Introduction DARPA has recently initiated a plan for a common spoken language task, to be developed independently by all members of the DARPA community, with the hope that it will provide a mechanism leading to appropriate formal evaluation procedures at the level of spoken language. The task that was selected for this purpose is the Air Travel Information System (ATIS) task, based on selected ...

The feedback effects of ATIS information for helping travelers choose optimal routes are researched. We build a dynamic route choice model for ATIS; analyze the adjustment and renewal process of travel route decision based on Bayesian method and logit choice model. Furthermore, through comparing whether can obtain ATIS information for users, the route choice rules are designed and practical sim...

Can we leverage the advances in statistical parsing over the last fifteen years to build more powerful “transportable” Natural Language Interfaces to Databases (NLIs)? To address this question, this paper reports on the PRECISE NLI, which uses a statistical parser as a “plug in”. The paper shows how a strong semantic model enables PRECISE to correct parser errors, and map from parsed questions ...

This paper presents a mixed approach to spoken language understanding that tries to make best use of the advantages of both statistical and knowledge-based algorithms. Results obtained on ATIS (Air Travel Information System) scenario transferred to Italian language are presented and discussed.

This paper describes the collection of authentic human–human air travel information data in Tok Pisin, the pidgin/creole language spoken in Papua New Guinea. Pros and cons of authentic data are discussed, as compared to data collected in more controlled settings like Wizard-of-Oz simulations. Some unexpected real-life phenomena that affect the data, and normally do not occur in corpora compiled...

Advanced Traveler Information Systems, which have for long been regarded as one of the most promising future applications of wireless vehicular networks for use in the field of Intelligent Transportation Systems (ITS), are effectively becoming part of today’s reality. Many drivers already access the information provided by such systems, for example checking for the state of the streets along a ...

We describe the rst sentence understanding system that is completely based on learned methods both for understanding individual sentences, and determinig their meaning in the context of preceding sentences. We describe the models used for each of three stages in the understanding: semantic parsing, semantic classi cation, and discourse modeling. When we ran this system on the last test (Decembe...

This paper presents a new paradigm of “challenge” evaluation of Spoken Language Understanding. This methodology aims at a quantitative assessment with a high diagnostic power, by opposition with standard ATIS-like frameworks. This paper details the methodology as well as the results of an evaluation campaign held by the French CNRS research agency. The benefits of this methodology are also disc...

A major domain of e-government research considers effective delivery of information to citizens (i.e. government-to-citizens services). These services involve all sets of communication and transaction between government, at various levels, and citizens. Research indicates that due to the expectations that people hold while utilizing public services available through digital means, there needs t...

This paper outlines a new telescope operations paradigm that is intended to achieve low operating costs with high operating ef-ciency and high scientiic productivity. The paradigm is based on the existing Principal Astronomer approach used in conjunction with atis, a language for commanding remotely located automatic telescopes. This paper describes a telescope management system called the Asso...