DARPA Speech & NLP Workshop '92.pdf

Workshop Overview (Marcus, p.3)
- Sponsored by DARPA/SISTO
- Focus: Evaluating and advancing spoken language systems

Session 1: Evaluating Spoken Language Systems (Allen, p.5)
- Multi-Site Data Collection for Spoken Language Corpus (MADCOW, MIT)
- DARPA ATIS Benchmark Test Results (NIST)
- Interactive Spoken Language System Evaluation (MIT)
- Subject-Based Evaluation Measures (SRI, MIT, UC Berkeley, Stanford)

Session 2: Spoken Language Systems II (Ward, p.41)
- Spontaneous Speech Collection for ATIS Domain (AT&T)
- Human-Machine Problem Solving with SLS (SRI)
- Spoken Language Processing at LIMSI (LIMSI-CNRS)
- Genuine Spoken Enquiries Collection using WOZ Techniques (RSRE)

Session 3: Spoken Language Systems III (Makhoul, p.65)
- Chronus System Progress Report & ATIS Benchmark Results (AT&T)
- BBN BYBLOS & HARC ATIS Benchmark Results (BBN)
- Speech Recognition in Open Tasks (CMU)
- MIT ATIS System Progress Report (MIT)
- Paramax ATIS System Improvements & Benchmark Results (Paramax)
- SRI ATIS Benchmark Test Results (SRI)

Session 4: Statistical Language Modeling (Joshi, p.101)
- Adaptive Language Modeling (IBM)
- Improvements in Stochastic Language Modeling (CMU)
- Simple Rule-Based Part-of-Speech Tagger (UPenn)
- Decision Tree Models for POS Tagging (IBM)
- Inside-Outside Reestimation from Partially Bracketed Corpora (AT&T, UPenn)
- Probabilistic Prediction & Picky Chart Parsing (Stanford, Paramax)
- History-based Grammars for Probabilistic Parsing (IBM)
- Stochastic Tree-Adjoining Grammars (UPenn)
- Constrained CFG Language Models Parameter Estimation (WashU, Brown, Bellcore)
- Analogical Parser for Restricted Domains (AT&T)
- Automatically Acquiring Phrase Structure (UPenn)

Session 5a: Acoustic Modeling (Murveit, p.161)
- Recent Topics in Speech Recognition Research (NTT)
- Vocabulary & Environment Adaptation in Vocabulary-Independent Speech Recognition (CMU, Apple)