-
Notifications
You must be signed in to change notification settings - Fork 0
/
Copy pathMultisensor-LD.html
1450 lines (1233 loc) · 51.4 KB
/
Multisensor-LD.html
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
353
354
355
356
357
358
359
360
361
362
363
364
365
366
367
368
369
370
371
372
373
374
375
376
377
378
379
380
381
382
383
384
385
386
387
388
389
390
391
392
393
394
395
396
397
398
399
400
401
402
403
404
405
406
407
408
409
410
411
412
413
414
415
416
417
418
419
420
421
422
423
424
425
426
427
428
429
430
431
432
433
434
435
436
437
438
439
440
441
442
443
444
445
446
447
448
449
450
451
452
453
454
455
456
457
458
459
460
461
462
463
464
465
466
467
468
469
470
471
472
473
474
475
476
477
478
479
480
481
482
483
484
485
486
487
488
489
490
491
492
493
494
495
496
497
498
499
500
501
502
503
504
505
506
507
508
509
510
511
512
513
514
515
516
517
518
519
520
521
522
523
524
525
526
527
528
529
530
531
532
533
534
535
536
537
538
539
540
541
542
543
544
545
546
547
548
549
550
551
552
553
554
555
556
557
558
559
560
561
562
563
564
565
566
567
568
569
570
571
572
573
574
575
576
577
578
579
580
581
582
583
584
585
586
587
588
589
590
591
592
593
594
595
596
597
598
599
600
601
602
603
604
605
606
607
608
609
610
611
612
613
614
615
616
617
618
619
620
621
622
623
624
625
626
627
628
629
630
631
632
633
634
635
636
637
638
639
640
641
642
643
644
645
646
647
648
649
650
651
652
653
654
655
656
657
658
659
660
661
662
663
664
665
666
667
668
669
670
671
672
673
674
675
676
677
678
679
680
681
682
683
684
685
686
687
688
689
690
691
692
693
694
695
696
697
698
699
700
701
702
703
704
705
706
707
708
709
710
711
712
713
714
715
716
717
718
719
720
721
722
723
724
725
726
727
728
729
730
731
732
733
734
735
736
737
738
739
740
741
742
743
744
745
746
747
748
749
750
751
752
753
754
755
756
757
758
759
760
761
762
763
764
765
766
767
768
769
770
771
772
773
774
775
776
777
778
779
780
781
782
783
784
785
786
787
788
789
790
791
792
793
794
795
796
797
798
799
800
801
802
803
804
805
806
807
808
809
810
811
812
813
814
815
816
817
818
819
820
821
822
823
824
825
826
827
828
829
830
831
832
833
834
835
836
837
838
839
840
841
842
843
844
845
846
847
848
849
850
851
852
853
854
855
856
857
858
859
860
861
862
863
864
865
866
867
868
869
870
871
872
873
874
875
876
877
878
879
880
881
882
883
884
885
886
887
888
889
890
891
892
893
894
895
896
897
898
899
900
901
902
903
904
905
906
907
908
909
910
911
912
913
914
915
916
917
918
919
920
921
922
923
924
925
926
927
928
929
930
931
932
933
934
935
936
937
938
939
940
941
942
943
944
945
946
947
948
949
950
951
952
953
954
955
956
957
958
959
960
961
962
963
964
965
966
967
968
969
970
971
972
973
974
975
976
977
978
979
980
981
982
983
984
985
986
987
988
989
990
991
992
993
994
995
996
997
998
999
1000
<?xml version="1.0" encoding="iso-8859-1"?>
<!DOCTYPE html PUBLIC "-//W3C//DTD XHTML 1.0 Strict//EN"
"http://www.w3.org/TR/xhtml1/DTD/xhtml1-strict.dtd">
<html xmlns="http://www.w3.org/1999/xhtml" lang="en" xml:lang="en">
<head>
<title>Multisensor Linked Data and Ontologies</title>
<meta http-equiv="Content-Type" content="text/html;charset=iso-8859-1"/>
<meta name="title" content="Multisensor Linked Data and Ontologies"/>
<meta name="generator" content="Org-mode"/>
<meta name="generated" content="2014-05-28"/>
<meta name="author" content="Vladimir Alexiev, Ontotext Corp"/>
<meta name="description" content=""/>
<meta name="keywords" content=""/>
<!-- configuration parameters --> <meta name='defaultView' content='slideshow' /> <meta name='controlVis' content='hidden' /> <!-- style sheet links --> <link rel='stylesheet' href='ui/default/slides.css' type='text/css' media='projection' id='slideProj' /> <link rel='stylesheet' href='ui/default/outline.css' type='text/css' media='screen' id='outlineStyle' /> <link rel='stylesheet' href='ui/default/print.css' type='text/css' media='print' id='slidePrint' /> <link rel='stylesheet' href='ui/default/opera.css' type='text/css' media='projection' id='operaFix' /><!-- S5 JS --> <script src='ui/jquery.js' type='text/javascript'></script> <script src='ui/org-slides.js' type='text/javascript'></script> <script src='ui/default/slides.js' type='text/javascript'></script>
<script type="text/javascript">
<!--/*--><![CDATA[/*><!--*/
function CodeHighlightOn(elem, id)
{
var target = document.getElementById(id);
if(null != target) {
elem.cacheClassElem = elem.className;
elem.cacheClassTarget = target.className;
target.className = "code-highlighted";
elem.className = "code-highlighted";
}
}
function CodeHighlightOff(elem, id)
{
var target = document.getElementById(id);
if(elem.cacheClassElem)
elem.className = elem.cacheClassElem;
if(elem.cacheClassTarget)
target.className = elem.cacheClassTarget;
}
/*]]>*///-->
</script>
</head>
<body>
<div id="preamble">
<div class='layout'>
<div id='controls'></div>
<div id='currentSlide'></div>
<div id='header'></div>
<div id='footer'><h1>Multisensor Linked Data and Ontologies</h1><h2>Vladimir Alexiev, Ontotext Corp • 2014-05-28</h2></div>
</div>
<div class='presentation'>
<div class='slide front'>
<div id='front-logo'></div>
<h1 class='front'>Multisensor Linked Data and Ontologies</h1><h3>Vladimir Alexiev, Ontotext Corp</h3></div>
</div>
<div id="content">
<h1 class="title">Multisensor Linked Data and Ontologies</h1>
<div id="outline-container-1-" class="outline-2">
<h2 id="sec-1-">Navigating This Presentation</h2>
<div class="outline-text-2" id="text-1-">
<ul>
<li>Ext links shown icon, eg <a href="http://www.ontotext.com">http://www.ontotext.com</a>
</li>
<li>This web-based presentation is made with <a href="http://meyerweb.com/eric/tools/s5/">S5</a> and plain text
</li>
<li><i>Press F11 for full-screen now</i> (on small screen or Opera)
</li>
<li>Navigating:
<ul>
<li><span style="text-decoration:underline;">Click, Right, Down, PageDown</span>: next slide (Opera: only PageDown)
</li>
<li><span style="text-decoration:underline;">Left, Up, PageUp</span>: prev slide (Opera: only PageUp)
</li>
<li><span style="text-decoration:underline;">Home</span>: first slide
</li>
<li><span style="text-decoration:underline;">End</span>: last slide
</li>
<li><i>nn</i> <span style="text-decoration:underline;">Enter</span>: go to slide <i>nn</i>
</li>
<li><span style="text-decoration:underline;">t</span>: toggle between slide view and single-page HTML view (for printing)
</li>
<li><span style="text-decoration:underline;">c</span> (or hover at bottom-right corner): show control panel<br/>
<img src="./img/s5-control-panel.png" width=150 alt="./img/s5-control-panel.png" />
<ul>
<li>toggle, prev, next icons
</li>
<li><b>table of contents</b> listbox
</li>
</ul>
</li>
</ul>
</li>
</ul>
</div>
</div>
<div id="outline-container-2-" class="outline-2">
<h2 id="sec-2-">Outline</h2>
<div class="outline-text-2" id="text-2-">
<p>2014-05-19 Mon
</p><ul>
<li>General linked data
<ul>
<li>FactForge and the datasets that it integrates
</li>
<li>Possible changes/additions
</li>
</ul>
</li>
<li>Use-case specific linked data
<ul>
<li>Energy Simulation; comparison to 2 automatic extractions
</li>
<li>Energy Thesauri
</li>
<li>Energy Datasets
</li>
<li>Use-case specific DBpedia subsets
</li>
<li>Household appliances
</li>
</ul>
</li>
<li>Linguistic Linked Data in FactForge
<ul>
<li>WordNet, Lingvo, Lexvo
</li>
</ul>
</li>
</ul>
<p>2014-05-21 Wed
</p><ul>
<li>Linguistic ontologies
<ul>
<li>NIF
</li>
<li>OLIA and its constituents
</li>
<li>LEMON, GOLD, ISOcat??
</li>
</ul>
</li>
<li>New Linguistic Linked Data
<ul>
<li>WordNet RDF
</li>
<li>Wiktionary
</li>
<li>BabelNet
</li>
<li>UBYlemon
</li>
</ul>
</li>
<li>News ontology
</li>
</ul>
</div>
</div>
<div id="outline-container-3-" class="outline-2">
<h2 id="sec-3-">General linked data</h2>
<div class="outline-text-2" id="text-3-">
<p>FactForge
</p><ul>
<li>DBpedia
</li>
<li>GeoNames
</li>
<li>FreeBase
</li>
<li>New York Times
</li>
<li>CIA World FactBook
</li>
<li>MusicBrainz (irrelevant)
</li>
<li>WordNet, Lingvoj, Lexvo
</li>
</ul>
<p>Possible changes/additions
</p><ul>
<li>Update datasets
</li>
<li>BaseKB instead of FreeBase
</li>
<li>OpenStreetMap
</li>
</ul>
</div>
</div>
<div id="outline-container-4-" class="outline-2">
<h2 id="sec-4-">FactForge</h2>
<div class="outline-text-2" id="text-4-">
<p><a href="http://factforge.net/"><img src="./img/factforge-large.png" alt="factforge-large.png"/></a>
</p><ul>
<li>An RDF warehouse of the 9 most central LOD datasets (red below)<br/>
<img src="./img/lod-datasets-2009-03-27-FactForge-LLD.jpg" alt="./img/lod-datasets-2009-03-27-FactForge-LLD.jpg" />
</li>
<li>A reason-able view over the web of data, which allows efficient linking and reasoning.
</li>
<li>Exposed to Multisensor as <a href="http://render.ontotext.com/">http://render.ontotext.com/</a> (allows writing).
</li>
<li><a href="http://www.ontotext.com/factforge/statistics3">Dataset statistics</a>
</li>
</ul>
</div>
</div>
<div id="outline-container-5-" class="outline-2">
<h2 id="sec-5-">PROTON</h2>
<div class="outline-text-2" id="text-5-">
<p>Problem: each of the LOD datasets comes with its own ontology.
</p><ul>
<li>Freebase isn't even structured according to RDF/RDFS principles.
</li>
</ul>
<p>Our approach:
</p><ul>
<li><a href="http://www.ontotext.com/proton-ontology">PROTON</a> (PROTo ONtology): a lightweight upper-level ontology
</li>
<li>Serves as a reference mapping layer so you can access the integrated LOD datasets in a uniform manner
</li>
<li>Examples at <a href="http://factforge.net/sparql">http://factforge.net/sparql</a> (contrasts access throuh PROTON vs DBpedia and GeoNames ontologies)
</li>
<li><a href="http://mklab2.iti.gr/multisensor/index.php/Render_OWLIM_Repository">RENDER Repository</a> page in the wiki, including presentation from Sofia meeting
</li>
</ul>
<p>References:
</p><ul>
<li><a href="http://www.dit.unitn.it/~p2p/OM-2010/om2010_Tpaper6.pdf">Mapping the central LOD ontologies to PROTON upper-level ontology</a>, Ontology Mapping Workshop at ISWC 2010, Shanghai, China, November 2010
</li>
<li><a href="http://www.ontotext.com/sites/default/files/publications/S3T-MuseumreasonableView_v7_cameraReady-30Jun.pdf">Reason-able View of Linked Data for Cultural Heritage</a>, Proceedings of S3T'2011, Burgas, Bulgaria, September 2011
</li>
<li>Creation and Integration of Reference Ontologies for Efficient LOD Management. In: Semi-Automatic Ontology Development: Processes and Resources, February 2012
</li>
<li><a href="http://ceur-ws.org/Vol-858/ore2012_paper14.pdf">OWLIM Reasoning over FactForge</a>. Proceedings of OWL Reasoner Evaluation Workshop (ORE'2012), collocated with IJCAR 2012, Manchester, UK
</li>
</ul>
</div>
</div>
<div id="outline-container-6-" class="outline-2">
<h2 id="sec-6-">DBpedia</h2>
<div class="outline-text-2" id="text-6-">
<p><a href="http://dbpedia.org"><img src="./img/dbpedia_logo.png" alt="dbpedia_logo.png"/></a><br/>
DBpedia is the central LOD dataset, serving as a source of stable URLs and a "crystallization point" for other LD.
</p><ul>
<li>English version: 4M things, 3.22M classified in a consistent ontology:
<ul>
<li>832k persons
</li>
<li>209k organizations (49k companies, 45k educational institutions)
</li>
<li>639k places (427k populated places)
</li>
<li>372k creative works (116k music albums, 78k films, 18.5k video games)
</li>
<li>226k species, 5.6k diseases
</li>
</ul>
</li>
<li>Localized versions (119 languages): 24.9M things
<ul>
<li>16.8M overlap (interlinked) with English Dbpedia, 8.1M (32%) are new
</li>
</ul>
</li>
<li>12.6M unique things with labels and abstracts
</li>
<li>24.6M links to images
</li>
<li>27.6M links to external web pages, 45M links into other RDF datasets
</li>
<li>67M links to Wikipedia categories, 41.2M to YAGO categories
</li>
<li>2.46B triples
<ul>
<li>470M from English Wikipedia
</li>
<li>1.98B from other language editions
</li>
<li>45M from links to external datasets
</li>
</ul>
</li>
<li><a href="http://wiki.dbpedia.org/Datasets39/DatasetStatistics?v=dqp">Detailed statistics</a>
</li>
</ul>
</div>
</div>
<div id="outline-container-7-" class="outline-2">
<h2 id="sec-7-">DBpedia English-German example</h2>
<div class="outline-text-2" id="text-7-">
<p><a href="http://dbpedia.org/resource/Wuppertal_Institute_for_Climate,_Environment_and_Energy">http://dbpedia.org/resource/Wuppertal_Institute_for_Climate,_Environment_and_Energy</a>
</p><table border=0 cellspacing=0 cellpadding=0 rules=all frame=box>
<caption></caption>
<colgroup><col class="left" /><col class="left" /><col class="left" />
</colgroup>
<thead>
<tr><th scope="col" class="left"><b>prop</b></th><th scope="col" class="left"><b>object</b></th><th scope="col" class="left"><b>comment</b></th></tr>
</thead>
<tbody>
<tr><td class="left">dbpprop:leitung</td><td class="left">dbpedia:Brigitte_Mutert-Breidbach</td><td class="left">resource (has own data)</td></tr>
<tr><td class="left"></td><td class="left">President</td><td class="left">say what?</td></tr>
<tr><td class="left"></td><td class="left">Vice President</td><td class="left">say what?</td></tr>
<tr><td class="left"></td><td class="left">Business Manager</td><td class="left">say what?</td></tr>
<tr><td class="left"></td><td class="left">Prof. Dr. Manfred Fischedick</td><td class="left">string (no own data)</td></tr>
<tr><td class="left"></td><td class="left">Prof. Dr. Uwe Schneidewind</td><td class="left">string (no own data)</td></tr>
<tr><td class="left">owl:sameAs</td><td class="left"><a href="http://de.dbpedia.org/resource/Wuppertal_Institut_fÐ âСr_Klima,_Umwelt,_Energie" width=50>http://de.dbpedia.org/resource/Wuppertal_Institut_fÐ âСr_Klima,_Umwelt,_Energie</a></td><td class="left">identity link to DE resource</td></tr>
</tbody>
</table>
To use owl:sameAs effectively, must use a powerful repo (eg OWLIM)
<ul>
<li>Semantics: "smush" the two resources, so they obtain the same statements
</li>
<li><a href="http://www.ontotext.com/owlim/owl-sameas-optimisation">OWLIM sameAs optimization</a>
</li>
</ul>
<p>
<a href="http://de.dbpedia.org/resource/Wuppertal_Institut_fÐ âСr_Klima,_Umwelt,_Energie">http://de.dbpedia.org/resource/Wuppertal_Institut_fÐ âСr_Klima,_Umwelt,_Energie</a>
</p><ul>
<li>(click "Back to old DBpedia" to see all data, this "NEW DBpedia" is no good)
</li>
</ul>
<table border=0 cellspacing=0 cellpadding=0 rules=all frame=box>
<caption></caption>
<colgroup><col class="left" /><col class="left" /><col class="left" />
</colgroup>
<tbody>
<tr><td class="left"><b>prop</b></td><td class="left"><b>object</b></td><td class="left"><b>comment</b></td></tr>
<tr><td class="left">dbpedia-owl:individualisedGnd</td><td class="left">2133644-1</td><td class="left">Nice, link to DNB GND</td></tr>
<tr><td class="left">dbpedia-owl:leader</td><td class="left">dbpedia-de:Uwe_Schneidewind</td><td class="left">Ok, all 3 are resources</td></tr>
<tr><td class="left"></td><td class="left">dbpedia-de:Manfred_Fischedick</td><td class="left"></td></tr>
<tr><td class="left"></td><td class="left">dbpedia-de:Brigitte_Mutert-Breidbach</td><td class="left"></td></tr>
<tr><td class="left">prop-de:leitung</td><td class="left">* Uwe Schneidewind * Manfred Fischedick * Brigitte Mutert-Breidbach</td><td class="left">All 3 in one string??</td></tr>
<tr><td class="left">owl:sameAs</td><td class="left">dbpedia-de:Wuppertal Institut fÐ âСr Klima, Umwelt, Energie</td><td class="left">same as itself? Very useful ;-)</td></tr>
</tbody>
</table>
</div>
</div>
<div id="outline-container-8-" class="outline-2">
<h2 id="sec-8-">FreeBase</h2>
<div class="outline-text-2" id="text-8-">
<p><a href="https://www.freebase.com/"><img src="./img/freebase-logo.png" alt="freebase-logo.png"/></a>
</p><ul>
<li>43.7M "topics": like WikiPedia pages or DBpedia resources,
typically lists of things (albums of a band, battles of a war, ingredients of a recipe…). 2.5B facts (triples)
</li>
<li>3.5x bigger coverage than English DBpedia, 2.3x bigger than all DBpedias
</li>
<li>Human-curated fact creation, so supposedly higher quality than DBpedia:<br/>
(WikiPedia is curated, DBpedia <b>mappings</b> are curated, but DBpedia is automatically extracted)
</li>
<li>Doesn't use RDFS modeling (e.g. no rdf:type). We've done significant work in mapping to RDFS and PROTON
</li>
</ul>
<p><img src="./img/freebase-numbers.png" alt="./img/freebase-numbers.png" />
</p></div>
</div>
<div id="outline-container-9-" class="outline-2">
<h2 id="sec-9-">GeoNames</h2>
<div class="outline-text-2" id="text-9-">
<p><a href="http://www.geonames.org/"><img src="./img/globe.gif" alt="globe.gif"/></a> <b>GeoNames</b>
</p><ul>
<li>8M place names (compare to 0.64M English DBpedia, estimated 1M all DBpedias, 1M FreeBase)
</li>
<li>Includes place names and alternate names in many languages
</li>
<li>Includes administrative hierarchy
</li>
<li>Includes coordinates (points only)
</li>
<li>I expect that most are interlinked to DBpedia
</li>
<li>Covers all countries, and a rich range of features
</li>
</ul>
<p><img src="./img/geonames-feature-counts.png" alt="./img/geonames-feature-counts.png" />
</p></div>
</div>
<div id="outline-container-10-" class="outline-2">
<h2 id="sec-10-">New York Times, CIA World FactBook</h2>
<div class="outline-text-2" id="text-10-">
<p><a href="http://data.nytimes.com">New York Times data</a>
</p><ul>
<li>5k people, 3k organizations, 2k locations, 0.5k concepts
</li>
<li>Other factual data
</li>
<li>Strong US bias
</li>
<li>(?) Explore deeper for useful News patterns/ontologies?
</li>
</ul>
<p><a href="https://www.cia.gov/library/publications/the-world-factbook/">CIA World FactBook</a>
</p><ul>
<li>General geopolitical data: history, people, government, economy, geography, communications, transportation, military, and transnational
</li>
<li>Covers most countries: 267 world entities
</li>
<li>Don't know whether the LD is kept up to date
</li>
</ul>
<p>Much smaller but still important resources
</p></div>
</div>
<div id="outline-container-11-" class="outline-2">
<h2 id="sec-11-">Possible Changes/Additions to FactForge</h2>
<div class="outline-text-2" id="text-11-">
<p>Although FactForge integrates the above described datasets, it has some disadvantages
</p><ul>
<li>Most importantly, it has not been updated in the last couple of years
</li>
<li>Eg DBpedia Live mirrors the minutely changes of Wikipedia.<br/>
But this is hard to implement, since we make corrections and mappings on top of DBpedia
</li>
<li>Potentially it can be extended with new datasets, if needed by the project:<br/>
OpenStreetMap/LinkedGeoData, WikiData, and Linguistic LD
</li>
</ul>
<p><a href="http://basekb.com/">BaseKB Lime</a> instead of FreeBase
</p><ul>
<li>Removes some 13M erroneous and 100M redundant triples
</li>
<li>Resilient towards errors in NTriples format: skips the line instead of rejecting the whole file ("parallelSuperEyeball" module)
</li>
<li><a href="https://github.com/paulhoule/infovore/wiki">Infovore Framework</a>: allows to transform a more recent Freebase RDF dump (Last one is from 201303)
</li>
</ul>
</div>
</div>
<div id="outline-container-12-" class="outline-2">
<h2 id="sec-12-">OpenStreetMap and LinkedGeoData</h2>
<div class="outline-text-2" id="text-12-">
<p><a href="http://www.openstreetmap.org"><img src="./img/openstreetmap.png" alt="openstreetmap.png"/></a> OpenStreetMap
</p><ul>
<li>An open manually curated map of the world
</li>
<li>Thousands of types of features, ways and areas; from the monumental to the micro scale:<br/>
continents, oceans, mountains,<br/>
cities, monuments, power stations,<br/>
roads, power lines, rivers, parks, areas,<br/>
pharmacies, stores, phone boxes, recycling bins…
</li>
<li>Integration of GPS tracks and easy editors for adding annotations
</li>
</ul>
<p><a href="http://linkedgeodata.org" width=50><img src="./img/linkedgeodata.png" alt="linkedgeodata.png"/></a>
</p><ul>
<li>Conversion of OSM to RDF: 15B triples
</li>
<li>Links to DBpedia, GeoNames, UN FAO Geopolitical data
</li>
<li>Last updated: just now (13 May 2014)
</li>
</ul>
</div>
</div>
<div id="outline-container-13-" class="outline-2">
<h2 id="sec-13-">WikiData</h2>
<div class="outline-text-2" id="text-13-">
<p><a href="http://www.wikidata.org/"><img src="./img/wikidata.png" alt="wikidata.png"/></a>
</p><ul>
<li>Provides atomic facts taht can be referenced in various WikiMedia sites
</li>
<li>Thus it is quite amenable to RDF use
</li>
<li>However, the data is still in its inception
</li>
<li>Furthermore, the facts can presumably be used from the WikiMedia sites, eg DBpedia
</li>
</ul>
</div>
</div>
<div id="outline-container-14-" class="outline-2">
<h2 id="sec-14-">Use-case specific linked data</h2>
<div class="outline-text-2" id="text-14-">
<p>For the specific use cases, we'll need to assemble specific datasets
</p><ul>
<li>The depth, breadth, content interlinking etc will be dictated by the needs of the use cases
</li>
</ul>
<p>We can elaborate, assemble and complement this only together!
</p><ul>
<li>Both to be goal-driven (what data do you need)
</li>
<li>And to find the data (what data you have stumbled upon)
</li>
<li>Ontotext will assess the data found and look at interlinking (also related to T4.3 alignment)
</li>
<li>Maybe some "crowd-sourcing" will be needed (eg DBpedia subsets)
</li>
</ul>
</div>
</div>
<div id="outline-container-15-" class="outline-2">
<h2 id="sec-15-">Energy Simulation</h2>
<div class="outline-text-2" id="text-15-">
<p>Identified all named entities <b>and concepts</b> in an Energy article by hand
</p><ul>
<li>Compared with REEGLE Tagging API (automatic annotation from a thesaurus dedicated to energy) and DBpedia Spotlight
</li>
<li>Shows the manual annotation is much richer & precise. Would require very powerful tools. Probably overly ambitious: this is a "programme maximum"
</li>
<li>Once we decide what is a feasible set of features to extract, we could write up Annotation Guidelines that can be used by manual annotators to build a Gold Standard corpus to be used for Machine Learning. What do you think (?)
</li>
</ul>
<p>Article: "TheGuardian_Goodbye nuclear power.docx"
</p><ul>
<li>Goodbye nuclear power: Germany's renewable energy revolution
</li>
<li>Tim Smedley - Guardian Professional, Friday 10 May 2013 17.58 BST
</li>
</ul>
<p><img src="./img/energy-simulation.png" alt="./img/energy-simulation.png" />
</p></div>
</div>
<div id="outline-container-16-" class="outline-2">
<h2 id="sec-16-">Simulation: Article Metadata/Attribution/Credits</h2>
<div class="outline-text-2" id="text-16-">
<p>"Standard" metadata, which doesn't mean it will be easy to extract.
</p><ul>
<li>DCTerms provides properties for a lot of these
</li>
<li>Added MARC Relators, esp. where there's no DCTerms field
</li>
</ul>
<table border=0 cellspacing=0 cellpadding=0 rules=all frame=box>
<caption></caption>
<colgroup><col class="left" /><col class="left" /><col class="left" />
</colgroup>
<tbody>
<tr><td class="left"><b>field</b></td><td class="left"><b>value</b></td><td class="left"><b>URL/canonic value</b></td></tr>
<tr><td class="left">creator</td><td class="left">Tim Smedley</td><td class="left"></td></tr>
<tr><td class="left">creator.affiliation</td><td class="left">Guardian Professional</td><td class="left"></td></tr>
<tr><td class="left">creator.role</td><td class="left">Author</td><td class="left"><a href="http://id.loc.gov/vocabulary/relators/aut">http://id.loc.gov/vocabulary/relators/aut</a></td></tr>
<tr><td class="left">created</td><td class="left">Friday 10 May 2013 17.58 BST</td><td class="left">"20130510T17:58:00"^^xsd:dateTime</td></tr>
<tr><td class="left">contributor</td><td class="left">Sean Gallup</td><td class="left"></td></tr>
<tr><td class="left">contributor.affiliation</td><td class="left">Getty Images</td><td class="left"></td></tr>
<tr><td class="left">contributor.role</td><td class="left">Photograph by</td><td class="left"><a href="http://id.loc.gov/vocabulary/relators/pht">http://id.loc.gov/vocabulary/relators/pht</a></td></tr>
<tr><td class="left">contributor</td><td class="left">Accenture</td><td class="left"></td></tr>
<tr><td class="left">contributor.role</td><td class="left">paid for by</td><td class="left"></td></tr>
<tr><td class="left">publisher</td><td class="left">Guardian Sustainable Business</td><td class="left"></td></tr>
<tr><td class="left">editorial</td><td class="left">The Guardian</td><td class="left"></td></tr>
<tr><td class="left">editorial.role</td><td class="left">controlled and overseen by (responsible party)</td><td class="left"><a href="http://id.loc.gov/vocabulary/relators/rpy">http://id.loc.gov/vocabulary/relators/rpy</a></td></tr>
<tr><td class="left">rights-holder</td><td class="left">Guardian News and Media Limited</td><td class="left"></td></tr>
<tr><td class="left">rights-holder.role</td><td class="left">copyright holder</td><td class="left"><a href="http://id.loc.gov/vocabulary/relators/cph">http://id.loc.gov/vocabulary/relators/cph</a></td></tr>
<tr><td class="left">copyright date</td><td class="left">2014</td><td class="left">"2014"^^xsd:gYear</td></tr>
</tbody>
</table>
</div>
</div>
<div id="outline-container-17-" class="outline-2">
<h2 id="sec-17-">Energy Simulation: Concepts (1)</h2>
<div class="outline-text-2" id="text-17-">
<ul>
<li>altLabels separated with ";". See "DBpedia Subsets re Energy" later on.
</li>
</ul>
<p>Generation, Transfer
</p><ul>
<li>nuclear power; nuclear capability (altLabel but only in the Energy context, else means nuclear weapons)
</li>
<li>renewable energy; renewables; Renewable Energy Sources; renewable-energy capacity
</li>
<li>wind power
</li>
<li>wind turbine
</li>
<li>wind farms; wind plants
</li>
<li>offshore wind farms; offshore wind plants
</li>
<li>intermittent wind energy
</li>
<li>biogas plant; plant [that] processes natural waste; biomass plants
</li>
<li>biomass facilities
</li>
<li>power lines, power cables
</li>
<li>electricity grid
</li>
<li>transportation [of energy]
</li>
<li>storage [of energy]
</li>
<li>micro-generation
</li>
<li>PV; photovoltaic
</li>
<li>photovoltaic plants
</li>
<li>co-generation; cogeneration
</li>
</ul>
</div>
</div>
<div id="outline-container-18-" class="outline-2">
<h2 id="sec-18-">Energy Simulation: Concepts (2)</h2>
<div class="outline-text-2" id="text-18-">
<p>Economics
</p><ul>
<li>renewable energy surcharge
</li>
<li>energy bill
</li>
<li>micro-ownership
</li>
<li>prosumer model; 'prosumer' model; prosumer (eg "the prosumer aspect")
</li>
<li>incentive systems
</li>
<li>ecological taxes
</li>
</ul>
<p>Other energy concepts
</p><ul>
<li>electricity
</li>
<li>heat
</li>
<li>environmentally-friendly
</li>
<li>insulation of buildings
</li>
<li>fertiliser
</li>
<li>greenhouse-gas (GHG) emissions
</li>
<li>energy consumption; consumption of energy
</li>
<li>consumer behaviour
</li>
<li>Energiewende, Energy Transformation, alternative energy transformation
</li>
<li>utility companies
</li>
<li>battery
</li>
</ul>
<p>Other
</p><ul>
<li>local farms
</li>
<li>houses
</li>
<li>Herculean task; Herculean
</li>
</ul>
</div>
</div>
<div id="outline-container-19-" class="outline-2">
<h2 id="sec-19-">Energy Simulation: Events, Places, Orgs, Persons</h2>
<div class="outline-text-2" id="text-19-">
<p>Events (from thesaurus)
</p><ul>
<li>Fukushima disaster
<ul>
<li>Fukushima: in this case is event not place: can be recognized by: "July 2011 (only three months after …)"
</li>
</ul>
</li>
</ul>
<p>Places
</p><ul>
<li>Germany; Germany's; economic powerhouse of Europe (alias)
</li>
<li>Lower Saxony
</li>
<li>Ebendorf, Germany
</li>
</ul>
<p>Organizations
</p><ul>
<li>German government
</li>
<li>Wuppertal Institute; Wuppertal Institute for Climate, Environment and Energy
</li>
<li>Big Four energy companies
</li>
<li>RWE
</li>
</ul>
<p>Persons
</p><ul>
<li>Angela Merkel
</li>
<li>Professor Dr Manfred Fischedick
<ul>
<li>Fischedick: requires coreference resolution
</li>
</ul>
</li>
</ul>
<p>Roles/positions
</p><ul>
<li>Professor Dr
</li>
<li>vice president
</li>
<li>scientific adviser
</li>
</ul>
</div>
</div>
<div id="outline-container-20-" class="outline-2">
<h2 id="sec-20-">Energy Simulation: Quantified Events</h2>
<div class="outline-text-2" id="text-20-">
<p>Quantities & Dates from text. Would be hard to recognize, but if possible will enable some quite interesting quantitative analyses.
</p><table border=0 cellspacing=0 cellpadding=0 rules=all frame=box>
<caption></caption>
<colgroup><col class="left" /><col class="left" /><col class="left" /><col class="left" /><col class="left" />
</colgroup>
<tbody>
<tr><td class="left"><b>subject</b></td><td class="left"><b>action</b></td><td class="left"><b>quantity</b></td><td class="left"><b>object</b></td><td class="left"><b>date</b></td></tr>
<tr><td class="left">Nuclear power</td><td class="left">produces</td><td class="left">nearly 20%</td><td class="left">Germany's energy</td><td class="left">now</td></tr>
<tr><td class="left">German government</td><td class="left">vowed to</td><td class="left"></td><td class="left">shut down</td><td class="left">July 2011</td></tr>
<tr><td class="left">German government</td><td class="left">shut down</td><td class="left">nuclear capability</td><td class="left"></td><td class="left">within 10 years</td></tr>
<tr><td class="left">German government</td><td class="left">cut</td><td class="left">by 40%</td><td class="left">greenhouse-gas (GHG) emissions</td><td class="left">by 2020</td></tr>
<tr><td class="left">German government</td><td class="left">cut</td><td class="left">and 80%</td><td class="left">greenhouse-gas (GHG) emissions</td><td class="left">by 2050</td></tr>
<tr><td class="left">renewables</td><td class="left">contribute</td><td class="left">80%</td><td class="left">Germany's energy</td><td class="left">by 2050</td></tr>
<tr><td class="left"></td><td class="left">drop</td><td class="left">20%</td><td class="left">energy consumption</td><td class="left">by 2020</td></tr>
<tr><td class="left"></td><td class="left">drop</td><td class="left">50%</td><td class="left">energy consumption</td><td class="left">by 2050</td></tr>
<tr><td class="left"></td><td class="left">Discussions about</td><td class="left"></td><td class="left">Energiewende</td><td class="left">started already in the 1980s</td></tr>
<tr><td class="left">renewable energy surcharge</td><td class="left">increase</td><td class="left">by 47%</td><td class="left">average family's energy bill</td><td class="left">in the past two years</td></tr>
<tr><td class="left">wind power</td><td class="left">contribute</td><td class="left">half of the 80%</td><td class="left">renewable energy target</td><td class="left">by 2050</td></tr>
<tr><td class="left">individuals or farmers in Germany</td><td class="left">INV is owned by</td><td class="left">over 50%</td><td class="left">renewable-energy capacity</td><td class="left"></td></tr>
<tr><td class="left">Big Four energy companies</td><td class="left">own</td><td class="left">just 6.5%</td><td class="left">renewable-energy capacity</td><td class="left">(according to 2010 figures)</td></tr>
<tr><td class="left"></td><td class="left">Reducing</td><td class="left">by half</td><td class="left">energy consumption</td><td class="left"></td></tr>
<tr><td class="left"></td><td class="left">discuss</td><td class="left"></td><td class="left">new forms of incentive systems</td><td class="left">in the next couple of years</td></tr>
<tr><td class="left">economic powerhouse of Europe</td><td class="left">decommission</td><td class="left">17</td><td class="left">nuclear power plants</td><td class="left">seven years left</td></tr>
<tr><td class="left">economic powerhouse of Europe</td><td class="left">cut</td><td class="left">by 40%</td><td class="left">GHG emissions</td><td class="left">seven years left</td></tr>
<tr><td class="left">economic powerhouse of Europe</td><td class="left">cut</td><td class="left">by 20%</td><td class="left">energy consumption</td><td class="left">seven years left</td></tr>
</tbody>
</table>
Annotations for such event networks can be visualized and manually created by <a href="http://brat.nlplab.org/">Brat Rapid Annotation Tool</a> (BRAT)
</div>
</div>
<div id="outline-container-21-" class="outline-2">
<h2 id="sec-21-">Compare with REEGLE Tagging API</h2>
<div class="outline-text-2" id="text-21-">
<p>How much can be recognized with a single glossary dedicated to energy (REEGLE)?
</p><ul>
<li>Paste the article text at <a href="http://api.reegle.info/service/try">REEGLE Tagging API</a> and compare the results to these above.<br/>
</li>
<li>Extracted Concepts: keywords from REEGLE thesaurus. Semantic info available, including synonyms, related concepts, higher level concepts
</li>
<li>Terms: Plain keywords extracted using statistical algorithms (frequency, position in text)
<table border=0 cellspacing=0 cellpadding=0 rules=all frame=box>
<caption></caption>
<colgroup><col class="left" /><col class="right" /><col class="left" /><col class="right" />
</colgroup>
<tbody>
<tr><td class="left"><b>Concept</b></td><td class="right"><b>Score</b></td><td class="left"><b>Term</b></td><td class="right"><b>Score</b></td></tr>
<tr><td class="left">energy</td><td class="right">74</td><td class="left">fischedick</td><td class="right">41</td></tr>
<tr><td class="left">wind</td><td class="right">32</td><td class="left">nuclear</td><td class="right">36</td></tr>
<tr><td class="left">renewable energies</td><td class="right">20</td><td class="left">germany's</td><td class="right">35</td></tr>
<tr><td class="left">electricity generation</td><td class="right">14</td><td class="left">energiewende</td><td class="right">28</td></tr>
<tr><td class="left">wind farms</td><td class="right">10</td><td class="left">germany</td><td class="right">22</td></tr>
<tr><td class="left">wind turbines</td><td class="right">9</td><td class="left">farms</td><td class="right">19</td></tr>
<tr><td class="left">emissions</td><td class="right">9</td></tr>
<tr><td class="left">biogas</td><td class="right">9</td></tr>
<tr><td class="left">biomass</td><td class="right">9</td></tr>
<tr><td class="left">natural disasters</td><td class="right">9</td></tr>
</tbody>
</table>
</li>
<li>Locations: Cities/countries, include latitude / longitude<br/>
<table border=0 cellspacing=0 cellpadding=0 rules=all frame=box>
<caption></caption>
<colgroup><col class="left" />
</colgroup>
<tbody>
<tr><td class="left">DE : Federal Republic of Germany (Country)</td></tr>
</tbody>
</table>
</li>
</ul>
<p>Conclusion: REEGLE covers perhaps 45% of the concepts and 1% of the entities.
</p><ul>
<li>Doesn't know specifics. Eg it sees the word "nuclear" is used often in the article, but doesn't know about nuclear energy
</li>
</ul>
</div>
</div>
<div id="outline-container-22-" class="outline-2">
<h2 id="sec-22-">DBpedia Spotlight Annotate Service</h2>
<div class="outline-text-2" id="text-22-">
<p><a href="https://github.com/dbpedia-spotlight/dbpedia-spotlight/wiki"><img src="./img/dbpedia-spotlight.jpg" alt="dbpedia-spotlight.jpg"/></a><br/>
<a href="#Https-github.com-dbpedia-spotlight-dbpedia-spotlight-wiki-Web-service">Web services</a>
</p><ul>
<li><b>/annotate</b>: just the top candidates. Supported formats:
<ul>
<li>HTML: text/html
</li>
<li>XML: text/xml (or no accept header; application/xml returns empty file)
</li>
<li>JSON: application/json
</li>
<li>RDFa: don't know how to invoke this since RDFa by itself doesn't have a mime type
</li>
<li>NIF: application/rdf+xml. This comes to 3Mb since it uses the full text as part of every URL and is therefore unusable: the text must be deployed on a web server before this format can be used
</li>
</ul>
</li>
<li><b>/candidates</b>: returns more data about potential enrichments (with score)<br/>
</li>
</ul>
<p>Annotations below are obtained with:
</p><pre class="example">
curl -H accept:text/html http://spotlight.dbpedia.org/rest/annotate
--data-urlencode text@EN_Guardian_nuclear_power.txt > Spotlight-Guardian.htm
</pre>
</div>
</div>
<div id="outline-container-23-" class="outline-2">
<h2 id="sec-23-">Compare with DBpedia Spotlight</h2>
<div class="outline-text-2" id="text-23-">
<p><img src="./img/Spotlight-performance.png" alt="./img/Spotlight-performance.png" />
</p></div>
</div>
<div id="outline-container-24-" class="outline-2">
<h2 id="sec-24-">Energy Datasets</h2>
<div class="outline-text-2" id="text-24-">
<p>Keeping the Energy article simulation in mind, what are possible sources for those entities and concepts?<br/>
</p>
<p>
Named entities:
</p><ul>
<li>I think we'll cover quite a lot of the domain of interest by using datasets already integrated in FactForge: DBPedia, GeoNames, FreeBase and NY Times
</li>
<li>There are better sources for energy-specific entities (e.g. Power Plants)
</li>
</ul>
<p>Concepts:
</p><ul>
<li>Below are examples from various thesauri
</li>
<li>When a resource is present in several thesauri, we can often find interlinks (alignments), and then enlarge the scope by exploring around the hierarchy.
</li>
<li>A potential problem in general theasuri is coverage: I don't know how much of the desired terms are covered, so we may have to pick and choose from various thesauri
</li>
</ul>
<p>DBPedia
</p><ul>
<li>DBpedia provides both entities and concepts/categories.
</li>
<li>Related concepts can be extracted by links navigation and using the Categories (see below)
</li>
<li>Person: <a href="http://de.wikipedia.org/wiki/Manfred_Fischedick">http://de.wikipedia.org/wiki/Manfred_Fischedick</a> (related to Wuppertal Institute)
</li>
<li>Event: <a href="http://en.wikipedia.org/wiki/Fukushima_Daiichi_nuclear_disaster">http://en.wikipedia.org/wiki/Fukushima_Daiichi_nuclear_disaster</a>.<br/>
<ul>
<li><a href="http://en.wikipedia.org/wiki/Fukushima">http://en.wikipedia.org/wiki/Fukushima</a> is a disambiguation (redirection) page that leads to the above (amongst other things). Thus redirection pages provide synonyms (altLabels)
</li>
</ul>
</li>
</ul>
</div>
</div>
<div id="outline-container-25-" class="outline-2">
<h2 id="sec-25-">General Thesauri</h2>
<div class="outline-text-2" id="text-25-">
<p>Example concept: "Wind Energy"
</p><ul>