Summary - Spatial Data Analysis: An Introduction for GIS Users - page 172

for estimating data values at locations where there is no sample. h is topic could have
been organized in many dif erent ways and the balance of coverage of dif erent meth-
ods could have been altered markedly. Nonetheless, it is hoped that it covers in reason-
able detail a sui cient variety of kinds of approaches, if not specii c methods, to
demonstrate clearly the diversity of approaches.
h ere is very much more that could be said about the issues raised in this topic and
some suggestions as to how to proceed and further develop knowledge of spatial data
analysis are given in Section 11.4. Appendix G provides a list of some problems and
the corresponding solutions detailed in this topic.

Other issues

11.2

h ere are many issues touched on within this topic that could have been developed
substantially, and references to additional sources are given to allow readers to expand
their knowledge. h ere are also many issues that could have been discussed, but which
were considered outside the remit of the topic. However, this topic is intended as a
starting point that provides pointers to other sources where necessary and the cover-
age of material within the topic is necessarily focused on particular issues.
Many specii c issues that some readers might like to have seen covered are omitted,
and this is necessarily the case because of limitations of space and because the key aim
of this topic is to introduce in a focused way a relatively limited array of key concepts
and methods. As an example, the topic discusses connectivity in terms of identifying
neighbouring areas. An important area of research, with many applications, concerns
connectivity, as well as the form of landscape areas. McGarigal and Marks (1995) pres-
ent a sot ware package, Fragstats, which is designed to quantify landscape structure (e.g.
by size or shape of landscape patches or their density over an area). Such approaches
are not yet a standard part of GIS sot ware and are, therefore, not discussed in this
topic. A more general omission is geographic data mining. h e theme of geographic
data mining (Shekar et al. , 2003; Miller, 2008) is an important one. Geographic knowl-
edge discovery, with, as a core component, geographic data mining is based on the
belief that there is new and useful knowledge to be extracted from the vast array
of geographic data sources now available (Miller, 2008). h e development and use of
dynamic models is central to many applications, but this is another topic excluded
from this topic. It is hoped that the sources cited will provide the necessary information
where some topics are covered only briel y in this topic or omitted completely.

Problems

11.3

h is topic presents a variety of solutions to problems, but it only makes passing
reference to other core issues that have an impact on spatial data and their analysis.
h ere are widely used alternatives to most of the methods detailed this topic and the