e Republic 
g. Julijske 
iation (eg. 
Ribnici — 
ntry names 
hort forms 
F), 
tu — Pr’ 
hy. The 
kind of 
gs are left 
n, whether 
ie level of 
about the 
anging the 
ffirmative. 
d together 
uently for 
neznik — 
names etc. 
al entities 
Ormoske 
, all these 
to be kept 
se of map 
'acters), 
le), 
ngle (real 
| in carto- 
designer 
ntity type 
base was 
tail based 
e uniform 
hree main 
Each geographical entity could have more than one name; there 
is a relation 1 : N. Each geographical name could also be used 
in more than one map annotation; there is a relation 1 : N, too. 
This reorganization of the database is going to remove the 
redundancy in the database. Some other attributes are added for 
the three main entity types, to be able to perform standar- 
dization of the most general level of the database. 
ENTITIES i 
    
   
    
     
Springs 
  
  
NAMES ANNOTATIONS 
ID 
noi [ID] [ID] 
TITY 
ID-EN IDNAME «R——————-—» — ID NAME 
"a a 
NAME STRING SOURCE 
LANGUAGE SCALE 
Rivers 
[ID] 
| 
STATUS TEXT HEIGHT 
[START. NODE] 
CORRECTNESS ROTATION 
[END NODE] 
TERRIER TYPE X MIN 
[DIRECTION] J 
ID_ENTITY Y_MIN 
[DISTANCE] = 
SPELLING X_MAX 
ID_ENTITY + = 
LE" PUNCTUATION Y. MAX 
Lakes RT ER 
[ID] 
[AREA] 
[PERIMETER] 
ID ENTITY + 
m 
Hu 
  
—áo 
Figure 2: Scheme of the REZI database. 
The new scheme of the REZI database (see Figure 2) is formed 
of two main relational tables (map annotations and 
geographical names), and a graphical part (geographical 
entities). Point, line and area features are provided in the 
graphical part, but this part of the database is only a supplement 
for the top level data (standardized names). 
8. HARMONIZATION AND MAINTENANCE OF 
TOPONYMIC DATABASE 
The harmonization of the database is going to be done during 
this year. Of course, some additional problems are to be 
expected. Some annotations comprehend more than one name 
(bilingual names, eg. PortoroZ/Portorose) that should be 
separated. As already mentioned, some names are combined 
with common nouns. Merging map annotations together have to 
be done very carefully. Having the same map inscription does 
not always mean that the map annotations represents the same 
geographical entity; there are many geographical entities with 
the same name. The analysis of repetition of the settlement 
names in Slovenia (Berk et al., 2002) showed that almost every 
fifth settlement (20 %) has at least one namesake. The most 
frequent names are: Pristava, Dolenja vas, Gradisce, Potok, and 
Ravne. All these appearances have to be treated as separate 
names even though they sound the same. Lots of work has to be 
done manually. 
There is no standardization that could freeze toponymic 
database once for all, because toponyms could always be 
changed. So, important thing there is the maintenance of the 
database. The analysis of changes of the settlement names in 
Slovenia (Berk et al., 2002) gave about 20 name changes per 
year in the period from 1948 to 1994. Some names changed 
5] 
even more than once (eg. Podvin pri Polzeli; from 1952 to 1963 
Podvin Vranski, and before 1952 Podvin). In the REZI database 
the name history is assured, too. It is provided that the new 
name changes are to be entered in the REZI database at least 
every six months. 
9. CONCLUSIONS 
The establishment of the toponymic database does not look 
very complicated at first sight. Using it only for cartography, it 
is only one additional layer of topographic (cartographic) data- 
base. For each level of detail (map scale), of course. But trying 
to uniform the use of toponyms on the national level — the final 
step is standardization of toponyms — one may run into 
troubles. To be able to solve them successfully, it is important 
to foresee, what could all go wrong. The early phase of the 
toponymic database creation should answere to the questions of 
toponym definition, linguistic, geographic and other 
classifications of toponyms, dealing with variant names, and 
providing the levels of eventual standardization of toponyms. In 
this way also the redundancy of names in the database should 
be removed. The uniform toponymic database should cover not 
only the needs of national cartography, but many other users, 
too. 
REFERENCES 
Berk, S., Radovan, D., Petrovi¢, D., 2002. Prenova registra 
zemljepisnih imen (The Renovation of the Register of Geo- 
graphical Names). Project documentation. Geodetski institut 
Slovenije. Ljubljana, 3 vol., 269 pp. 
Furlan, M., GloZancev, A., Sivic - Dular, A., 2001. Pravopisno 
ustrezen zapis zemljepisnih in stvarnih lastnih imen v Registru 
zemljepisnih imen in Registru prostorskih enot (Orthographi- 
cally Correct Writing of Geographical and Other Proper 
Names in the Register of Geographical Names and the Register 
of Spatial Units). Geodetska uprava Republike Slovenije, 
Ljubljana, 58 pp. 
GU RS, 2001. Zgosceni imenik zemljepisnih imen Slovenije — 
Concise Gazetteer of Slovenia. Geodetska uprava Republike 
Slovenije, Ljubljana, 48 pp. 
Kladnik, D., 1999. Imenik tujih imen v slovenskem jeziku (The 
List of Foreign Names in the Slovene Language). Geografski 
institut Antona Melika. Znanstvenoraziskovalni center Slo- 
venske akademije znanosti in umetnosti, Ljubljana, 139 pp. 
Radovan, D., Ferlan, M., Petrovi¢, D., 1996. Evidenca zemlje- 
pisnih imen: konceptualni, logicni in fizicóni model z navodilom 
za vzpostavitev (The Evidence of Geographical Names: Logical 
and Phisical Model with the Guidelines for Establishment). 
Institut za geodezijo in fotogrametrijo FGG, Ljubljana, 67 pp. 
Radovan, D., Rojc, B., Ferlan, M., Ambrozi¢, M., Sumrada, R., 
Rozman, J., 1993. Priprava tehnoloskih osnov in vzpostavitve 
enotne baze evidence zemljepisnih imen (The Arrangement of 
Technological Knowledge and Establishment of the Uniform 
Evidence of Geographical Names). Technical report. Institut za 
geodezijo in fotogrametrijo FAGG, 2 vol., 199 pp.