TREC-6 1997 Spoken Document Retrieval Track Overview and Results

نویسندگان

  • John S. Garofolo
  • Ellen M. Voorhees
  • Vincent M. Stanford
  • Karen Spärck Jones
چکیده

This paper describes the 1997 TREC-6 Spoken Document Retrieval (SDR) Track which implemented a first evaluation of retrieval of broadcast news excerpts using a combination of automatic speech recognition and information retrieval technologies. The motivations behind the SDR Track and background regarding its development and implementation are discussed. The SDR evaluation collection and topics are described and summaries and analyses of the results of the track are presented. Finally, plans for future SDR tracks are described. Since this was the first implementation of an evaluation of SDR, the evaluation itself as well as the evaluated technology should be considered experimental. The results of the first SDR Track were very encouraging and showed us that SDR could be successfully implemented and evaluated. However, the results of the SDR Track should be considered preliminary since the 50-hour spoken document collection used was very small for retrieval experiments (even though it was considered extremely large for speech recognition purposes.) Nonetheless, with thirteen groups participating in the TREC-6 SDR Track, a considerable amount of experience was gained in implementing and evaluating the SDR task. This experience will greatly benefit the next 1998 TREC-7 SDR Track.

برای دانلود متن کامل این مقاله و بیش از 32 میلیون مقاله دیگر ابتدا ثبت نام کنید

ثبت نام

اگر عضو سایت هستید لطفا وارد حساب کاربری خود شوید

منابع مشابه

ETH TREC-6: Routing, Chinese, Cross-Language and Spoken Document Retrieval

ETH Zurich's participation in TREC-6 consists of experiments in the main routing task, both manual and automatic runs in the Chinese retrieval track, cross-language retrieval in each of German, French and En-glish as part of the new cross-language retrieval track, and experiments in speech recognition and retrieval under the new spoken document retrieval track. This year our routing experiments...

متن کامل

The TREC-6 Spoken Document Retrieval Track

The Text REtrieval Conference (TREC) workshops provide a forum for di erent groups to compare retrieval systems on common retrieval tasks. The 1997 TREC workshop will feature a Spoken Document Retrieval task for the rst time. This paper motivates the task and describes the measures to be used to evaluate the e ectiveness of the retrieval methodologies. 1. The Text REtrieval Conference The Text ...

متن کامل

AT&T at TREC-7 SDR Track

AT&T participated in the Spoken Document Retrieval (SDR) track of TREC-7. Our speech retrieval system uses modern Information Retrieval (IR) methods in conjunction with in-house automatic speech recognition. The novel feature of our TREC-7 work is the use of document expansion to reduce the performance loss due to ASR errors. Results show that retrieval from automatic transcriptions of speech i...

متن کامل

1998 TREC-7 Spoken Document Retrieval Track Overview and Results

This paper describes the 1998 TREC-7 Spoken Document Retrieval (SDR) Track which implemented an evaluation of retrieval of broadcast news excerpts using a combination of automatic speech recognition and information retrieval technologies. The motivations behind the SDR Track and background regarding its development and implementation are discussed. The SDR evaluation collection and topics are d...

متن کامل

Short Queries, Natural Language and Spoken Document Retrieval: Experiments at Glasgow University

This paper contains a description of the methodology and results of the three TREC submissions made by the Glasgow IR group (glair). In addition to submitting to the ad hoc task, submissions were also made to NLP track and to the SDR speech `pre-track'. Results from our submissions reveal that some of our approaches have performed poorly (i.e. ad hoc and NLP track), but we have also had success...

متن کامل

ذخیره در منابع من


  با ذخیره ی این منبع در منابع من، دسترسی به آن را برای استفاده های بعدی آسان تر کنید

برای دانلود متن کامل این مقاله و بیش از 32 میلیون مقاله دیگر ابتدا ثبت نام کنید

ثبت نام

اگر عضو سایت هستید لطفا وارد حساب کاربری خود شوید

عنوان ژورنال:

دوره   شماره 

صفحات  -

تاریخ انتشار 1997