Knowledge Graphs

1. 00:00Welcome to knowledge graphs. This is lecture number six Advanced Knowledge Graph Applications.
2. 00:06As always we start with the lecture overview.
3. 00:09So what we are going to do in this lecture is of course we
4. 00:13want to look at more advanced applications. First of all we
5. 00:17want to look deeper into knowledge graphs and to understand you know
6. 00:21how to characterize a knowledge graph. So how can we make differences,
7. 00:26how can we say this knowledge graph should be preferred above another.
8. 00:30And there the first thing we look into it is the graph of the
9. 00:34knowledge graph. This means we are looking for structural properties
10. 00:38of that and try to compare. So this graph in the knowledge graphs
11. 00:43can actually be of huge size. In order to approximate that we
12. 00:47will start looking at the knowledge graph embeddings.
13. 00:52And then we are going to use this knowledge graph embeddings
14. 00:55for knowledge graph completion because it is not necessary that
15. 00:59knowledge graphs can capture all the information, all the knowledge in the universe.
16. 01:05So in these two lectures you will only deal with one knowledge
17. 01:10graph but what if we have more than one knowledge graph where
18. 01:14which is depicting the similar information?
19. 01:18However it is complementary to each other. So in order to deal
20. 01:22with such kind of problems we will talk about knowledge graph mappings and alignment.
21. 01:27And in the last sections of the lecture we will talk about
22. 01:31information systems and how knowledge graph can support these
23. 01:34kind of information systems. As first we are going to talk about
24. 01:38semantic search. This is the extension of traditional search
25. 01:42with knowledge graph. So you know this already, nowadays search
26. 01:46engines always are based on knowledge graphs and in semantic
27. 01:49search you will see how this works under the hood.
28. 01:53And then we come up in the very last chapter with something
29. 01:57which is still ongoing research and this is so called exploratory research and
30. 02:03then of course recommender systems. You all know recommender
31. 02:05systems. But exploratory search is a generalization of pure recommendation. This means
32. 02:11what if I not exactly know, if I don't exactly know how
33. 02:16to phrase my search? If I try to search in a complete unknown domain
34. 02:22where I had first to gather knowledge to be able to search? Then
35. 02:26often I want to be guided through the search space and
36. 02:30this kind of exploration then is referred to often as exploratory
37. 02:34search and we will talk about that in the last part of the lecture.
38. 02:40But first let's talk about the graph in knowledge graphs.

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