The Meaning of Geographic Expressions

1
The Meaning of Geographic Expressions

• Dr Kristin Stock
• kristin.stock_at_nottingham.ac.uk

2
Q. These houses look like

1. Comfortable homes
2. A retirement village
3. A slum
4. Mansions

3
Q. I would describe this geographic feature as

1. A hill
2. An island
3. A volcano
4. A cay

4
Little round planetIn a big universeSometimes
it looks blessedSometimes it looks
cursedDepends on what you look at obviouslyBut
even more it depends on the way that you
seeBruce Cockburn
5
What does it all mean?
6

• How Do People Think?

• How Do Computers Think?

7
Bridging the gap...
8
What is a geographic expression?

• National Parks near Nottingham.
• Walking paths in the Peak District.
• The river is opposite the museum.
• The library is around the corner from the market
  square.
• The castle is near here.
• Which rivers go through the Lake District?
• Which mountains in the Alps are steep?
• Where is a large lake near Nottingham?

9
Geographic Expressions aremade up of...

• Geographic Features
• Spatial Relations
• Geographic Qualifiers

10
Geographic Features

• Objects that have a location relative to the
  earth.
• Natural, person-made or administrative.
• Categories or instances (instance a specific
  object).
• Usually nouns in English.

11
Spatial Relations

• Words that describe the relationship in space
  between two features.
• Usually verbs in English.

intersects
touches
on
covers
near
contains
crosses
in
next to
12
Geographic Qualifiers

• Further restrict or qualify a geographic feature.
• Sometimes vague in interpretation.
• Usually adjectives in English.

13
Q. The parts of these expressions shown in red
are

• National Parks near Nottingham.
• Walking paths in the Peak District.
• The river is opposite the museum.
• The library is around the corner from the market
  square.
• Is the castle near here?
• Which rivers go through the Lake District?
• Which mountains in the Alps are steep?
• Where is a large lake near Nottingham?

1. Geographic object categories
2. Geographic object instances
3. Spatial relations
4. Geographic qualifiers

14
Q. An example of a spatial relation is

1. around the corner
2. go through
3. tallest
4. Nottingham
5. River
6. step-mother

15
Q. large is

1. A geographic object category
2. A spatial relation
3. A geographic qualifier
4. A geographic object instance
5. An administrative relation

16
Spatial Relations Exercise
17
(No Transcript)
18
Q. I would describe this geographic feature as a

1. Creek
2. Beck
3. River
4. Stream
5. Road

19
Q. I would describe this geographic feature as a

1. Hill
2. Hillock
3. Mountain
4. Tor
5. Munro
6. Pingo

20
How do the meanings of geographic expressions
vary?

• Culture.
• Background.
• Education.
• Environment.
• Context (dynamic).

21
So, how do people think about geographic features?

• Differently from each other (to varying degrees).
• Sometimes vaguely.
• Sometimes context sensitively.

22
But, how do computers think about geographic
information?
23
Geographic Expressions

• Have a precise, fixed meaning
• The same for everyone
• Not context sensitive
• Not vague.
• Have rigid, crisp (not fuzzy) physical
  boundaries.

24
Geographic Features

• Census output area polygons, lower level
  super-output area polygons, ward polygons.
• County polygons.
• National Park polygons.
• Road network.
• Buildings from Historical Digimap.

25
Spatial Relations

• Query operators allowed by ArcMap
• Intersect
• Completely contain
• Share a line segment with
• Touch the boundary of
• Have their centroid in
• Like geographic features, spatial operators in
  GIS are also crisp and rigid.

26
Q. Which spatial operator would you use to find
the buildings in Smith Street?

1. Contains
2. Crosses
3. Overlaps
4. Touches
5. Within
6. Near

27
Q. How would you query geographic
qualifiers(steep, large) in a GIS?
Geographic Qualifiers

1. I would look for an attribute containing the
   relevant information.
2. I would look at the spatial relations to try to
   deduce the qualifier values.
3. I would ask the person next to me.
4. I wouldnt.

28
Summary

• People think uniquely, context-sensitively,
  sometimes vaguely.
• Computers think precisely, crisply, uniformly.
• When using current mainstream GIS, people have to
  bridge the gap.
• We are working on ways to help computers bridge
  the gap...