Adding Neurons to your Assistants

Paris NLP Meetup #4
How to use Deep Learning and Neural Networks for chatbots Natural Language Processing

Transcript

1. Adding Neurons to Your Assistants Christophe Bourguignat Matthieu Bizien @zelrosAI @ParisNLP
2. What we want to solve at Zelros
3. Datasource Connection Predictive Modeling Dialogs & NLP Configuration AI Education INTELLIGENT VIRTUAL ASSISTANT PLATFORM
4. Understanding Natural Language Understanding Source : http://nlp.stanford.edu/~wcmac/papers/20140716-UNLU.pdf
5. Our playground Today : retrieval based systems - what works today in practice for conversational agents Given a user sentence and context, find the best answer among a pre-defined set of intents Tomorrow : generative models, self-learning
6. Bots are not born equal Success
7. Bots are not born equal Success Error
8. Bots are not born equal Success Error Fallback
9. A bit of history http://disi.unitn.it/~riccardi/papers/specom97.pdf
10. Things are going fast 01/15 09/16 11/16 12/16 06/16
11. Why we want to build our own NLU system More fun ! Data Privacy Performances for our use-cases Our own roadmap ...
12. BUNT, The first public benchmarker for NLU APIs https://github.com/zelros/bunt
13. Approach 1 : Supervised N-grams Intent Utterance X Y NAME What is your name? N-gram 1 NEED What do you need? N-gram 2 NEED Do you need anything? N-gram 2 ✔ Work in practice ❌ Out-Of-Vocabulary words? ❌ Fallback Detection?
14. Approach 2 : Word2Vec Step 1: Distance function What is your name Utterance 0.34 0.32 0.01 0.31 0.27 0.21 0.21 0.04 0.42 0.22 ... ... ... .. .. 0.5 0.12 0.5 0.44 0.41 What was your mood Sentence 0.34 0.12 0.01 0.21 0.17 0.21 0.51 0.04 0.52 0.32 ... ... ... .. .. 0.5 0.22 0.5 0.24 0.36 M E A N M E A N Cosine Similarity
15. Approach 2 : Word2Vec Step 2: Classifier Distance Function Nearest Neighbors + Best Intent ⚠ Work in practice ✔ Out-Of-Vocabulary words? ❌ Fallback Detection?
16. Going further: ML without data Distance Function Better algorithm Learn the Distance Function Fallback detection Nearest Neighbors + Best Intent
17. ML without data Available datasets: QA/QC 15 500 Questions Supervised : label = type of intent • 6 coarse classes : abbreviation, description, entity, human, location, numeric value • 50 fine classes Q A Q C Sentence Coarse class Fine class How did serfdom develop in and then leave Russia ? DESC manner What films featured the character Popeye Doyle ? ENTY cremat How can I find a list of celebrities ' real names ? DESC manner What is the full form of .com ? ABBR exp What contemptible scoundrel stole the cork from my lunch ? HUM ind https://cogcomp.cs.illinois.edu/Data/QA/QC/
18. ML without data Available datasets: SNLI 570k human-written English sentence pairs Label = entailment, contradiction, or neutral (judgments of five turkers) Text Judgments Hypothesis A man inspects the uniform of a figure in some East Asian country. contradiction C C C C C The man is sleeping An older and younger man smiling. neutral N N E N N Two men are smiling and laughing at the cats playing on the floor. A black race car starts up in front of a crowd of people. contradiction C C C C C A man is driving down a lonely road. A soccer game with multiple males playing. entailment E E E E E Some men are playing a sport. https://nlp.stanford.edu/projects/snli/
19. ML without data Available datasets: Quora Sentence 1 Duplicate Sentence 2 What is the step by step guide to invest in share market in india? False What is the step by step guide to invest in share market? Why am I mentally very lonely? How can I solve it? False Find the remainder when [math]23^{24}[/math] is divided by 24,23? How do we prepare for UPSC? True How do I prepare for civil service? How can I be a good geologist? True What should I do to be a great geologist? 400 000 pairs of questions Supervised : label = are they duplicates?
20. ML without data Available datasets: Quora
21. Siamese Network
22. Siamese Network Creation of a sentence embedding What is your name 0.34 0.32 0.01 0.31 0.21 0.21 0.04 0.42 ... ... ... .. 0.5 0.12 0.5 0.44 Embedding 0.27 0.22 .. 0.41 Lambda (lambda x: K.max(x, axis=1), output_shape=(300,)) TimeDistributed(Dense(300, activation=’relu’)) Embedding(nb_tokens+1, 300, input_length=25, trainable=False)
23. Siamese Network A simple architecture What is your name Embedding 0.34 0.32 0.01 0.31 0.27 0.21 0.21 0.04 0.42 0.22 ... ... ... .. .. 0.5 0.12 0.5 0.44 0.41 What was your mood Embedding 0.34 0.12 0.01 0.21 0.17 0.21 0.51 0.04 0.52 0.32 ... ... ... .. .. 0.5 0.22 0.5 0.24 0.36 Lambda (lambda x: K.max(x, axis=1), output_shape=(300,)) TimeDistributed(Dense(300, activation=’relu’)) Embedding(nb_tokens+1, 300, input_length=25, trainable=False) Lambda (lambda x: K.max(x, axis=1), output_shape=(300,)) TimeDistributed(Dense(300, activation=’relu’)) Embedding(nb_tokens+1, 300, input_length=25, trainable=False) SIMILARITY 1/(1 + |h-h’|²) Same weights
24. Siamese Network Learned Similarity https://github.com/bradleypallen/ker as-quora-question-pairs
25. Siamese Network Going further Dropout? BiLSTM? GRU? Mean Pooling? Batch Norm? More Layers? N-char? Maxout ?
26. Siamese Network Conclusion What is your name Embedding 0.34 0.32 0.01 0.31 0.27 0.21 0.21 0.04 0.42 0.22 ... ... ... .. .. 0.5 0.12 0.5 0.44 0.41 What was your mood Embedding 0.34 0.12 0.01 0.21 0.17 0.21 0.51 0.04 0.52 0.32 ... ... ... .. .. 0.5 0.22 0.5 0.24 0.36 S I M I L A R I T Y
27. Non-siamese network Example Bilateral Multi-Perspective Matching for Natural Language Sentences
28. Better Algorithm: SVM Distance Function Better algorithm Learn the Distance Function Fallback detection Nearest Neighbors + Best Intent ✅
29. Better Algorithm: SVM J.P. Vert
30. Better Algorithm: SVM J.P. Vert So the Siamese Networks are kernels! So we can use SVM J.P. Vert
31. Fallback Detection Distance Function Better algorithm Learn the Distance Function Fallback detection Nearest Neighbors + Best Intent ✅ ✅
32. Fallback Detection If Probability < Threshold Then Fallback External dataset of unrelated sentences
33. Learned Distance Function Better algorithm Learn the Distance Function Fallback detection SVM + Best Intent Or Fallback ✅ ✅ ✅
34. Thanks ! @chris_bour @MatthieuBizien @ParisNLP