Differences
This shows you the differences between two versions of the page.
Both sides previous revision Previous revision Next revision | Previous revision Next revision Both sides next revision | ||
draft [2009/07/15 13:35] 147.228.47.142 |
draft [2009/07/15 13:56] ptacek |
||
---|---|---|---|
Line 8: | Line 8: | ||
Re: progress: there is progress in the following: | Re: progress: there is progress in the following: | ||
- | - language model re-training for the collected dialogue data | + | - evaluation of the ASR performance using the WoZ data (WP5.1) |
+ | - language model re-training for the collected dialogue data (using also data sources external to COMPANIONS) (WP5.1) | ||
+ | - implementation of the real-time speaker adaptation (WP5.1) | ||
- additional dialogue transription for ASR is ongoing (WP52.? T5.2.1) | - additional dialogue transription for ASR is ongoing (WP52.? T5.2.1) | ||
- DM has been transferred from USFD to Prague (WP5.3) | - DM has been transferred from USFD to Prague (WP5.3) | ||
Line 19: | Line 21: | ||
- DA set is being prepared, also based on the sample dialogues (WP5.2) | - DA set is being prepared, also based on the sample dialogues (WP5.2) | ||
- preliminary DA tagger (on std DAMSL-SWBD tagset) working (~35% error rate) (WP5.2) | - preliminary DA tagger (on std DAMSL-SWBD tagset) working (~35% error rate) (WP5.2) | ||
+ | - | ||
- integration work is ongoing (CU/ZCU, internally at CU) | - integration work is ongoing (CU/ZCU, internally at CU) | ||
but no functioning full demo yet (beyond what we've presented in Madrid) | but no functioning full demo yet (beyond what we've presented in Madrid) | ||
Line 49: | Line 52: | ||
- | ===== Automatic Speech Recognition ===== | + | |
+ | |||
+ | ===== Automatic Speech Recognition | ||
features: improved language models, real-time speaker adaptation | features: improved language models, real-time speaker adaptation | ||
+ | performance indicator: WER | ||
+ | |||
+ | |||
- | ===== Speech Reconstruction ===== | + | ===== Speech Reconstruction |
features: omit filler phrases, irrelevant speech events, false starts, repetitions, | features: omit filler phrases, irrelevant speech events, false starts, repetitions, | ||
imlementation(zahrnout tuhle info?): moses natrenovany na korpusu | imlementation(zahrnout tuhle info?): moses natrenovany na korpusu | ||
Line 58: | Line 66: | ||
XXX Mirek | XXX Mirek | ||
- | ===== Morphology Analyzer and POS tagging ===== | + | ===== Morphology Analyzer and POS tagging |
features: XXX Mirek/ | features: XXX Mirek/ | ||
performance indicator: accuracy | performance indicator: accuracy | ||
- | ===== Syntactic Parsing ===== | + | ===== Syntactic Parsing |
features: induce dependencies and labels | features: induce dependencies and labels | ||
performance indicator: f-measure | performance indicator: f-measure | ||
Line 69: | Line 77: | ||
- | ===== Semantic Parsing ===== | + | ===== Semantic Parsing |
features: meaning representation with semantic roles (69 roles), coordinations, | features: meaning representation with semantic roles (69 roles), coordinations, | ||
performance indicator: f-measure | performance indicator: f-measure | ||
- | ===== Information Extraction ===== | + | ===== Information Extraction |
features: template based identification of predicates | features: template based identification of predicates | ||
covering predicates from before-mentioned set of DAFs. | covering predicates from before-mentioned set of DAFs. | ||
Line 79: | Line 87: | ||
- | ===== Named Entities Recognition ===== | + | ===== Named Entities Recognition |
features: detect person names, geographical locations (organizations myslim nepotrebne) | features: detect person names, geographical locations (organizations myslim nepotrebne) | ||
performance indicator: f-measure | performance indicator: f-measure | ||
- | ===== Dialog Act Tagging ===== | + | ===== Dialog Act Tagging |
features: tagset derived from DAMSL-SWBD, DA is a key feature driving the decision, what to say next. | features: tagset derived from DAMSL-SWBD, DA is a key feature driving the decision, what to say next. | ||
performance indicator: accuracy | performance indicator: accuracy | ||
- | ===== Sentiment Analysis ===== | + | ===== Sentiment Analysis |
features: za tohle bych vydaval klasifikator, | features: za tohle bych vydaval klasifikator, | ||
performance indicator: f-measure | performance indicator: f-measure | ||
+ | |||
+ | |||
Line 96: | Line 106: | ||
===== Complete System Evaluation ===== | ===== Complete System Evaluation ===== | ||
T5.2.7 tohle zminuje, nick webb to pro nas asi neudela | T5.2.7 tohle zminuje, nick webb to pro nas asi neudela | ||
- | performance indicator: | + | performance indicator: |
- | ===== Dialog Manager ===== | + | ===== Dialog Manager |
features: reply types, using (language independed) predicates (prakticky to znamena, ze pojmenuju testy na prechodech v dafech anglicky) | features: reply types, using (language independed) predicates (prakticky to znamena, ze pojmenuju testy na prechodech v dafech anglicky) | ||
performance indicator: rucni hodnoceni prijatelnosti vybrane akce | performance indicator: rucni hodnoceni prijatelnosti vybrane akce | ||
- | ===== Natural Language Generation ===== | + | ===== Natural Language Generation |
features: variations, underspecified input (dott format), emotional markup (natvrdo v dafech a templatech u hodnoticich vet) | features: variations, underspecified input (dott format), emotional markup (natvrdo v dafech a templatech u hodnoticich vet) | ||
performance indicator: BLEU score | performance indicator: BLEU score |