• Publications
  • Influence
Prosody-based automatic segmentation of speech into sentences and topics
TLDR
We investigate the use of prosody (information gleaned from the timing and melody of speech) for speech segmentation tasks, and evaluate performance on two speech corpora, Broadcast News and Switchboard. Expand
  • 487
  • 45
  • PDF
Multi-Domain Joint Semantic Frame Parsing Using Bi-Directional RNN-LSTM
TLDR
This paper proposes a holistic multi-domain, multi-task (i.e. slot filling, domain and intent detection) modeling approach to estimate complete semantic frames for all user utterances addressed to a conversational system, demonstrating the distinctive power of deep learning methods, namely bi-directional recurrent neural network (RNN) with long-short term memory (LSTM) cells to handle such complexity. Expand
  • 252
  • 34
  • PDF
MultiWOZ 2.1: Multi-Domain Dialogue State Corrections and State Tracking Baselines
TLDR
We use crowdsourced workers to fix the state annotations and utterances in the MultiWOZ dataset and show joint state tracking performance on the corrected state annotations. Expand
  • 65
  • 28
  • PDF
The ICSI/UTD Summarization System at TAC 2009
TLDR
We describe improvements to our 2008 system that result in a top-performing summarization system for the TAC 2009 Summarization task. Expand
  • 88
  • 18
  • PDF
The ICSI Summarization System at TAC 2008
TLDR
The ICSI multi-document summarization system relies on a general framework that casts summarization as a global optimization problem with an integer linear programming solution. Expand
  • 84
  • 15
  • PDF
A Hybrid Hierarchical Model for Multi-Document Summarization
TLDR
We formulate extractive summarization as a two step learning problem building a generative model for pattern discovery and a regression model for inference. Expand
  • 140
  • 14
  • PDF
Statistical Morphological Disambiguation for Agglutinative Languages
TLDR
We present statistical models for morphological disambiguation in agglutinative languages, with a specific application to Turkish. Expand
  • 167
  • 14
  • PDF
End-to-End Memory Networks with Knowledge Carryover for Multi-Turn Spoken Language Understanding
TLDR
An architecture using end-to-end memory networks to model knowledge carryover in multi-turn conversations, where utterances encoded with intents and slots can be stored as embeddings in the memory and the decoding phase applies an attention model to leverage previously stored semantics for intent prediction and slot tagging simultaneously. Expand
  • 106
  • 14
  • PDF
Dialog State Tracking: A Neural Reading Comprehension Approach
TLDR
We formulate dialog state tracking as a reading comprehension task to answer the question $what\ is\ the state\ of\ the\ current dialog?$ after reading conversational context. Expand
  • 49
  • 14
  • PDF
What is left to be understood in ATIS?
TLDR
In this paper, our goal is not experimenting with domain specific techniques or features which can help with the remaining SLU errors, but instead exploring methods to realize this utility via extensive error analysis. Expand
  • 153
  • 13
  • PDF