Automation of Telecom Italia directory assistance service: field trial results

Billi, R.; Canavesio, F.; Rullent, Claudio

doi:10.1109/ivtta.1998.727685

Cited by 20 publications

(9 citation statements)

References 9 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Our procedures have been used to automatically detect new formulations, discover why existing formulations exhibit high rejection rates, and to compute the percentage of routed versus automated calls per phone number. 1…”

Section: Discussionmentioning

confidence: 99%

See 1 more Smart Citation

Learning new user formulations in automatic directory assistance

Popovici¹,

Andorno²,

Laface³

et al. 2002

IEEE International Conference on Acoustics Speech and Signal Processing

View full text Add to dashboard Cite

Telecom Italia has deployed since the beginning of year 2001 a nationwide automatic Directory Assistance (DA) system that routinely serves customers asking for residential and business listings. DA for business listings is a challenging task: one of its main problems is that customers formulate their requests for the same listing with a great variability. Since it is difficult to reliably predict a priori the user formulations, in this paper we propose a procedure for detecting, from field data, user formulations that were not foreseen by the designers. These formulations can be added, as variants, to the denominations already included in the system to reduce its failures. We show that using a large database associating phonetic transcriptions of user utterances with the phone number provided by the automatic service, a completely unsupervised approach detects most of the old formulations. Furthermore, our procedure is able to filter a huge amount of calls routed to the operators, and to detect a limited number of phonetic strings that are good candidate to be included as new formulation variants in the system vocabulary.

show abstract

Section: Discussionmentioning

confidence: 99%

“…Whenever the automatic system is unable to terminate the transaction with the customer, the call is routed to a human operator. A description of the system, related to the management of the residential calls, has been presented in [1].…”

Section: Introductionmentioning

confidence: 99%

Learning new user formulations in automatic directory assistance

Popovici¹,

Andorno²,

Laface³

et al. 2002

IEEE International Conference on Acoustics Speech and Signal Processing

View full text Add to dashboard Cite

show abstract

“…Institutions on both sides of the Atlantic have deployed experimental systems during mid-late 90's. The early studies focused mainly on residential DA (Billi et al 1998;Kamm et al 1995;Lennig et al 1994), and speech recognition was the major topic of research -as long as personal names were correctly recognized, the search could be a simple database lookup. As a result, the dialog strategies centered on limiting the scope (hence perplexity) of the target listing space for ASR and the confidence measures mostly relied on features from the ASR.…”

Section: History Of Voice Searchmentioning

confidence: 99%

Semantic Frame‐Based Spoken Language Understanding

Wang¹,

Deng²,

Acero³

2011

Spoken Language Understanding

View full text Add to dashboard Cite

“…In the last 10 years, one of the most active H.M. Chang ( ) AT&T Labs-Research, Florham Park, USA e-mail: harry_chang@labs.att.com fields in ASR-driven applications in the US telecom industry is the automation of operator-based DA services (often referred to as the 411 service in the US). However, in spite of the tremendous efforts in this area (Lennig et al 1995;Seide and Kellner 1997;Billi et al 1998;Schramm et al 2000;Boves et al 2000;Scharenborg et al 2001;Natarajan et al 2002;Jan et al 2003;Levin and Mané 2005;Alphonso and Chang 2006), the directory assistance automation (DAA) systems deployed today still perform at a level that is far below most of the trained human operators employed by major telecom service providers in the United States (Rudolph 2008).…”

Section: Introductionmentioning

confidence: 99%

Comparing machine and human performance for caller’s directory assistance requests

Chang

2007

Int J Speech Technol

View full text Add to dashboard Cite

To understand how to systematically construct the machine models for automating Directory Assistance (DA) that are capable of reaching the performance level of human DA operators, we conducted a number of studies over the years. This paper describes the methods used for such studies and the results of laboratory experiments. These include a series of benchmark tests configured specifically for DA related tasks to evaluate the performance of state-ofthe-art and commercially-available Hidden Markov Model (HMM) based Automatic Speech Recognition (ASR) technologies. The results show that the best system achieves a 57.9% task completion rate on the city-state-recognition benchmark test. For the most frequently-requested-listing benchmark test, the best system achieves a 40% task completion rate.

show abstract

Automation of Telecom Italia directory assistance service: field trial results

Cited by 20 publications

References 9 publications

Learning new user formulations in automatic directory assistance

Learning new user formulations in automatic directory assistance

Semantic Frame‐Based Spoken Language Understanding

Comparing machine and human performance for caller’s directory assistance requests

Contact Info

Product

Resources

About