The Whole Truth and Nothing But the Truth: Faithful and Controllable Dialogue Response Generation...
685
17.6
Microsoft Research335 тыс
Следующее
07.07.23 – 83649:01Популярные
286 дней – 99 68116:22
05.12.23 – 14213:18Опубликовано 7 июля 2023, 18:47
The Whole Truth and Nothing But the Truth: Faithful and Controllable Dialogue Response Generation with Dataflow Transduction and Constrained Decoding
In a real-world dialogue system, generated text must be truthful and informative while remaining fluent and adhering to a prescribed style. Satisfying these constraints simultaneously is difficult for the two predominant paradigms in language generation: neural language modeling and rule-based generation. We describe a hybrid architecture for dialogue response generation that combines the strengths of both paradigms. The first component of this architecture is a rule-based content selection model defined using a new formal framework called dataflow transduction, which uses declarative rules to transduce a dialogue agent's actions and their results (represented as dataflow graphs) into context-free grammars representing the space of contextually acceptable responses. The second component is a constrained decoding procedure that uses these grammars to constrain the output of a neural language model, which selects fluent utterances. Our experiments show that this system outperforms both rule-based and learned approaches in human evaluations of fluency, relevance, and truthfulness.
See more at microsoft.com/en-us/research/v...
In a real-world dialogue system, generated text must be truthful and informative while remaining fluent and adhering to a prescribed style. Satisfying these constraints simultaneously is difficult for the two predominant paradigms in language generation: neural language modeling and rule-based generation. We describe a hybrid architecture for dialogue response generation that combines the strengths of both paradigms. The first component of this architecture is a rule-based content selection model defined using a new formal framework called dataflow transduction, which uses declarative rules to transduce a dialogue agent's actions and their results (represented as dataflow graphs) into context-free grammars representing the space of contextually acceptable responses. The second component is a constrained decoding procedure that uses these grammars to constrain the output of a neural language model, which selects fluent utterances. Our experiments show that this system outperforms both rule-based and learned approaches in human evaluations of fluency, relevance, and truthfulness.
See more at microsoft.com/en-us/research/v...
Свежие видео
10 дней – 5970:34
11 дней – 125 1653:44
11 дней – 4 4551:01
13 дней – 7700:56
15 дней – 1 6775:12
16 дней – 19 0381:13Случайные видео
109 дней – 548 8557:04
201 день – 202 21311:35
352 дня – 8 8335:50
03.08.23 – 21 8321:18
16.10.22 – 157 9419:56
10.03.17 – 2911:50
3 дня – 930 32014:49
4 дня – 2531:08
5 дней – 4312:58
10 дней – 3 3864:02
10 дней – 744:42
11 дней – 2 010 55531:13
3 дня – 1223:08
6 дней – 1 2717:11