Title :
Unsupervised induction and filling of semantic slots for spoken dialogue systems using frame-semantic parsing
Author :
Yun-Nung Chen ; Wang, Wei Yu ; Rudnicky, Alexander I.
Author_Institution :
Sch. of Comput. Sci., Carnegie Mellon Univ., Pittsburgh, PA, USA
Abstract :
Spoken dialogue systems typically use predefined semantic slots to parse users´ natural language inputs into unified semantic representations. To define the slots, domain experts and professional annotators are often involved, and the cost can be expensive. In this paper, we ask the following question: given a collection of unlabeled raw audios, can we use the frame semantics theory to automatically induce and fill the semantic slots in an unsupervised fashion? To do this, we propose the use of a state-of-the-art frame-semantic parser, and a spectral clustering based slot ranking model that adapts the generic output of the parser to the target semantic space. Empirical experiments on a real-world spoken dialogue dataset show that the automatically induced semantic slots are in line with the reference slots created by domain experts: we observe a mean averaged precision of 69.36% using ASR-transcribed data. Our slot filling evaluations also indicate the promising future of this proposed approach.
Keywords :
grammars; interactive systems; natural language processing; pattern clustering; speech recognition; unsupervised learning; ASR-transcribed data; frame semantic theory; frame-semantic parser; natural language; predefined semantic slot; real-world spoken dialogue dataset; spectral clustering based slot ranking model; spoken dialogue system; unified semantic representation; unlabeled raw audio; unsupervised filling; unsupervised induction; Computational modeling; Hidden Markov models; Iron; Manuals; Probabilistic logic; Semantics; Vectors; Unsupervised slot induction; semantic representation; semantic slot filling;
Conference_Titel :
Automatic Speech Recognition and Understanding (ASRU), 2013 IEEE Workshop on
Conference_Location :
Olomouc
DOI :
10.1109/ASRU.2013.6707716