Semantic Knowledge Bases from Web Sources thumbnail
Pause
Mute
Subtitles
Playback speed
0.25
0.5
0.75
1
1.25
1.5
1.75
2
Full screen

Chapter list

Outline for Part IV - 100:00
Outline for Part V - 100:00
Semantic Knowledge Basesfrom Web Sources00:00
Outline for Part II - 100:07
Outline for Part III00:17
RDFS-Ontologies00:34
There’s not only DBpedia& YAGO00:47
There’s a whole Web of Ontologies00:57
SPARQL01:26
Two Paradigms in Information Extraction (IE)01:56
SPARQL – Example - 102:01
All slides for download…02:03
URIs02:13
Outline - 102:32
URL- like URIs - 102:36
SPARQL – Example - 202:38
Labels03:02
SOFIE: Facts & Patterns Consistency - 103:42
SPARQL – More Features03:47
Entities & Classes04:05
Classes04:30
URL- like URIs - 204:44
Goal: Turn Web into Knowledge Base04:46
SPARQL: Extensions from W3C05:23
Binary Relations05:35
Entailment06:01
URL- like URIs - 306:18
SPARQL: Extensions from Research (1)06:29
Higher-arity Relations & Reasoning06:38
Approach: Knowledge Harvesting07:13
SOFIE: Facts & Patterns Consistency - 207:16
SPARQL: Extensions from Research (2) - 107:33
SOFIE Example07:45
Relations07:59
Outline for Part III08:22
SPARQL: Extensions from Research (2) - 208:25
Unary Relations08:29
WordNet Thesaurus - 109:21
Triples with URIs09:28
RDF+SPARQL: Systems09:29
WordNet Thesaurus - 210:06
Outline for Part IV - 210:17
OWL10:18
WordNet Thesaurus - 310:32
Why ranking is essential10:33
Knowledge for Intelligence10:43
Tapping on Wikipedia Categories - 111:17
Namespace Prefixes11:17
Soft Rules vs. Hard Constraints11:41
Extending Entities with Keywords12:02
Tapping on Wikipedia Categories - 212:03
Storing data12:20
Tapping on Wikipedia Categories - 312:27
OWL Undecidability12:35
Mapping: Wikipedia → WordNet - 112:45
Digression 1: Graph Authority Measures13:21
OWL-DL13:24
Mapping: Wikipedia → WordNet - 213:28
Cool URIs - 113:49
Application 1: Semantic Queries on Web - 114:13
Event Entities14:42
Pattern Harvesting, Revisited15:05
Keyword-Based Entity Search: Principles15:19
Mapping: Wikipedia → WordNet - 315:31
Digression 2: Language Models (LMs)15:50
Reification16:32
Mapping: Wikipedia → WordNet - 416:49
Language Models for Text: Example17:17
RDFS: Summary17:29
Cool URIs - 217:40
Language Models for Text: Smoothing17:49
We’re all one Graph18:10
PROSPERA Architecture18:10
YAGO Concept Mappings18:12
Outline for Part II - 218:12
Application 1: Semantic Queries on Web - 218:15
YAGO Consistency Checks18:37
Outline - 218:42
Cyc18:42
Application 1: Faceted Search18:42
Entity Search with LM Ranking18:43
Standard Vocabulary18:51
Application 2: Deep QA in NL19:09
Goal: Comprehensive & Consistent - 119:12
Cyc: Language19:24
Goal: Comprehensive & Consistent - 219:28
Dublin Core19:44
Learning More Mappings19:49
Outline for Part IV - 420:15
Trivially Parallel: Pattern Mining20:20
What makes a fact „good“?20:32
Creative Commons20:39
Cyc: Example of Content20:53
Long Tail of Class Instances - 121:27
Harder to Parallelize: Consistency Reasoning21:40
Long Tail of Class Instances - 222:14
Application 3: Machine Reading22:18
Cyc: Summary22:20
Schema.org23:06
WordNet23:35
PROSPERA Results23:41
WordNet: Content24:11
Outline - 324:16
Linked Data Problem24:22
Entity Disambiguation24:39
How can we implement this?24:43
Named-Entity Disambiguation24:43
Outline for Part III - 525:06
Linked Data Solution25:28
Open-Domain IE, History25:34
WordNet: Semantic Relations25:34
Individual Entity Disambiguation26:04
LMs: From Entities to Facts26:28
WordNet: Summary26:30
Open-domain IE, Methodology26:38
Mentions, Meanings, Mappings27:10
Wikipedia27:29
Joint Disambiguation27:33
The Linking Data Project27:39
Wikipedia: Articles and Attributes27:52
ReadTheWeb - 128:00
LMs for Triples and Triple Patterns28:22
The Linked Data Cloud28:27
AIDA – Disambiguating Names in YAGO228:50
Mention-Entity Graph - 129:08
ReadTheWeb - 229:24
Wikipedia: Summary29:26
Mention-Entity Graph - 229:32
Features for Disambiguation29:54
Outline for Part II - 329:56
Knowledge Bases from Wikipedia30:01
Basic idea30:14
ReadTheWeb - 331:16
LMs for Composite Queries31:22
Objective Function31:44
WikiNet31:51
NELL: Never-Ending Language Learning - 131:57
Joint Disambiguation as Graph Problem32:15
Extensions: Keywords - 132:18
Existing Ontologies32:25
Graph Algorithm32:49
WikiNet: Summary33:42
NELL: Never-Ending Language Learning - 233:50
Mention-Entity Graph - 333:59
DBpedia - 134:07
Outline for Part V - 234:13
Outline for Part III - 234:14
And the rest of the Web?34:32
Mention-Entity Graph - 434:34
DBpedia - 234:45
NELL Example Output - 134:49
Mention-Entity Graph - 534:54
Joint Mapping35:09
Extensions: Keywords - 235:11
NELL Example Output - 235:18
Binary Relations – Which Sources to Pick?35:39
DBpedia - 336:06
TextRunner36:13
Microdata36:14
DBpedia - 436:33
Picking Low-Hanging Fruit (First)36:41
AIDA Accurate Online Disambiguation - 136:53
LMs for Keyword-Augmented Queries36:53
AIDA Accurate Online Disambiguation - 236:57
Application 4: Annotation of Web Data36:59
Creating an Entity37:00
DBpedia - 537:03
Deterministic Pattern Matching37:10
YAGO37:33
TextRunner Example Ouput - 138:00
Naming an Entity38:04
YAGO: Classes38:06
Wrapper Induction38:17
TextRunner Example Ouput - 238:40
Application 4: Map Annotation38:44
YAGO: Consistency Checks38:47
Tapping on Web Tables - 138:49
Extensions: Query Relaxation38:53
Item Properties39:01
Spectrum of Machine Knowledge - 139:07
Item Properties with URIs39:35
Omnivore39:47
YAGO: Annotations40:00
Tapping on Web Tables - 240:06
Inner Nodes - 140:20
Recovering the Semantics of Web Tables40:31
Outline for Part III - 640:34
Spectrum of Machine Knowledge - 240:44
Higher-arity Relations –Space & Time40:45
YAGO: Summary40:51
Inner Nodes - 241:11
Microdata Summary41:27
French Marriage Problem (Revisited)41:54
Extensions: Diversification42:14
Relational Fact Extraction From Plain Text - 142:20
Facebook & Annotated HTML42:35
Spectrum of Machine Knowledge - 342:43
Freebase - 143:02
Challenge: Temporal Knowledge Harvesting43:42
Freebase - 244:09
Outline for Part IV - 444:16
Search Engines & Annotated HTML44:19
Relational Fact Extraction From Plain Text - 244:51
What we have seen so far45:08
This Tutorial45:28
Readings for Part I45:32
Difficult Dating45:33
DIPRE/Snowball45:37
Outline - 245:41
Other Query Interfaces45:50
Freebase: Community45:52
Implicit Dating - 146:11
Freebase: Summary46:38
Natural Language Queries46:40
Implicit Dating - 246:59
Outline for Part V - 347:15
TARSQI: ExtractingTime Annotations47:36
DIPRE/Snowball/QXtract47:40
PowerAqua (Open University, UK)47:44
Help from NLP: Dependency Parsing - 148:04
References48:05
Outline - 448:09
13 Relations between Time Intervals48:14
Outline for Part II - 448:26
Read the Web/NELL - 148:38
Possible Worlds in Time - 148:47
Summary49:19
Example: Querix (Uni Zurich)49:21
Spectrum of Machine Knowledge (1)49:26
Help from NLP: Dependency Parsing - 250:01
Help from NLP: Dependency Parsing - 350:23
Open-Domain Gathering of Facts51:04
Natural Language Queries51:13
Possible Worlds in Time - 251:52
Faceted Search51:56
Read the Web/NELL - 252:15
Declarative Extraction Frameworks - 152:44
Outline for Part III - 652:49
Faceted Search: http://dpbedia.neofonie.de/ - 152:56
Declarative Extraction Frameworks - 253:18
Faceted Search: http://dpbedia.neofonie.de/ - 253:32
Multilingual Lexical Knowledge53:37
Faceted Search: http://dpbedia.neofonie.de/ - 353:43
Pattern-Based Harvesting Summary54:07
Outline for Part III - 354:32
Faceted Search: http://dpbedia.neofonie.de/ - 454:58
Applications for Sequence Labeling55:03
Faceted Search: http://dpbedia.neofonie.de/ - 555:18
Read the Web/NELL - 355:23
Probabilistic Extraction Models55:42
Faceted Search: http://dpbedia.neofonie.de/ - 655:52
Knowledge from Many Languages55:53
Faceted Search: http://dpbedia.neofonie.de/ - 756:02
Faceted Search56:11
Wolfram Alpha56:15
Probabilistic Models for Sequence Labeling 56:31
Open Problems and Challenges in IE (I)56:41
Wolfram Alpha: Content56:54
Hidden Markov Models – HMMs57:22
AutoSPARQL: Learning Queries from Examples57:30
True Knowledge58:12
Open Problems and Challenges in IE (II)58:24
HMMs: Inference & Learning59:00
Wolfram Alpha & TrueKnowledge59:20
Active Learning from Examples59:31
Visual Query Formulation59:56
Spectrum of Machine Knowledge (2)01:00:18
iSPARQL, http://dbpedia.org/isparql/ - 101:00:35
Maximum Entropy Markov Models – MEMMs - 101:01:00
iSPARQL, http://dbpedia.org/isparql/ - 201:01:05
Visual Query Formulation01:01:15
Which interface is best (for casual users)?01:01:35
Maximum Entropy Markov Models – MEMMs - 201:01:58
Spectrum of Machine Knowledge (3)01:02:01
Outline for Part II - 501:02:11
References for Part II01:02:17
Outline - 301:02:21
ImageNet: Visual WordNet - 101:02:23
Directed Models and Label Bias01:02:51
Outline for Part IV - 501:03:25
Open Problems and Challenges – Part IV01:03:29
ImageNet: Visual WordNet - 201:03:47
Photos of Entities in the Long Tail - 101:04:06
Conditional Random Fields – CRFs01:05:07
Photos of Entities in the Long Tail - 201:06:24
CRF Extensions01:06:53
Outline for Part III - 401:08:17
KB Building: Achievements & Challenges01:08:45
More Ontological Rigor01:09:08
French Marriage Problem - 101:10:07
French Marriage Problem - 201:11:30
Reasoning about Fact Candidates01:12:23
KB Applications: Achievements & Challenges01:12:38
Markov Logic Networks - 101:13:40
Markov Logic Networks - 201:13:40
Markov Logic Networks - 301:16:49
Grand Challenge: Web-Scale KB Construction - 101:17:57
Markov Logic Networks - 401:19:06
Grand Challenge: Web-Scale KB Construction - 201:19:31
Markov Logic Networks - 501:19:46
Overall Take-Home01:20:06
Related Alternative Probabilistic Models01:20:24
FactorIE01:21:12
Outline - 501:21:31
The End01:21:32
Thanks01:21:44
Bidirectional Joint Segmentation & Disambiguation01:23:03
SOFIE: Reasoning for KB Growth01:25:30