449 
CIP A 2003 XIX th International Symposium, 30 September - 04 October, 2003, Antalya, Turkey 
-, ¡- r 1 im 
- Pi Ì <* 
Figure 8 - Dating precision 
3.5 Selection by location 
This type of query allows the selection of different entities 
which are on different layers of the G.I.S. 
» # a * 9 a. 
modified or destroyed. In some cases the same building, or 
road, is mentioned in more than a document by using 
different name but other information seem to demonstrate the 
coincidence between the different objects described. 
In order to help the archaeologies to solve the correct 
identification of the elements, each information, also the less 
meaningful ones (names of the sites, etc.), has been recorded 
in a separate structure of the database. 
A specific algorithm has been developed in order to merge in 
an appropriate way all the acquired information about an 
object (e.g. a building, a road, a site, etc.) and to visualize the 
areas of different location likelihood. 
The specialist can choose, between all the selected areas the 
more interesting in order to make deeper investigations. 
4. PHOTO INTERPRETATION, REMOTE SENSING 
AND FIELD SURVEYS 
After the relational database structure design and 
implementation, the second step consisted in an 
archaeological survey, aimed to find out and locate all the 
structures mentioned in bibliographical sources. For almost 
everyone of these structures was extremely difficult to find 
* ** c ** * * v- *• r: r: 4 + >» <* n «t >• 
Figure 9 - Select by location of the quartz mines in a 
village coherent with geology information 
The selection can be done by intersection (e.g. the selection 
of an area crossed by a line), distance (e.g. the selection of al 
the entities with a fixed distance from a specified element), 
pattern (e.g. the selection of the entities completely enclosed 
in a different entity), superimposition (e.g. the section of the 
entities completely superimposed on other entities or which 
have line or point shared with other entities). 
A common application in the project has been the control of 
the material extracted form the mines and the geology of the 
area. Thanks to archaeological field survey we has been able 
to analyze the materials and to assume where were the mines 
they came from. The pattern analysis made by merging the 
document information and the actual geology confirmed or 
disproved the conjecture 
Another interesting application of the selection by location 
has been the selection of the building which could be used for 
activities needing water (see fig. 10): only the buildings far 
less than an appropriate distance can accomplish this 
hypothesis. 
3.6 Location likelihood 
One of the more difficult parts of the research has been the 
correct localisation of the sites and buildings mentioned by 
the documents. 
For examples, many buildings are well preserved and still 
used, but in many cases the building has been completely 
□ % 
Figure 10 — Selection by location of buildings with a distance 
form the river less than 50, 150, 200 and 250 m 
out the geospatial location. Sometime very precise details 
have been extracted from the documents but it was not 
possible to geo-referencing them by points on cartographic 
maps. Sometimes it was also possible to guess the area where 
the structure were probably located (see par. 3.6) 
In these cases, the data have been stored waiting for more 
accurate information (e.g. deriving by field survey) and G.I.S.