Successfully reported this slideshow.
We use your LinkedIn profile and activity data to personalize ads and to show you more relevant ads. You can change your ad preferences anytime.

KR in the age of Deep Learning

1,159 views

Published on

Updates an earlier talk and adds some new work trying to lay groundwork for a future world of "DL and Symbols" (as Gary Marcus calls it)

Published in: Technology
  • Be the first to comment

KR in the age of Deep Learning

  1. 1. Tetherless World Constellation, RPI KR in the age of Deep Learning, Watson, and the Semantic Web Jim Hendler Tetherless World Professor of Computer, Web and Cognitive Sciences Director, Institute for Data Exploration and Applications Rensselaer Polytechnic Institute http://www.cs.rpi.edu/~hendler @jahendler (twitter) Major talks at: http://www.slideshare.net/jahendler
  2. 2. Tetherless World Constellation, RPI Talk derives in part from a recent book (More info at Springer booth) (Thanks Alice!)
  3. 3. Tetherless World Constellation, RPI Outline • Several important AI technologies have moved through “knees in the curve” bringing much of the attention to AI again – Deep Learning (& ML in general) – Watson (& “cognitive computing”) – Semantic Web (& the knowledge graph) • But what about KR – What it is, why it still matters • And how can these come together – Which comes with a lot of important challenges
  4. 4. Tetherless World Constellation, RPI A) Deep Learning “phase transition” in capabilities of neural networks w/machine power
  5. 5. Tetherless World Constellation, RPI Trained on lots of categorized images Imagenet: Duck Imagenet: Cat
  6. 6. Tetherless World Constellation, RPI Impressive results Increasingly powerful techniques have yielded incredible results in the past few years
  7. 7. Tetherless World Constellation, RPI B) Watson
  8. 8. Tetherless World Constellation, RPI The Watson DeepQA Pipeline
  9. 9. Tetherless World Constellation, RPI Watson is based on ”Associative knowledge” © IBM, used with permission.
  10. 10. Tetherless World Constellation, RPI Impressive Results Watson showed the power of “associative knowledge”
  11. 11. Tetherless World Constellation, RPI C) Semantic Web
  12. 12. Tetherless World Constellation, RPI From Semantic Web to the Knowledge Graph
  13. 13. Tetherless World Constellation, RPI Based on a large “knowledge graph” mined from extracted and learned data
  14. 14. Tetherless World Constellation, RPI Impressive results Google finds embedded metadata on >30% of its crawl – Guha, 2015 Google “knowledge vault” reported to have over 1.6 billion “facts” (links)
  15. 15. Tetherless World Constellation, RPI Summary: AI has done some way cool stuff Summary (simplifying tremendously) • Deep Learning: neural learning from data with high quality, but imperfect results • Watson: Associative learning from data with high quality but imperfect results • Semantic Web/Knowledge Graph: Graph links formation from extraction, clustering and learning As much as many of us “GOFAI” folks wish it, this stuff cannot be ignored but, there are still problems…
  16. 16. Tetherless World Constellation, RPI Many intermediate steps (P. Norvig, WWW 2016, 4/16)
  17. 17. Tetherless World Constellation, RPI Why did knowledge graph need “”Human Judgments”? Association ≠ Correctness
  18. 18. Tetherless World Constellation, RPI GOFAI: Knowledge Representation? • A knowledge representation (KR) is most fundamentally a surrogate, a substitute for the thing itself, used to enable an entity to determine consequences by thinking rather than acting, i.e., by reasoning about the world rather than taking action in it. • It is a set of ontological commitments, i.e., an answer to the question: In what terms should I think about the world? • It is a fragmentary theory of intelligent reasoning, expressed in terms of three components: (i) the representation's fundamental conception of intelligent reasoning; (ii) the set of inferences the representation sanctions; and (iii) the set of inferences it recommends. • It is a medium for pragmatically efficient computation, i.e., the computational environment in which thinking is accomplished. One contribution to this pragmatic efficiency is supplied by the guidance a representation provides for organizing information so as to facilitate making the recommended inferences. • It is a medium of human expression, i.e., a language in which we say things about the world. R. Davis, H. Shrobe, P. Szolovits (1993)
  19. 19. Tetherless World Constellation, RPI KR: Human Expression Cute kid story: first two words
  20. 20. Tetherless World Constellation, RPI Telling cats from ducks doesn’t need KR !
  21. 21. Tetherless World Constellation, RPI “Saying things about the world” does "If I was telling it to a kid, I'd probably say something like 'the cat has fur and four legs and goes meow, the duck is a bird and it swims and goes quack’. " How would you explain the difference between a duck and a cat to a child? Woof
  22. 22. Tetherless World Constellation, RPI KR: Surrogate knowledge? Which could you sit in? What is most likely to bite what? Which one is most likely to become a computer scientist someday? …
  23. 23. Tetherless World Constellation, RPI “Surrogate” knowledge Which could you sit in? What is most likely to bite what? Which one is most likely to become a computer scientist someday? How would they go about doing it?
  24. 24. Tetherless World Constellation, RPI KR: Recommended vs. Possible inference Which one would you save if the house was on fire?
  25. 25. Tetherless World Constellation, RPI Recommended vs. Possible inference Which one would you save if the house was on fire? Would you use a robot baby-sitter without knowing which of the three possibilities it would choose?
  26. 26. Tetherless World Constellation, RPI KR systems in AI need grounded symbols • Logic- and rule- based systems – Ground in “model theory” with a notion of truth and falsity • Probabilistic Reasoning – P(A|B) requires A, B map to “meaningful” concepts, P to be a “real” probability • Constraint Satisfaction, etc – Finding an interpretation satisfying a set of boolean (T,F) constraints (Note: Yes, I am simplifying, blurring distinctions, ignoring much cutting edge work… happy to discuss later)
  27. 27. Tetherless World Constellation, RPI The challenge • If we want to implement KR systems on top of neural and associative learners we have an issue – The numbers coming out of Deep Learning and Associative graphs are not probabilities – They don’t necessarily ground in human-meaningful symbols • ”sub-symbolic” learning … • Association by clustering … • Errorful extraction …
  28. 28. Tetherless World Constellation, RPI The challenges • Can we avoid throwing out the reasoning baby with the grounding bathwater? – We still need planning systems – We still want to be able to define the rules that a system should follow – We want to be able to interact with and understand these systems • Even if computers don’t need to be symbolic communicators, WE DO!!!
  29. 29. Tetherless World Constellation, RPI Starting Place: Rethinking grounding – Formal Explanation vs. post hoc justification • Eg. Even if we cannot use a formal decomposition to explain the reasoning, can we produce a justification that explains it – Reasoning systems that “know” some of their axioms may be simply wrong • Eg.F1 of .9 doesn’t mean answers are 90% correct, it is (simplifying) more like 9 out of 10 answers are right, the others aren’t. – Nailing context …
  30. 30. Tetherless World Constellation, RPI Vision Challenge 1 What is the relationship between this man and this woman?
  31. 31. Tetherless World Constellation, RPI Vision Challenge 1 What is the relationship between this man and this woman? Deep learning produced Scene Graph w/relationships (Klawonn & Heims, 2018)
  32. 32. Tetherless World Constellation, RPI Vision and Knowledge Challenge What is the relationship between this man and this woman? Deep learning produced Scene Graph w/relationships (Klawonn, 2018)
  33. 33. Tetherless World Constellation, RPI Vision and Knowledge Challenge What is the relationship between this man and this woman? Deep learning produced Scene Graph w/relationships (Klawonn, 2018) Seeing the bride adds significant information that cannot be easily learned w/o background knowledge
  34. 34. Tetherless World Constellation, RPI This kind of context matters!
  35. 35. Tetherless World Constellation, RPI Human-Aware AI • Context is key – AI learning systems still perform best in well-defined contexts (or trained situations, or where their document set is complete, etc.) – Humans are good at recognizing context and deciding when extraneous factors don’t make sense • Or add extra “inferencing” (the bride example)
  36. 36. Tetherless World Constellation, RPI Learning inferences
  37. 37. Tetherless World Constellation, RPI In noisy data
  38. 38. Tetherless World Constellation, RPI Summary of talk (minus moose) • Modern AI is making some huge strides – Eg. DL, Associative Learning, Knowledge Graphs, … • But the need for KR has not gone away – Eg. Surrogacy, Recommended Inference, Human communication • The integration challenge will require rethinking some key AI ideas – Grounding, explanation, context …. • But we need to do it.
  39. 39. Tetherless World Constellation, RPI Questions?

×