Knowledge Graphs - Foundations and Applications

1. 00:00Welcome to Knowledge Graphs. This is Lecture Five Ontological Engineering for Smarter
2. 00:06Knowledge Graphs. We are Harald Sack, Sasha Bruns Mary Ann Tan and Tabea Tietz
3. 00:12as always. First we will give you a brief overview about the
4. 00:16content of this week's lecture.
5. 00:19As you might remember in the last week we have talked about our
6. 00:22web Ontology language A and how to create ontologies. Based on our
7. 00:27we have clearly seen that all has restrictions and specific limits
8. 00:32and about these limits we will talk in the very first part of this lecture
9. 00:36and we will show you how we will be able to go beyond these
10. 00:39limits because we are able to express even more semantic expressivity.
11. 00:46The drawback of course is if you go beyond then you pay of course again then
12. 00:52with in terms of complexity that means things quickly become undecidable. However
13. 00:58what we are going to do is we are going to show you rules. So
14. 01:01we are going to introduce also the Semantic Web rule language Swirl
15. 01:06And what we are doing here is we are leaving the real
16. 01:09of the open world assumption we have been so far and adding
17. 01:13a bit of the closed world assumption rule word that is known
18. 01:16from expert systems for example
19. 01:20Ok A and recall also our D I K W Pyramid from Knowledge Management.
20. 01:26So for us to be able to move up from information to knowledge
21. 01:31we have to infuse our information with semantics.
22. 01:34In order to do this we need an ontology
23. 01:38and in five point two we will tell you how to design your own ontology.
24. 01:43We will tell you the methodologies that we use
25. 01:47to do this in Five Point three. We then help you
26. 01:52to design better ontologies
27. 01:56and you will see it's not so easy to define. Let's say a good ontology.
28. 02:01It's also difficult to evaluate ontologies in that sense because
29. 02:05there are many criteria we have to fulfill.
30. 02:08So what we need one A sound methodologies that enable us to develop
31. 02:13better ontologies. And for that we are looking at Ontological Engineering. All these
32. 02:18methodologies that might help us to improve. Also, the automatic
33. 02:24and semi automatic creation of ontologies like for example,
34. 02:28not only ontological Design or also ontology learning or ontology, alignment
35. 02:34and stuff like that. And finally, after we have created a good ontology.
36. 02:41What's missing so far is then of course, yeah,
37. 02:44it's Data A and it's not data in that sense. Of course it's not
38. 02:48knowledge. So we have to populate the ontology. So and this is Knowledge Graph Construction
39. 02:53which we are dealing with in chapter Five Point
40. 02:57Five. Okay, to summarize in Five Point Six, we will tell you the best practices
41. 03:03in Ontological Engineering
42. 03:06to have something practical for your daily life.
43. 03:10Us always. At the end of each lecture we have prepared additional hands-on for you.
44. 03:15We start out with Knowledge Graph Construction from unstructured text
45. 03:20using natural language processing techniques such as named
46. 03:23entity recognition and relation extraction.
47. 03:26Then we will show you the new tool which is called Open Refine
48. 03:31which is really helpful when you have to transform structured
49. 03:34data like C three files into Rdf and it's also and also how to
50. 03:40connect your entities with entities. from Wiki data and additional hands on
51. 03:46is a swirl of one a Semantic Web rune language, whispered protocol ontology editor
52. 03:51where we will take a look how the rules can be helpful for deducing new knowledge.
53. 03:57A And so let's start and the fact First lecture is beyond the
54. 04:01limits of all.

To enable the transcript, please select a language in the video player settings menu.