9th International Colloquium on Grammatical Inference (ICGI), Saint-Malo 2008

Grammatical Inference: news from the Machine Translation front

author: François Yvon, Computer Sciences Laboratory for Mechanics and Engineering Sciences (LIMSI), National Center for Scientific Research (CNRS)
published: Oct. 9, 2008, recorded: September 2008, views: 151

Slides

Slides
0:00	Grammatical Inference: News from the Machine Translation Front
2:57	Outline - 1
3:28	Outline - 2
4:02	Outline - 3
4:14	Outline - 4
4:30	Some Problems with Machine Translation
6:27	Mainstream Statistical Machine Translation - 1
6:32	Mainstream Statistical Machine Translation - 2
6:59	Mainstream Statistical Machine Translation - 3
7:30	Mainstream Statistical Machine Translation - 4
7:38	Mainstream Statistical Machine Translation - 5
7:47	Mainstream Statistical Machine Translation - 6
7:51	Mainstream Statistical Machine Translation - 7
8:11	Mainstream Statistical Machine Translation - 8
8:17	Mainstream Statistical Machine Translation - 9
8:27	Take a Set of Parallel Sentences - 1
8:54	Take a Set of Parallel Sentences - 2
8:59	Take a Set of Parallel Sentences - 3
9:38	Take a Set of Parallel Sentences - 4
11:13	Take a Set of Parallel Sentences - 5
11:17	Take a Set of Parallel Sentences - 6
11:23	Take a Set of Parallel Sentences - 7
12:17	Training 1.a: Build Word Alignments - 1
13:08	Training 1.a: Build Word Alignments - 2
13:25	Training 1.a: Build Word Alignments - 3
13:27	Training 1.a: Build Word Alignments - 4
13:42	Training 1.a: Build Word Alignments - 5
13:48	Training 1.a: Build Word Alignments - 6
14:02	Training 1.a: Build Word Alignments - 7
14:16	Training 1.a: Build Word Alignments - 8
14:28	Training 1.a: Build Word Alignments - 9
14:37	Training 1.a: Build Word Alignments - 10
14:56	Training 1.a: Build Word Alignments - 11
15:13	Training 1.a: Build Word Alignments - 12
15:42	Training 1.a: Build Word Alignments - 13
16:11	Training 1.a: Build Word Alignments - 14
16:54	Training 1.a: Build Word Alignments - 15
17:25	Training 1.a: Build Word Alignments - 16
17:26	Training 1.a: Build Word Alignments - 17
17:36	Training 1.b : Accumulate “Phrases” and their Statistics - 1
18:48	Training 1.b : Accumulate “Phrases” and their Statistics - 2
19:15	Training 1.b : Accumulate “Phrases” and their Statistics - 3
19:28	Training 1.b : Accumulate “Phrases” and their Statistics - 4
19:51	Training 1.b : Accumulate “Phrases” and their Statistics - 5
20:12	Training 1.b : Accumulate “Phrases” and their Statistics - 6
20:18	Training 1.b : Accumulate “Phrases” and their Statistics - 7
20:39	Training 1.b : Accumulate “Phrases” and their Statistics - 8
20:47	A Real World Phrase-Table - 1
21:37	A Real World Phrase-Table - 2
21:57	A Real World Phrase-Table - 3
22:13	Training 1.b : Accumulate “Phrases” and their Statistics - 8
22:14	Training 1.b : Accumulate “Phrases” and their Statistics - 9
22:16	Training 1.b : Accumulate “Phrases” and their Statistics - 10
22:17	Training 1.b : Accumulate “Phrases” and their Statistics - 11
22:54	Training 1.b : Accumulate “Phrases” and their Statistics - 12
23:06	Training 1.b : Accumulate “Phrases” and their Statistics - 13
23:29	Training 1.b : Accumulate “Phrases” and their Statistics - 14
23:41	Training 1.b : Accumulate “Phrases” and their Statistics - 15
23:42	Training 1.b : Accumulate “Phrases” and their Statistics - 16
24:33	Training 2: Learn a Target Language Model - 1
24:41	Training 2: Learn a Target Language Model - 2
24:42	Training 2: Learn a Target Language Model - 3
24:43	Training 2: Learn a Target Language Model - 4
24:43	Training 2: Learn a Target Language Model - 5
25:31	Training 2: Learn a Target Language Model - 6
26:30	Training 2: Learn a Target Language Model - 7
26:50	Training 3: Tune the Score Function - 1
28:00	Training 3: Tune the Score Function - 2
28:12	Training 3: Tune the Score Function - 3
28:47	Training 3: Tune the Score Function - 4
28:48	Training 3: Tune the Score Function - 5
28:58	Decoding, An Optimisation Problem - 1
29:05	Decoding, An Optimisation Problem - 2
29:37	Decoding, An Optimisation Problem - 3
29:47	Decoding, An Optimisation Problem - 4
30:00	Decoding, An Optimisation Problem - 5
30:16	Decoding, An Optimisation Problem - 6
30:51	Decoding, An Optimisation Problem - 7
31:04	Decoding, An Optimisation Problem - 8
31:05	Get Some Numbers - 1
31:40	Get Some Numbers - 2
31:55	Get Some Numbers - 3
32:16	Get Some Numbers - 4
32:26	Get Some Numbers - 5
32:35	Get Some Numbers - 6
32:46	Get Some Numbers - 7
32:49	Get Some Numbers - 8
33:10	Get Some Numbers - 9
33:32	Get Some Numbers - 10
34:06	A Step Back: Finite-State SMT - 1
34:35	A Step Back: Finite-State SMT - 2
34:40	A Step Back: Finite-State SMT - 3
35:43	A Step Back: Finite-State SMT - 4
36:30	A Step Back: Finite-State SMT - 5
37:15	A Step Back: Finite-State SMT - 6
37:21	Approaches to Reordering - 1
37:37	Approaches to Reordering - 2
38:15	Approaches to Reordering - 3
38:21	Approaches to Reordering - 4
38:22	Approaches to Reordering - 5
39:08	Approaches to Reordering - 6
39:12	Approaches to Reordering - 7
39:13	Approaches to Reordering - 8
39:25	Approaches to Reordering - 9
39:39	Approaches to Reordering - 10
39:48	Approaches to Reordering - 11
39:49	Approaches to Reordering - 12
39:51	Approaches to Reordering - 11
40:12	Approaches to Reordering - 12
40:24	IBM Style Constraints - 1
41:56	IBM Style Constraints - 2
42:36	IBM Style Constraints - 3
42:54	Approaches to Reordering - 12
43:19	Inversion Transduction Grammars (ITGs)
44:17	ITG’s Permutations - 1
44:43	ITG’s Permutations - 2
44:50	ITG’s Permutations - 3
44:55	ITG’s Permutations - 4
44:57	ITG’s Permutations - 5
45:04	ITG’s Permutations - 6
45:46	Approaches to Reordering - 12
45:49	Learning Reordering Rules - 1
46:10	Learning Reordering Rules - 2
46:13	Learning Reordering Rules - 3
46:15	Learning Reordering Rules - 4
46:57	Learning Reordering Rules - 2
47:22	Learning Reordering Rules - 3
47:41	Learning Reordering Rules - 4
47:46	Learning Reordering Rules - 5
48:22	Approaches to Reordering - 12
49:37	Why It Works - 1
49:49	Why It Works - 2
50:09	Why It Works - 3
50:15	Why It Works - 4
50:32	Why It Works - 5
50:41	Why It Works - 6
51:12	Why It Works - 7
51:17	Why It Works - 8
52:53	Why It Fails - 1
53:24	Why It Fails - 2
54:17	Why It Fails - 3
54:33	Why It Fails - 4
55:22	Why It Fails - 5
56:20	Temporary Conclusions - 1
56:21	Temporary Conclusions - 2
56:22	Temporary Conclusions - 3
56:23	Temporary Conclusions - 4
57:29	Questions?

Report a problem or upload files

If you have found a problem with this lecture or would like to send us extra material, articles, exercises, etc., please use our ticket system to describe your request and upload the data.
Enter your e-mail into the 'Cc' field, and we will keep you updated with your request's status.

Lecture popularity: You need to login to cast your vote.