour, D M., & Noakes, D (2005) Integrating Uncertain Preferences into Status Quo Analysis with Applications to an Environmental Conflict Group Decision and Negotiation, 14(6), 461-479 Fisher, P F (2000) Sorites paradox and vague geographies Fuzzy Sets and Systems, 113, 7-18 Foody, G M (1995) Land cover classification by an artificial neural network with ancillary information International Journal of Geographical Information Systems, 9(5), 527-542 Haimes,Y Y (2004) Risk Modeling, Assessment, and Management Hoboken, NJ: Wiley Han, J., & Kamber, M (2006) Data Mining Concepts and Techniques Boston: Morgan Kaufmann Heuvelink, G (1998) Error Propagation in Environmental Modeling with GIS London: Taylor and Francis Kahneman, D., Slovic, P., & Tversky, A (1982) Judgment under Uncertainty: Heuristics and Biases Cambridge: Cambridge University Press Kallimanis, A S., Sgardelis, S P., & Halley, J M (2002) Accuracy of fractal dimension estimates for small samples of ecological distributions Landscape Ecology, 17(3), 281-297 Klir, G., & Yuan, B (1995) Fuzzy Sets and Fuzzy Logic: Theory and Applications: Pearson Education Li, K W., Hipel, K W., Kilgour, D M., & Fang, L (2004) Preference Uncertainty in the Graph Model for Conflict Resolution IEEE Transaction on Systems, Man, and Cybernetics- Part A: Systems and Humans, 34(4 July) Linkov, I., & Burmistrov, D (2003) Model Uncertainty and Choices Made by Modelers: Lessons Learned from the International Atomic Energy Agency Model Intercomparisons Risk Analysis 23(6), 1297–1308 Longley, P A., Goodchild, M F., Maguire, D J., & Rhind, D W (2001) Geographic Information Systems and Science Chichester, UK: John Wiley & Sons Morgan, M G (1998) Uncertainty Analysis in Risk Assessment Human and Ecological Risk Assessment 4(1), 25–39 Nychka, D., Wikle, C., & Royle, J A (2001) Multiresolution models for nonstationary spatial covariance functions Statistical Modelling, 2(4), 315-331 337 GeoCache Figure 12 shows the hit ratio performances realized with a random dataset queries Here we can see that in general, GDSF and B&BGDSF, outperform LRU and LFU When the cache size becomes larger, GDSF roughly scale like B&BGDSF But with smaller cache size, our B&BGDSF algorithm performs better than the other cache replacement policies This is because the Branch and Bound algorithm associated to GDSF only evicts the most irrelevant cache results with smallest cost and which size sum is greater but closest to the new query, which must be put in cache Thus, the cache is better managed For example with a cache size equal to 20 Mega Bytes, B&BGDSF performs about 8% better than GDSF Figure 13 shows the cache coverage ratio in percentage Cache coverage hit shows the effect of partial matching in semantic caching This information is not readable in traditional cache hit ratio It is defined as follows (Lee & Chu, 1998): Given a query set consisting of I queries q1, …qI, let Ai be the number of answers found in the cache for the query qi, and let Bi be the total number of answers for the query qi for 1≤ i ≤I ∑ CCR = I i =1 QCRi I where QCRi = Ai if Bi > Bi Here again, we can notice that in general GDSF and B&BGDSF outperform LRU and LFU But it is interesting to see that compared to Figure 12, when the number of hit ratio is equal for GDSF and B&BGDSF with a cache size of 50 MB or 60MB, the number of cache coverage ratio is better using the B&BGDSF algorithm Figure 16 shows the performance gain which is defined as follows (Brinkhoff et al., 2002): gain = disk _ acces _ for _ a _ given _ Policy −1 disk _ access _ of _ B & BGDSF where Policy stands for one of the three cache replacement policies: GDSF, LRU, and LFU The performance gain shows the reduction in the number of disk access using the proposed cache replacement policy, compared to others In general, we can notice that the gain is positive using our proposed algorithm However, depending on the cache size, the gain with GDSF is more or less important, especially when the cache size becomes small or large For example in Figure 12 with a cache size equal to 20 or 100 MB, the performances of GDSF are the same or less than B&BGDSF However, if we look at the gain distribution, the values obtained are about or 18 % Especially with a cache size of 100 MB, the hit ratio for GDSF and B&BGDSF are roughly the same, but the gain obtained is different about 18% These results show that the cache is better managed using B&BGDSF For a cache size equal to 30 Mega Bytes, we can see that the gain compared to GDSF is roughly equal to 23% But for a cache size between 40 and 90 MB, the gain compared to GDSF is roughly the same or less than our proposed algorithm But the Hit Ratio and Cache Coverage Ratio show that better cache management is obtained using the B&BGDSF algorithm Moreover, for a cache size of 40 or 60 MB in Figure 12, LRU performs like B&BGDSF; but in Figure 16 the gain obtained compared to LRU is the same or better (for example with a cache size of 40 MB) For a cache size of 50 MB peaks of gain, around 37% with LRU and 25% with LFU are obtained Figures 14, 15, and 17 show other experimentations performed with another query set where the most recent queries are repeated In this case, we notice that LRU outperforms GDSF and LFU, but B&BGDSF outperforms GDSF, LFU, and LRU, especially when the cache size is smaller (from 10 to 50 MB) This is due to a better selection of cache queries to evict These results show that the cache is better managed using the B&BGDSF compared to GDSF, LFU and LRU cache replacement policies The cache hit ratio (Figures 12 and 14) show that B&BGDSF 365 GeoCache outperforms GDSF, LRU, and LFU; but when the performances of B&BGDSF are roughly equal to another cache replacement algorithm, the cache coverage ratio (Figures 13 and 15) show that a better cache management is obtained using the proposed algorithm This is confirmed by the performance gain in disk access presented in Figures 16 and 17 C onc lus ion In order to optimize the I/O-performance of a database system, many techniques have been considered One of the most used in spatial database is the buffering technique associate to the LRU strategy and its variants Related works generally focus on spatial data stored in object-relational databases With the popularity of Internet and XML, the OGC proposed the GML semi-structured language, which is an extension of XML for the modeling, transport, and storage of geographical information including both spatial and non-spatial fragments of geographical data The GML language is more and more used in many distributed architectures In order to handle efficiently geographical queries stored in GML documents, we have proposed an appropriate cache structure and cache replacement policy The proposed cache structure aims at considerably reduce the memory used to store geographical queries Only non redundant spatial fragments of a geographical query are stored Thus, much more queries can be stored in cache This architecture is associated to a semantic cache where inference rules between spatial predicates are exploited for a better cache management Associated with this semantic cache, a new cache replacement policy is proposed It takes into account the cost implied by the different predicates and gives better performances than existing cache replacement policies like GDSF, LRU, and LFU The proposed cache replacement policy is based on Branch and Bound and GDSF algorithms It only evicts the 366 most irrelevant cache results according to the cost and size of data stored in cache The hit ratio, cache coverage ratio and performance gains show that the proposed B&BGDSF replacement policy outperforms GDSF, LRU and LFU algorithms In perspective, this work will be extended to spatiotemporal data stored in semi-structured documents in order to optimize spatiotemporal queries R eferences Aboulnaga, A., & Naughton, J (2000) Accurate estimation of the cost of spatial selections IEEE International Conference on Data Engineering (ICDE ’00) (pp 123-134), San Diego, CA Abrams, M., Standbridge, C R., Adbulla, G., Williams, S., & Fox, E A (1995) Caching proxies: Limitations and potentials WWW-4, Boston Conference Arlitt, M., Friedrich, R., Cherkasova, L., Dilley, J., & Jin, T (1999) Evaluating content management techniques for Web proxy caches In Proceedings of the Workshop on Internet Server Performance (WISP), Atlanta, GA Beckmann, N., Kriegel, H P., Schneider, R., & Seeger, B (1990) An efficient and robust access method for points and rectangles In Proceeding ACM SIGMOD International Conference on Management of Data (pp 322-331), Atlantic City, NJ Boucelma, O., Messid, M., & Lacroix, Z (2002) A WFS-based mediation system for GIS interoperability The 10th ACM International Symposium on Advances in Geographical Information Systems (ACM GIS), McLean, VA Brinkhoff, T (2002) A robust and self-tuning page-replacement strategy for spatial database systems The 8th International Conference on Extending Database Technology (EDBT 2002) GeoCache (pp 533-552), Prague, Czech Republic, 2002 LNCS, Vol 2287, Springer-Verlag 1737 (available from http://www.npaci.edu/DICE/ Pubs/) Brinkhoff, T., Horn, H., Kriegel, H P., & Schneider, R (1993) A storage and access architecture for efficient query processing in spatial databases In Proceedings of the 3rd International Symposium on Large Spatial Databases (pp 357-376), Singapore LNCS, Vol 692, Springer Kellerer, H., Pferschy, U., & Pisinger, D (2004) Knapsack problems Springer Cao, P., & Irani, S (1997) Cost-aware WWW proxy caching algorithms Proceedings of USENIX Symposium on Internet Technologies and Systems (USITS) (pp 193-206), Monterey, CA, December Chen, L., Wang, S., & Rundensteiner, E A (2004) Replacement strategies for XQuery caching systems Data and Knowledge Engineering, 49(2), 145-175 Chidlovskii, B., Roncancio, C., & Schneider, M L (1999) Semantic CacheMechanism for heterogeneous Web querying In Proceedings of the 8th World-Wide WebConference (WWW8) Chou, H., & DeWitt, D (1985) An evaluation of buffer management strategies for relational database systems.In Proceedings of the 11th VLDB Conference Corocoles, J E., & Gonzalez, P (2003) Querying spatial resources An approach to the semantic geospatial Web CAISE 2003 Workshop, Web, e-Business, and the Semantic Web (WES: Foundations, Models, Architecture, Engineering and Applications, LNCS, Springer-Verlag Dar, S., Franklin, M J., Jonsson, B T., Srivastava, D., & Tan, M (1996) Semantic data caching and replacement Proceedings of the 22nd VLBD Conference, Bombay, India Gupta, A., Marciano, R., Zaslavsky, I., & Baru, C (1999) Integrating GIS and imagery through XML-based information mediation In P Agouris & A Stefanidis (Eds.), Integrated spatial databases: Digital images and GIS, LNCS, Vol Leclercq, E., Djamal, B., & Yétongnon, K (1999) ISIS: A semantic mediation model and an agent-based architecture for GIS interoperability Proceedings of the International Database Engineering and Applications Symposium IDEAS (pp 81-92), Montreal, IEEE Computer Society Lee, D., & Chu, W (1998) A semantic caching schema for wrappers in Web databases The 1st ACM International Workshop on Web Information and Data Management (WIDM), Washington DC Lee, D., & Chu, W (1999) A semantic caching via query matching for Web sources In Proceedings of the 8th ACM International Conference on Information and Knowledge Management (CIKM), Kansas City, Missouri, USA Lorenzetti, P., & Rizzo, L (1996) Replacement policies for a proxy cache Technical report, Universita di Pisa Luo, O., Naughton, J F., Krishnamurthy, R., Cao, P., Li, Y (2000) Active query caching for database {Web} servers ACM SIGMOD Workshop on the Web and Databases, WebDB Meier, W (2002) eXist: An open source native XML database In A Chaudri, M Jeckle, E Rahm, R Unland (Eds.), Web, Web-Services, and Database Systems NODe 2002 Web- and Database-Related Workshops Springer LNCS Series, 2593 Erfurt, Germany O’Neil, E J., O’Neil, P E., & Weikum, G (1993) The LRU-K Page replacement algorithm for database disk buffering In SIGMOD (pp 297-306) Open GIS Consortium (1999) Inc., OpenGIS Simple Features Specification for SQL Revision 1.1 OpenGIS Project Document 99-049 367 GeoCache Open GIS Consortium (2003) Inc., OpenGIS® Geography Markup Language (GML) Implementation Specification, version 3.00, Document 02-023r4 Pagel, B., Six, H., Toben, H., Widmayer, P (1993) Towards an Analysis of Range Query Performance in Spatial Data Structures PODS 214-221 Savary, L., & Zeitouni, K (2003) Spatial data warehouse—A prototype In The 2nd EGOV International Conference (pp 335-340), Prague LNCS 2739 Savary, L., & Zeitouni, K (2005) Automated linear geometric conflation for spatial data warehouse integration process The 8th AGILE Conference on GIScience Estirl, Portugal Silvano, M., & Paolo, T (1990) Knapsack problems: Algorithms and computer implementations John Wiley & Sons Sindoni, G., Tininini, L., Ambrosetti, A., Bedeschi, C., De Francisci, S., Gargano, O., Molinaro, R., Paolucci, M., Patteri, P., Ticca, P (2001) SIT-IN: A real-life spatio-temporal information system VLDB 2001 (pp 711-712) Stoimenov, L., Djordjevic-Kajan, S., & Stojanovic, D (2000) Integration of GIS data sources over the Internet using mediator and wrapper technology MELECON 2000, 10th Mediterranean Electrotechnical Conference (Vol 1, 334-336) Tanenbaum, A S (1992) Modern operating systems Prentice-Hall Theodoridis, Y., & Sellis, T (1996) A model for the prediction of r-tree performance In Proceedings of the 15th ACM Symp Principles of Database Systems (pp 161-171) Vatsavai, R R (2002) GML-QL: A spatial query language specification for GML.UCGIS Summer, Athens, Georgia Voisard, A., & Juergens, M (1999) Geographical information extraction: Querying or quarrying? In M Goodchild, M Egenhofer, R Fegeas, C Kottman (Eds.), Interoperating geographical information systems, New York: Kluwer Academic Publishers Yang, O., Zhang, H., & Zhang, H (2003) Taylor series prediction: A cache replacement policy based on second-order trend analysis The 34th Annual Hawaii International Conference on System Sciences, Maui, Hawaii Zhang, M., Javed, S., Shaheen, A., & Gruenwald, L (2001) A prototype for wrapping and visualizing geo-referenced data in distributed environments using the XML technology ACMGIS at VA (pp 27-32) This work was previously published in International Journal of Data Warehousing and Mining, Vol 3, Issue 1, edited by D Taniar, pp 67-88, copyright 2007 by IGI Publishing (an imprint of IGI Global) 368 369 Chapter XLIV Cognitive Mapping and GIS for Community-Based Resource Identification Lyn Kathlene Colorado State University, USA Abstr act This chapter describes and analyzes the effectiveness of two methodological techniques, cognitive mapping and geographical information systems (GIS), for identifying social service resources It also examines the processes used to integrate hand-drawn map information into geocoded data points and provides recommendations for improving efficiency and precision As a first step to integrate Jefferson County social service delivery into community-based child welfare “systems of care” (SOC), both formal and informal services had to be identified Cognitive mapping, a process by which participants draw visual representations of geographical areas, was conducted with 247 participants in Jefferson County, Colorado Over 3,500 resources were identified and entered into a GIS to analyze the availability, capacity, and distribution of social services in the county and within communities Identification of community resources via cognitive mapping and GIS analysis provide: (1) a comprehensive database of existing services; (2) a basis to build communication networks and cooperation among government and community providers; (3) the ability to create an efficient system that avoids duplication of efforts; (4) an understanding of the geographical distribution of resources; (5) the identification of resources lacking in the county and specific communities; and (6) knowledge differences among diverse participant groups Copyright © 2009, IGI Global, distributing in print or electronic forms without written permission of IGI Global is prohibited Cognitive Mapping and GIS for Community-Based Resource Identification Introduct ion In December, 2003, the Colorado Institute of Public Policy (CIPP) at Colorado State University was contracted by Jefferson County, Colorado, Division of Human Services, to conduct a resource identification analysis The project was one component in the first year of a five-year Health and Human Services — Children’s Bureau grant to create a “systems of care” (SOC) in child welfare social service delivery.1 Jefferson County, Colorado, was one of eight pilot sites awarded an SOC grant The CIPP component was to identify services available at the community-level and discover services that were lacking SOC is a major paradigm shift in social service delivery It removes the locus of authority away from one individual, the social service worker, and replaces it with a group of service providers, family, and community members to develop collectively a comprehensive plan to move the child and family out of crisis The provision of services are to be coordinated, community-based, culturally competent and individualized (Stroul, 1986) To integrate Jefferson County social service delivery into community-based comprehensive child welfare SOC, both formal and informal services had to be identified Informal services are of particular interest since these are likely the least well-known (there was no official directory) and serve populations at a community SOC level (rather than county-wide) For definition purposes, informal services were identified for participants as private or not-for-profit programs, including services such as church soup kitchens, non-profit agencies providing transportation services for the elderly, and in-home daycare providers not registered with the county Formal services are public programs at the state, county, and local level, such as Jefferson County Mental Health Services, Title XX daycare providers, public schools, public transportation, and park and recreation programs 370 To identify existing resources at the community level, cognitive mapping, a process by which participants draw visual representations of geographical areas, was conducted with 247 participants in Jefferson County, Colorado Participant groups in the mapping included social service, non-profit, and faith-based providers, social service clients (Temporary Assistance for Needy Families (TANF) recipients, youths, foster care providers, and adoptive parents), residents and ethnic/racial enclaves (Latino, Eastern European, Native American, and African American) In addition, all resources listed in the Jefferson County resource guides were included in the resource database Over 3,800 unique resources were identified and entered into a GIS — ArcMap, a component of ArcView — to analyze the availability, capacity, and distribution of social services in the county and within communities Census data was overlaid to identify high-need areas and ethnic enclaves Here, a novel application of GIS for designing improved social service delivery systems is described The chapter also discusses complications involved in working with human service agencies, and reconsiders the processes developed to merge cognitive mapping information into ArcMap C ognitive Mapping: O rigin and U ses Cognitive mapping has not been comprehensively integrated with GIS to the degree discussed in this project, although elementary integration has been done in some previous projects (Fulton, Horan, & Serrano, 1997; Kathlene, 1997; Kathlene & Horan, 1998; Horan, Serrano, & McMurran, 2001).2 The potential usefulness to the human services sectors through projects that use the combined methodologies is substantial To better understand this “fit,” a brief review of cognitive mapping follows Cognitive mapping did not originate from research on humans Rather, the term “cognitive map” was originally used to describe the mental Cognitive Mapping and GIS for Community-Based Resource Identification representations that rats develop as they navigate the same maze multiple times (Tolman, 1948) Quickly, researchers became interested in using the information from cognitive maps in the human context, and the resulting collection of methods became known as cognitive mapping Later, the term expanded to include mental depictions of more abstract entities, like ideas or chains of events Since its inception, cognitive mapping has been used as an approach to a number of real-world issues Cognitive mapping’s first practical application was in the field of urban planning when Kevin Lynch (1960) found that certain places or elements in a city generated a positive emotional reaction among its residents This type of research, he felt, could be used to better design cities so as to make them more memorable Jack Nasar (1988) extended Lynch’s study by including a model of how city attributes influenced the affective responses of residents and visitors Cognitive mapping is also used to evaluate mental representations of smaller areas Mapping of specific neighborhoods (Quaiser-Pohl, Lehmann, & Eid, 2004; Uzzell, Pol, & Badenas, 2002), college campuses (Hardwick, Wooldridge, & Rinalducci, 1983; Holahan & Dobrowolny, 1978; Sholl, 1987), and buildings (Moeser, 1988; O’Laughlin & Brubaker, 1998; O’Neill, 1991) adds to knowledge about how spatial abilities develop and are utilized, the skills related to map creation, and the mapping abilities of various groups Numerous overlying maps of these smaller areas have been used to program autonomous mobile robots (Yoshino, 1991) Almost anyone can successfully complete a mapping exercise Studies have found the quality of female and male’s maps are similar (Evans, 1980; Magana & Norman, 1980; O’Laughlin & Brubaker, 1998); nearly all ages can successfully participate (Quaiser-Pohl, Lehmann, & Eid, 2004); and map quality is not influenced by drawing skill (Evans, 1980; Hardwick, Wooldridge, & Rinalducci, 1983) The only criteria that affects map quality is familiarity with the target area, where quality increases with familiarity (Evans, 1980; Fridgen, 1987; Unger & Wandersman, 1985) Familiarity, however, can distort maps For example, locations of importance to the participant are often drawn larger or more centrally than other map elements (Holahan & Dobrowolny, 1978; Kathlene, 1997) Finally, at the coding stage, the use of multiple coders without knowledge of the study hypothesis is commonly used (Daniels & Johnson, 2002); yet, studies have found nearly perfect agreement among multiple coders (Quaiser-Pohl, Lehmann, & Eid, 2004) For this project, the above issues did not present problems Each of the participants was directed to draw the area they self-identified as their most recognizable; therefore, they had high levels of familiarity The maps required particularly little drawing skill as participants were only asked to draw boundaries (for which they are provided a ruler), mark their home with an “X,” and identify landmarks with squares Since the objective in the study was to identify resources of importance to the participant; having the maps drawn to scale was irrelevant because the exact geographic coordinates for each location are determined during the coding process Coding the maps involved recording objective location information, thereby removing the problem of coder subjectivity In short, information gleaned from the maps was not dependent upon the drawing skills of the participants Ad apt at ion of Mapp ing Methods for R esource Ident if ic at ion Each use of mapping described above required a unique adaptation of the method Participants in these studies were asked to sketch maps, arrange pictures of pieces of paper according to their location or importance, locate points on a map, circle areas on a map, recognize features on an 371 Cognitive Mapping and GIS for Community-Based Resource Identification aerial photograph, or label features on a blank map The literature effectively suggests that as many methods of mapping exist as purposes for using it In this project, participants attending a mapping workshop were first provided background on the overall goals of the SOC project as well as their role as a mapper in achieving those goals The hope was that understanding the importance of the project would encourage conscientious participation Then, participants were provided with a large blank sheet of paper and asked to indicate the street or physical boundaries within which they most commonly travel Street maps of the area were available to help them in this and later tasks Workshops consisted of three to sixteen people, and at each session they were encouraged to work cooperatively with their fellow workshop participants Following a visualization exercise to assist them in accessing non-verbal memories, they were asked to fill in details such as streets, natural features, and resources they commonly use It was explained that these maps were not meant to be artistic works; simple boxes and lines drawn with their rulers would suffice to represent objects and locations Once the maps were complete, participants were asked to label various types of resources with colors corresponding to commercial businesses, community resources, government agencies, community strengths, and community weaknesses They also completed a survey asking for detailed description of the important resources and features on their maps and to provide any additional information regarding resources lacking in their area Each map was then entered into a GIS database Figure is an example of a cognitive map drawn for the Jefferson County Project G IS An alys is T echn ique The GIS for this cognitive mapping project provides aggregated spatial analysis of the maps 372 GIS combines the thousands of data points collected from the participant-drawn maps, which greatly enhances the value of cognitive mapping information In addition, each distinct physical address has dozens of descriptors that identify not only who provided the information but also secondary data such as the participants’ perceptions of a particular resource or service (Longley, Goodchild, Maguire, & Rhind, 2001) The data is accessible in Excel-type tables that can be easily queried and displayed on a digital street map (Longley, Goodchild, Maguire, & Rhind, 2001) This analysis allows for preliminary interpretation of the data set, such as the dispersion or clustering of resources in geographic areas (Heagerty & Lele, 1998; Reich & Davis, 2003) GIS can incorporate external data sources, such as census data, which can be layered with the cognitive mapping data Integrating the cognitive mapping data with GIS software creates nearly limitless ways to analyze the data and allows for easily-interpreted visual results Technical Capacity for the Project For this project the particular GIS capabilities were: • • • • Visual representation of addresses and areas via data points, data lines, and data shapes The ability to layer data points for sorting, categorizing, and selection purposes Accompanying tables of descriptive variables for each data layer with the ability to search and sort individual data points based on particular variable(s) The ability to export the data to conduct advanced spatial analysis The cognitive mapping process was greatly enhanced, especially for analysis purposes, by using GIS to create unique, specialized parameters (for a theoretical model, see Jordan, Raubal, Gartrell, Cognitive Mapping and GIS for Community-Based Resource Identification Figure Example of hand drawn cognitive map & Egenhofer, 1998) The potential outcomes of such a union include: • • • The capability to export a large quantity of data in an Excel-type format that can be searched and sorted based on any given number of criteria The ability to organize and sort data in a spatial manner (i.e in the form of maps) The opportunity to create an easily-updated database for organizing additional data gathered by the cognitive mapping process and through other resource identification processes Structure of the GIS for the Project The structure of the GIS system for this project was based on three principles: The structure needed to be flexible and allow for the maximum amount of data to be stored and easily accessed The structure needed to be relatively simple such that an inexperienced GIS user can understand the sorting, searching, and selecting potential of the system The GIS system needed to be able to organize data from a variety of sources and be easily updated over time 373 Cognitive Mapping and GIS for Community-Based Resource Identification The first step to organizing the data was to create basic social service or resource categories (in GIS terms these categories are referred to as layers) Each layer needed to be broad enough to encompass a variety of unique data points (identified resources) but narrow enough to create a good classification system for a beginning search and find procedure These basic layers were the essential framework of the GIS system and the search capabilities of the system that providers could use; therefore, they had to match either the basic social service field breakdown of resource types and/or be intuitively organized Ultimately, 24 layers were created for the project including Housing, Mental Health Services, Transportation, Health Care, Emergency Services, Domestic Violence, and Victim Services Within each layer were key fields with similar categories across all the layers For example, the data field “Resource Type” exists in each of the 24 layers and contains only one of five values: Governmental, Community Non-Profit, Faith-Based, Private, or Unknown This allowed queries and linkages to be made across layers As an illustration, a user could search for all the community non-profit sponsored resources or services in two or more of the layers and combine these results on one map Search or selection functions work on a layer by layer level in ArcMap Within the Housing layer there were attributes such as Resource Name, Resource Address, and Resource Hours of Availability These attributes are easily sorted For example, there should only be one Lakewood Housing Authority (LHA), but if there are two locations for LHA, they can be distinguished from one another in the data table due to the different values entered in the Resource Address cell However, not all searches are done based on a known name or location; therefore, a more extensive list of attributes must exist to properly describe each resource Some examples of other attribute identifiers were Resource Type (government-sponsored, private, faith-based, or community non-profit) or Funding Source Another type of attribute is a binary descriptor such as Used by Respondent Yes/No This would apply to data points identified by clients and useful in various types of qualitative analysis Below is a simplified example of a layer table structure and associated attributes It is easy to see in Table the large number of possible searches based on only seven attributes For example, a provider might want to find all government-sponsored housing resources The search would use “Resource Type=Government.” The first two rows would be selected from the search If housing resources provided by community non-profits is of interest, the search would be written as “Respondent Category=Client and Resource Type=Community Non-Profit and Used by Respondent=Yes.” In the table above, this would select only the last row Table Housing layer attribute table Survey ID Resource Name Resource type Hou r s of Availability Funding Type Used by Respondent 0105 Client LHA Government 9am-5pm Government No 0431 Provider LHA Government 9am-5pm Government N/A 0302 Community Member Lakeside Apartments Private N/A N/A Yes 0746 374 Respondent Category Client Allison Care Center Community Non-Profit N/A Private Yes Cognitive Mapping and GIS for Community-Based Resource Identification Each row represents one unique data point on the GIS map and therefore represents one unique resource Resources can be identified and placed on the map multiple times, as in the above example, with the first and second row both being the Resource Type “LHA.” What distinguishes the two data points, however, is not the location on the map but the multiple attributes in the table listed above Here, LHA was identified once by a provider and once by a client, information that indicates common knowledge Alternatively, if only clients identify the Allison Care Center, a knowledge gap in Jefferson County Human Services providers and resource manuals is revealed, which can be easily remedied Creating a comprehensive database with separate rows of multiple criteria allows for flexibility in the analysis of resources ranging from a simple resource search engine to a complex spatial analysis tool to numerically describe the layout of resources and services in Jefferson County G IS An alys is of C o mmun ity R esources To the GIS analysis, the resources identified by the mapping and focus group participants were verified using the 2004 Denver Metro phone book (www.dexonline.com) and/or MapQuest (www mapquest.com) The addresses were entered into an Excel spreadsheet, the type of resource and service was coded into one or more of the 24 types of resources, and the participant group and demographics were entered.3 The spreadsheet was imported into ArcMap and geo-coded to place the identified resources with known addresses on a common geographical coordinate system The 24 layers of resource and service types were created by: (1) working with participants in four pilot sessions to understand how they conceptualized services; (2) presenting the pilot workshop information to Jefferson County Human Service employees to add categories that aligned with their current resource guides and experiences; and (3) aligning with the Aires taxonomy used by Colorado 2-1-1 system.4 The resulting twenty-four layers allowed for some resources/services to be coded into more than one layer, thereby increasing the ability of information seekers to find the desired resources For example, a service dedicated to providing transportation to elderly individuals would be included in Aging and Adult/Senior Services as well as Transportation Services The dual coding in this project and used by the 2-1-1 system produces a comprehensive searchable database that can reach a specific resource through multiple avenues At the onset of the project, the research team decided that a method for determining when the mapping process had reached a saturation level of redundant information was needed A novel method, a “repeat rate,” was created The repeat rate was set at 80% based on the time and cost involved in identifying additional resources The repeat rate estimates the projected number of new points to be expected from each additional map So, for example, on the first map every identified resource is unique and a new point in the database, the repeat rate is zero By the 100th map, if 50% of the points collected were previously identified on any of the previous 99 maps, the non-repeated points for that map would be 50% At some point, theoretically, no new information will be gleaned from additional maps, resulting in a 100% repeat rate In practical terms, full saturation cannot be reached; but more importantly, a trade-off must be made between additional information and cost to acquire the information In this project, there was an average of 20 resource points per map Each map took approximately four hours to process (from address coding the points to data entry into ArcMap) An 80% threshold repeat rate was chosen as a break-even level At this rate, only one additional new resource would be identified per one hour of coding time Less than one additional new resource per hour of work was deemed an inefficient use of resources 375 Cognitive Mapping and GIS for Community-Based Resource Identification S e lected R esu lts While the cognitive mapping process identified formal and informal resources, without the ability to aggregated and stratified analyses the collected data would only have provided a list of resources known by the mappers To create a comprehensive SOC child welfare system requires detailed knowledge about the geographical distribution of the density of community resources and the types available in the community Additionally, to understand and correct the community-level information gaps among providers, community members and clients, a geographical analysis stratified by resource categories and information sources (e.g., county resource guides, social service workers, community providers, faith-based providers, and clients) was needed GIS is able to provide such geographical analysis Importantly, data presented in visual form facilitated information dissemination among a wide variety of stakeholders Seven Jefferson County cities were the focus of the mapping project Clients and community resident groups were drawn from Lakewood, Arvada, Wheat Ridge, Golden, Littleton, and the mountain communities of Conifer and Evergreen A total of 42 workshops were held at locations in each of these communities Special sessions were held with members of the dominant ethnic groups, identified using census data imported into ArcMap The groups included Native American, Russian, and Latino community residents,5 with Latinos being the largest ethnic population in Jefferson County To determine the areas in Jefferson County with the highest growth rate of Latinos, 1990 and 2000 census data was imported into ArcMap and the percent change was calculated The highest concentrations of the Latino population were found in the central eastern area of Jefferson County (Lakewood and Arvada) on the border of Denver County Figure shows the Latino growth patterns from 1990 to 2000 Numbers on the map are the percentage of the area’s Latino population in the 2000 Census 376 Current and past users of Jefferson County social services were identified by Jefferson County Since the database was confidential, Jefferson County would not allow CIPP to pull a stratified sample In addition, over half the client records were missing complete address information Because of these difficulties, all complete client files with zip codes in the study area were mailed invitations by Jefferson County to participate in the study Fortunately, the clients, foster, and adoptive parents who participated were distributed throughout the study area The following map shows the home location cross-streets of the participants and the extrapolated hypothetical boundaries of their identified communities, ranging from one-third mile to one mile around their residence The buffer area was based on actual community boundaries drawn on the maps, which ranged from larger than one mile to a handful that were smaller than one-third mile Figure indicates that the mapping participants who were current or previous users of Jefferson County social services had collective knowledge of nearly all of the study area (see Appendix A for demographic information of the participants) Figure shows the 3,845 resources identified and mapped The number of unique points is 1,819 after removing the repeated information Identified resources are distributed across all the Jefferson County study areas with a scattering throughout Denver County and a dense clustering along the main arterial east-west street, Colfax Next are selected categories of resources to illustrate how GIS was used to inform SOC planning efforts • Childcare: One hundred and eleven different childcare providers were identified As Figure shows, there is very little overlap in knowledge between the Jefferson County resource guides and the participants in the mapping workshops This could indicate: (1) clients and providers are not familiar with the childcare resources which Jefferson County Cognitive Mapping and GIS for Community-Based Resource Identification Figure Latino population growth in Jefferson County, 1990-2000 has identified; and (2) Jefferson County is not effectively disseminating this information Many of the childcare resources identified in the Jefferson County resource guides are before-school and after-school care programs In contrast, the cognitive mapping participants identified pre-school childcare facilities rather than before and after-school programs Based on this analysis, the current guides are lacking in pre-school childcare resources It is likely this information is housed with a specific organization rather than listed within Jeffco’s resource guides; however, a consolidated database will fa- • cilitate resource integration, information referral, and client access Education: There were 366 education resources identified This is the category in the Jefferson County guides with the most resources As would be expected, the Jefferson County Public Schools Resource Guide has all the primary and secondary public schools, as well as a few private schools in the county Figure shows there is very little additional information provided by the mapping sessions with regard to education resources in Jefferson County; however, the mapping identified a significant number of 377 Cognitive Mapping and GIS for Community-Based Resource Identification Figure Client home locations and range boundaries • 378 resources in Denver that are known and used by Jefferson County residents The Jefferson County education resources not identified in the resource guides were typically services such as breast-feeding classes located at hospitals, ESL classes available at community centers and/or libraries, and other educational services available at libraries, these important educational services are missing in Jefferson County’s guides Health care resources: Of the 190 health care resources, 140 were identified by the cognitive mapping process with only onethird of the services identified by both participants and the Jefferson County resource guides As seen in the Figure 7, however, the Jefferson County resource guides also have information about several health care services in Denver Most of the Denver and Jefferson County services were also identified by the cognitive mapping process Information from the cognitive mapping process increases the resource base by two-thirds, adding significantly to health care knowledge of Jefferson County Three of the 24 resource categories have been provided above for sake of brevity However, there were several trends that appeared in almost all of the resource categories First, the Jefferson Cognitive Mapping and GIS for Community-Based Resource Identification Figure All Jefferson County, Colorado, resources identified by participants and resource guides Figure Childcare resources identified by participants vs Jefferson County resource guides 379 ... campuses (Hardwick, Wooldridge, & Rinalducci, 1 983 ; Holahan & Dobrowolny, 19 78; Sholl, 1 987 ), and buildings (Moeser, 1 988 ; O’Laughlin & Brubaker, 19 98; O’Neill, 1991) adds to knowledge about how... M., & Madelaine, J (19 98) The geographical Anteserver: a Client/Server Architecture for GIS EOGEO’ 98 workshop Salzburg, Austria, (pp 78- 85) Wessel, M., & Haarslev, V (19 98) VISCO: Bringing Visual... 23(6), 1297–13 08 Longley, P A., Goodchild, M F., Maguire, D J., & Rhind, D W (2001) Geographic Information Systems and Science Chichester, UK: John Wiley & Sons Morgan, M G (19 98) Uncertainty Analysis in Risk Assessment