Dialog act modeling for virtual personal assistant applications using a small volume of labeled data and domain knowledge
نویسندگان
چکیده
Recently, virtual personal assistant (VPA) applications have been employed in mobile devices, which provide a natural and convenient interface between human and machines. As the VPA services become popular, consumers demand for a wider service than their scope, so the rapid development becomes more important. This paper introduces a dialog act modeling approach for VPA applications, which is an extension of a Latent Dirichlet Allocation model. This approach enables the rapid and costeffective development by reducing human efforts for manual labeling and the development of a fail-safe product by incorporating domain knowledge such as dictionaries, cross-lingual data, and logic rules. The experimental results showed that a reliable and high-performance dialog act model was built only with a small volume of labeled data and domain knowledge.
منابع مشابه
Dialog with a Personal Assistant
This paper describes a new generic architecture for dialog systems enabling communication between a human user and a personal assistant based on speech acts. Dialog systems are often domain-related applications. That is, the system is developed for specific applications and cannot be reused in other domains. A major problem concerns the development of scalable dialog systems capable to be exten...
متن کاملA Portuguese spoken and multi
This paper presents an overview of the spoken and multimodal dialog Portuguese corpora collected in the context of the FASiL (Flexible and Adaptive Spoken Language and Multi-Modal Interfaces) project. The project developed a Virtual Personal Assistant application in the Personal Information Management domain, exploiting the state-of-theart of speech and multi-modal technology. The FASiL corpora...
متن کاملDesigning a Message Handling Assistant Using the BDI Theory and Speech Act Theory
This thesis introduces a new approach to designing a Message Handling Assistant (MA). It presents a model of an MA and an intention extraction function for text messages, such as emails and Newsgroups articles. Based on a speech act theory and the belief-desire-intention (BDI) theory of rational agency, we define a generic MA. By interpreting intuitive descriptions of the desired behaviours of ...
متن کاملDetecting out-of-domain utterances addressed to a virtual personal assistant
Conversational understanding systems, especially virtual personal assistants (VPAs), perform “targeted” natural language understanding, assuming their users stay within the walled gardens of covered domains, and back-off to generic web search otherwise. However, users usually do not know the concept of domains and sometimes simply do not distinguish the system from simple voice search. Hence it...
متن کاملHypotheses ranking and state tracking for a multi-domain dialog system using multiple ASR alternates
In this paper, we present an approach to improve the accuracy of multi-domain multi-turn spoken dialog system (SDS) by including alternate results from automatic speech recognition (ASR). Often, even if the top ranked result from the ASR is not correct, the correct result may still be available in the NBest list or in the word confusion network (WCN). Thus, the SDS performance can be improved b...
متن کامل