Data driven methods for utterance semantic tagging
نویسندگان
چکیده
The proliferation of mobile devices, along with advances in speech and natural language processing technologies, have given birth to a new wave of personal assistance applications that enable users to quickly and more naturally perform many tasks through voice on their smart devices. This paper focuses on a natural language understanding (NLU) solution for one such application. We adopted a data-driven approach, aiming to take advantage of large volume of deployment data for continued learning and system improvement. In this paper, we compare two different statistical models---a hidden Markov model and a maximum entropy Markov model---for the task of semantic slot extraction, and we present empirical results on real user data.
منابع مشابه
SEIMCHA: a new semantic image CAPTCHA using geometric transformations
As protection of web applications are getting more and more important every day, CAPTCHAs are facing booming attention both by users and designers. Nowadays, it is well accepted that using visual concepts enhance security and usability of CAPTCHAs. There exist few major different ideas for designing image CAPTCHAs. Some methods apply a set of modifications such as rotations to the original imag...
متن کاملInvestigating the Contribution of Distributional Semantic Information for Dialogue Act Classification
This paper presents a series of experiments in applying compositional distributional semantic models to dialogue act classification. In contrast to the widely used bag-ofwords approach, we build the meaning of an utterance from its parts by composing the distributional word vectors using vector addition and multiplication. We investigate the contribution of word sequence, dialogue act sequence,...
متن کاملSyntactic annotation of spontaneous speech: application to call-center conversation data
Both frameworks are based on the automatic semantic analysis of Human-Human spoken conversations. The semantic interpretation of a spoken utterance can be split into a two-level process: a tagging process projecting lexical items into basic conceptual constituents and a composition process that takes as input these basic constituents and combine them in a possibly complex semantic interpretatio...
متن کاملIdentification of utterance intention in Japanese spontaneous spoken dialogue by use of prosody and keyword information
This paper describes the study on the identification of utterance intention in Japanese spontaneous dialogue. The procedure of tagging the dialog act which was labeled by hand was evaluated by the analysis of the prosodic information and keyword recognition for the dialogues of scheduling and travel arrangement domains. It was shown that the integration of prosody and keywords relevant to illoc...
متن کاملTowards the Annotation of Penn TreeBank with Information Structure
Information Structure (IS) determines the “communicative” segmentation of the meaning of an utterance, which makes it central to the semantics–syntax– intonation interface and therefore also to NLP. Despite this relevance, IS has not received much attention in the context of the majority of the reference treebanks for data-driven NLP that already contain a semantic and syntactic layers of annot...
متن کامل