1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
|
<?xml version='1.0' encoding="UTF-8"?>
<part id="tracker-ontologies" xmlns:xi="http://www.w3.org/2003/XInclude">
<title>Defining ontologies</title>
<partintro>
<para>
An ontology defines the entities that a Tracker endpoint can store, as
well as their properties and the relationships between different entities.
</para>
<para>
Tracker internally uses the following ontologies as its base, all ontologies
defined by the user of the endpoint are recommended to be build around this
base:
</para>
<formalpara>
<title><systemitem>XML Schema (XSD).</systemitem></title>
<para>Defining basic types.</para>
</formalpara>
<formalpara>
<title><systemitem>Resource Description Framework (RDF).</systemitem></title>
<para>Defining classes, properties and inheritance</para>
</formalpara>
<formalpara>
<title><systemitem>Nepomuk Resource Language (NRL).</systemitem></title>
<para>Defining resource uniqueness and cardinality</para>
</formalpara>
<formalpara>
<title><systemitem>Dublin Core (DC).</systemitem></title>
<para>Defining common superproperties for documents</para>
</formalpara>
<formalpara>
<title><systemitem>Nepomuk Annotation Ontology (NAO).</systemitem></title>
<para>Implementing tagging and annotations</para>
</formalpara>
<para>
Ontologies are RDF files with the .ontology extension, Tracker parses all
ontology files from the given directory. The individual ontology files may
ontologies may not be self-consistent (i.e. use missing definitions), but
all the ontology files as a whole must be.
</para>
<para>
Tracker loads the ontology files in alphanumeric order, it is advisable
that those have a numbered prefix in order to load those at a consistent
order despite future additions.
</para>
</partintro>
<chapter id="creating-ontology">
<title>Creating an ontology</title>
<section id="defining-namespaces">
<title>Defining a namespace</title>
<para>
A namespace is the topmost layer of an individual ontology, it will
contain all classes and properties defined by it. In order to define
a namespace you can do:
</para>
<programlisting><xi:include href="examples/ontologies/defining-namespaces-1.txt" parse="text"/></programlisting>
</section>
<section id="defining-classes">
<title>Defining classes</title>
<para>
Classes are the base of an ontology, all stored resources must define
themselves as "being" at least one of these classes. They all derive
from the base rdfs:Resource type. To eg. define classes representing
animals and plants, you can do:
</para>
<programlisting><xi:include href="./examples/ontologies/defining-classes-1.txt" parse="text"/></programlisting>
<para>
By convention all classes use CamelCase names, although class names
are not restricted. The allowed charset is UTF-8.
</para>
<para>
Declaring subclasses is possible:
</para>
<programlisting><xi:include href="./examples/ontologies/defining-classes-2.txt" parse="text"/></programlisting>
<para>
With such classes defined, resources may be inserted to the endpoint,
eg. with the SPARQL:
</para>
<programlisting><xi:include href="./examples/ontologies/defining-classes-3.rq" parse="text"/></programlisting>
<para>
Note that multiple inheritance is possible, resources will just inherit
all properties from all classes and superclasses.
</para>
</section>
<section id="defining-properties">
<title>Defining properties</title>
<para>
Properties relate to a class, so all resources pertaining to that class
can define values for these.
</para>
<programlisting><xi:include href="./examples/ontologies/defining-properties-1.txt" parse="text"/></programlisting>
<para>
The class the property belongs to is defined by rdfs:domain, while the
data type contained is defined by rdfs:range. By convention all
properties use lowerCamelCase names, although property names are not
restricted. The allowed charset is UTF-8.
</para>
<para>
The following basic types are supported:
</para>
<formalpara>
<title><systemitem>xsd:boolean</systemitem></title>
</formalpara>
<formalpara>
<title><systemitem>xsd:string</systemitem></title>
</formalpara>
<formalpara>
<title><systemitem>xsd:integer</systemitem></title>
<para>Ranging from -2^63 to 2^63-1.</para>
</formalpara>
<formalpara>
<title><systemitem>xsd:double</systemitem></title>
<para>Able to store a 8 byte IEEE floating point number.</para>
</formalpara>
<formalpara>
<title><systemitem>xsd:date and xsd:dateTime</systemitem></title>
<para>
Able to store dates and times since January 1, 1970 UTC, with
millisecond resolution.
</para>
</formalpara>
<para>
Of course, properties can also point to resources of the same or
other classes, so stored resources can conform a graph:
</para>
<programlisting><xi:include href="./examples/ontologies/defining-properties-2.txt" parse="text"/></programlisting>
<para>
There is also inheritance of properties, an example would be a property
in a subclass concretizing a more generic property from a superclass.
</para>
<programlisting><xi:include href="./examples/ontologies/defining-properties-3.txt" parse="text"/></programlisting>
<para>
SPARQL queries are expected to provide the same result when queried
for a property or one of its superproperties.
</para>
<programlisting><xi:include href="./examples/ontologies/defining-properties-4.rq" parse="text"/></programlisting>
</section>
<section id="defining-cardinality">
<title>Defining cardinality of properties</title>
<para>
By default, properties are multivalued, there are no restrictions in
the number of values a property can store.
</para>
<programlisting><xi:include href="./examples/ontologies/defining-cardinality-1.rq" parse="text"/></programlisting>
<para>
Wherever this is not desirable, cardinality can be limited on properties
through nrl:maxCardinality.
</para>
<programlisting><xi:include href="./examples/ontologies/defining-cardinality-2.txt" parse="text"/></programlisting>
<para>
This will raise an error if the SPARQL updates in the endpoint end up
in the property inserted multiple times.
</para>
<programlisting><xi:include href="./examples/ontologies/defining-cardinality-3.rq" parse="text"/></programlisting>
<note>
Tracker does not implement support for other maximum cardinalities
than 1.
</note>
<!--
XXX: explain how cardinality affects subproperties, superproperties
-->
</section>
<section id="defining-uniqueness">
<title>Defining uniqueness</title>
<para>
It is desirable for certain properties to keep their values unique
across all resources, this can be expressed by defining the properties
as being a nrl:InverseFunctionalProperty.
</para>
<programlisting><xi:include href="./examples/ontologies/defining-uniqueness-1.txt" parse="text"/></programlisting>
<para>
With that in place, no two resources can have the same value on the
property.
</para>
<programlisting><xi:include href="./examples/ontologies/defining-uniqueness-2.rq" parse="text"/></programlisting>
<!--
XXX: explain how inverse functional proeprties affect sub/superproperties
-->
</section>
<section id="defining-indexes">
<title>Defining indexes</title>
<para>
It may be the case that SPARQL queries performed on the endpoint are
known to match, sort, or filter on certain properties more often than others.
In this case, the ontology may use tracker:domainIndex in the class definition:
</para>
<programlisting><xi:include href="./examples/ontologies/defining-indexes-1.txt" parse="text"/></programlisting>
<para>
Classes may define multiple domain indexes.
</para>
<note>
Be frugal with indexes, do not add these proactively. An index in the wrong
place might not affect query performance positively, but all indexes come at
a cost in disk size.
</note>
</section>
<section id="defining-fts-indexes">
<title>Defining full-text search properties</title>
<para>
Tracker provides nonstandard full-text search capabilities, in order to use
these, the string properties can use tracker:fulltextIndexed:
</para>
<programlisting><xi:include href="./examples/ontologies/defining-fts-indexes-1.txt" parse="text"/></programlisting>
<para>
Weighting can also be applied, so certain properties rank higher than others
in full-text search queries. With tracker:fulltextIndexed in place, sparql
queries may use full-text search capabilities:
</para>
<programlisting><xi:include href="./examples/ontologies/defining-fts-indexes-2.rq" parse="text"/></programlisting>
</section>
<section id="predefined-elements">
<title>Predefined elements</title>
<para>
It may be desirable for the ontology to offer predefined elements of a
certain class, which can then be used by the endpoint.
</para>
<programlisting><xi:include href="./examples/ontologies/predefined-elements-1.txt" parse="text"/></programlisting>
<para>
Usage does not differ in use from the elements of that same class that
could be inserted in the endpoint.
</para>
<programlisting><xi:include href="./examples/ontologies/predefined-elements-2.rq" parse="text"/></programlisting>
</section>
</chapter>
<chapter id="accompanying-metadata">
<title>Accompanying metadata</title>
<para>
Ontology files are optionally accompanied by description files, those have
the same basename, but the ".description" extension.
</para>
<programlisting><xi:include href="./examples/ontologies/example.description" parse="text"/></programlisting>
</chapter>
<chapter id="updating-ontology">
<title>Updating an ontology</title>
<para>
As software evolves, sometimes changes in the ontology are unavoidable.
Tracker can transparently handle certain ontology changes on existing
databases.
</para>
<formalpara>
<title><systemitem>Adding a class.</systemitem></title>
</formalpara>
<formalpara>
<title><systemitem>Removing a class.</systemitem></title>
<para>
All resources will be removed from this class, and all related
properties will disappear.
</para>
</formalpara>
<formalpara>
<title><systemitem>Adding a property.</systemitem></title>
</formalpara>
<formalpara>
<title><systemitem>Removing a property.</systemitem></title>
<para>
The property will disappear from all elements pertaining to the
class in domain of the property.
</para>
</formalpara>
<formalpara>
<title><systemitem>Changing rdfs:range of a property.</systemitem></title>
<para>
The following conversions are allowed:
</para>
<variablelist>
<varlistentry><listitem>xsd:integer to xsd:bool, xsd:double and xsd:string</listitem></varlistentry>
<varlistentry><listitem>xsd:double to xsd:bool, xsd:integer and xsd:string</listitem></varlistentry>
<varlistentry><listitem>xsd:string to xsd:bool, xsd:integer and xsd:double</listitem></varlistentry>
</variablelist>
</formalpara>
<formalpara>
<title><systemitem>Adding and removing tracker:domainIndex from a class.</systemitem></title>
</formalpara>
<formalpara>
<title><systemitem>Adding and removing tracker:fulltextIndexed from a property.</systemitem></title>
</formalpara>
<formalpara>
<title><systemitem>Changing the tracker:weight on a property.</systemitem></title>
</formalpara>
<formalpara>
<title><systemitem>Removing nrl:maxCardinality from a property.</systemitem></title>
</formalpara>
<!--
XXX: these need documenting too
add intermediate superproperties
add intermediate superclasses
remove intermediate superproperties
remove intermediate superclasses
-->
<para>
However, there are certain ontology changes that Tracker will find
incompatible. Either because they are incoherent or resulting into
situations where it can not deterministically satisfy the change
in the stored data. Tracker will error out and refuse to do any data
changes in these situations:
</para>
<variablelist>
<varlistentry><listitem>
Properties with rdfs:range being xsd:bool, xsd:date, xsd:dateTime,
or any other custom class are not convertible. Only conversions
covered in the list above are accepted.
</listitem></varlistentry>
<varlistentry><listitem>
You can not add rdfs:subClassOf in classes that are not being
newly added. You can not remove rdfs:subClassOf from classes.
The only allowed change to rdfs:subClassOf is to correct
subclasses when deleting a class, so they point a common
superclass.
</listitem></varlistentry>
<varlistentry><listitem>
You can not add rdfs:subPropertyOf to properties that are not
being newly added. You can not change an existing
rdfs:subPropertyOf unless it is made to point to a common
superproperty. You can however remove rdfs:subPropertyOf from
non-new properties.
</listitem></varlistentry>
<varlistentry><listitem>
Properties can not move across classes, thus any change in
rdfs:domain forbidden.
</listitem></varlistentry>
<varlistentry><listitem>
You can not add nrl:maxCardinality restrictions on properties that
are not being newly added.
</listitem></varlistentry>
<varlistentry><listitem>
You can not add nor remove nrl:InverseFunctionalProperty from a
property that is not being newly added.
</listitem></varlistentry>
</variablelist>
<para>
The recommendation to bypass these situations is the same for all,
use different property and class names and use SPARQL to manually
migrate the old data to the new format if necessary.
</para>
<para>
High level code is in a better position to solve the
possible incoherences (e.g. picking a single value if a property
changes from multiple values to single value). After the manual
data migration has been completed, the old classes and properties
can be dropped.
</para>
</chapter>
</part>
|