Directory Assistance Research Articles

Modelling customer satisfaction in Telecom New Zealand

Presents some techniques developed by Telecom New Zealand where first the component factors that most influence the overall satisfaction rating of Telecom’s service are found. Second, these components are quantified and the improvement required for each component to meet overall excellence targets is determined. Gives a successful example of International Directory Assistance which shows that these techniques have significantly raised customer overall satisfaction in just five months.

A field evaluation of the Italian “Automated reverse directory assistance” service

The paper describes a field evaluation of the automated ‘reverse directory assistance’ service presently in use in Italy in which information about names and addresses is provided by a TTS system. A simulation of the service using a natural voice was also run to get comparative data. Both services were accessed from an office room and a call-box on the street. Different evaluation metrics, such as intelligibility, task completion, task correctness, transaction success, and user's reactions were used. The aim of the work was to evaluate TTS synthesis in real world use and to make a comparison between laboratory data and data on system performance in a real application. Such a comparison suggested that in laboratory tests more attention should be dedicated to simulate more closely the conditions that can be predicted in real world use, by including important aspects that are generally not taken into consideration in laboratory tests and that are likely to have a large influence on TTS system performance such as environmental noise, prosody, and task complexity. The results also underline the importance of field evaluations to get an overall view of the usability of a service in real applications and with users who are as similar as possible to actual users.

Advances in human-interface engineering for reverse directory assistance (ACNA) services

After years of productive research, speech synthesis is now profitably automating services by answering queries via constrained dialogs, directly accessing individual computer databases, and speaking text created from disparate sources of information. Directory-based services, such as Automated Customer Name and Address (ACNA), requires synthesis with high intelligibility and name pronunciation accuracy. Current synthesis technology achieves those goals. However, even the best of current speech technology is not good enough to mindlessly “drop” into complex services. Customized directory preprocessing is still necessary to transform listing data, which commonly contains unconventional abbreviations, unlabeled acronyms, and scrambled word ordering, into a sentence suitable for synthesis. This article describes state-of-the-art directory preprocessing programs that have led to successful implementations for synthesis services in 2 major U.S. telephone companies (Ameritech and Bell Atlantic). Of course, the basic capabilities of the synthesizer, such as pronunciation accuracy, speech quality and naturalness, play a large role. Efforts ensured locality terms were pronounced in accordance with local custom. Finally, for prompts and other fixed messages, this article describes experiments that determined whether the naturalness of recorded speech offsets the undesirable discontinuity between recorded and synthesized utterances.

Cumulative Trauma Disorders among Telephone Directory Assistance Operators

Cumulative Trauma Disorders among Telephone Directory Assistance Operators

Searching the Internet

Your first experiences using the Internet search and retrieval programs will be profitable if you understand the environment within which they were designed to work. When used properly, they represent the Internet's directory assistance that will aid you in information discovery.

Method and apparatus for automation of directory assistance using speech recognition

In a telecommunications system, automatic directory assistance uses a voice processing unit comprising a lexicon of lexemes and data representing a predetermined relationship between each lexeme and calling numbers in a locality served by the automated directory assistance apparatus. The voice processing unit issues messages to a caller making a directory assistance call to prompt the caller to utter a required one of said lexemes. The unit detects the calling number originating a directory assistance call and, responsive to the calling number and the relationship data computes a probability index representing the likelihood of a lexeme being the subject of the directory assistance call. The unit employs a speech recognizer to recognize, on the basis of the acoustics of the caller's utterance and the probability index, a lexeme corresponding to that uttered by the caller.

How (Fast) Can I Help You? Tone of Voice and Telephone Operator Efficiency in Interactions1

This study examined how tone of voice related to directory assistance operators' s]peed at handling customer calls. Brief (5‐s) audio exerpts of telephone operator conversation were presented to judges who rated these on 5 personality dimensions and 5 vocal qualities. When these measures were correlated with a key measure of job performance, namely call duration, results showed that operators who were faster at handling calls were rated as significantly more enthusiastic, sympathetic, confident, professional, and friendly (average r= ‐.56). The voices of faster operators were also rated as significantly more changing (vs. monotonous) and as containing clearer (vs. slurred) pronunciation. Results are discussed in terms of the role that tone of voice plays in telephone operator‐customer interaction, and telephone interactions more generally.

Speech recognition issues for directory assistance applications

Telephone companies in the United States handle over 6 billion Directory Assistance (DA) calls each year. Automation of even a portion of DA calls could significantly reduce the cost of DA services. This paper explores two factors affecting successful automation of DA: (a) the effect of directory size on speech recognition performance, and (b) the complexity of existing DA call interactions. Speech recognition performance for a set of 200 spoken names was measured for directories ranging from 200 to 1.5 million unique names. Recognition accuracy decreased from 82.5% for a 200-name directory to 16.5% for a 1.5 million name directory. In part because high recognition accuracy is not easily achievable for these very large, low-context directories, it is likely that initial implementations of DA automation will focus on a small percentage of calls, requiring a smaller vocabulary. To maximize potential savings, listings that are most frequently requested constitute the optimal vocabulary. To identify critical issues in automating frequent DA requests, approximately 13,000 DA calls from an office near a major metropolitan area in the United States were studied. In this sample, 245 listings covered 10% of the call volume, and 870 listings covered 20% of the call volume.

Directory assistance automation in Bell Canada: Trial results

Speech recognition was used to automate directory assistance in a 6-month trial with Bell Canada's public customers. The bilingual application gave the caller a choice at the beginning of the dialog to continue in English or French. Over 89% of calls were either partially or fully automated. Customer and operator reactions to the system were positive. Bell-Northern Research's flexible vocabulary recognizer, using a vocabulary of 1,700 city names and synonyms, performed well under real world conditions. Economically significant operator work time savings were demonstrated.

Interactive voice technology at work: The CSELT experience

This paper is a survey of the speech technologies and applications developed at CSELT, some of which are employed in real services deployed in the Italian telephone network. With the rise of significant speech recognition and text-to-speech applications, the activity of our lab encompasses now a broader set of activities, from new algorithmic approaches to speech product engineering and application development. In particular, the paper gives an overview of the products originated from our speech technology research. It describes two operative applications, namely a voice dialing service for large name directories, which is installed in the CSELT PABX, and an automated network service for directory assistance, which is now accessible to all the Italian telephone customers.

Do Human Factors Color Recommendations have any Practical Value?

This study surveyed 342 directory-assistance (DA) operators regarding the text and background colors they used to display directory listings on their computer workstations. Color combinations were observed in use, recorded, and scored based on independent, published subjective ratings of color comfort and edge sharpness. Operators also answered a questionnaire that included information about: how often they changed colors and if they had experienced eye problems. No relationships were found between self-reported eye problems and observed color choices. However, operators using color combinations that were scored low on either comfort, edge sharpness, or both, were more likely to change the colors on their screens more frequently. This result provides evidence that independent ratings can provide practical guidance in making color recommendations.

Obtaining telephone numbers for a rural Medicaid population: issues for outreach and research.

This study examined the accuracy and costs of determining whether rural, low-income Medicaid recipients did or did not have a phone, and of obtaining phone numbers for those who did. For a random sample of 209 families, we compared phone information obtained from phone books and directory assistance with information obtained from department of social services (DSS) records. DSS records identified 51% of the sample as having phones, compared with 19%-25% for phone books and directory assistance. For identifying families as having no home phone or a phone with a number that matched the one in the DSS record, phone books or directory assistance corresponded with DSS records in 52%-57% of the sample. Using phone books or directory assistance was up to 3.2 times more costly than using DSS records. The study highlighted the need to establish policies to promote the exchange of information between social services and public health agencies and researchers.

Speech Compression in Attendant Services: Analysis of a Queueing System with Delay-Dependent Service Times

We study a multiserver queueing system in which delays influence service times. The system arises in a proposed application of speech compression technology to telephone directory assistance. Customers who must wait for service record their service requests and these are compressed and played to a server as soon as one becomes available (even if the customer is still talking). The longer the delay, the larger the portion of the request that may be compressed, and hence the larger the service time reduction. We develop a model that predicts various performance measures of the system relative to the corresponding system without speech compression.

Large-vocabulary continuous speech recognition algorithm applied to a multi-modal telephone directory assistance system

This paper describes an accurate and efficient algorithm for very-large-vocabulary continuous speech recognition. It is based on a two-stage LR parser with hidden Markov models (HMMs) as phoneme models. To improve recognition accuracy, it uses the forward and backward trellis likehood. To improve search efficiency, it uses adjusting windows and merges candidates that have the same allophonic phoneme sequences and grammatical state, and then merges candidates at the meaning level. This algorithm was applied to a telephone directory assistance system that contains more than 70,000 subscribers (about 80,000 words) to evaluate its speaker-independent speech recognition capabilities. For eight speakers, the algorithm achieved a speech understanding rate of 65% for spontaneous speech. The results show that the system performs well in spite of the large word perplexity. This paper also describes a multi-modal dialog system that uses our large-vocabulary speech recognition algorithm.

The effect of different work-rest schedules on fatigue and performance of a simulated directory assistance operator's task.

The directory assistance operators' task with a visual display terminal was simulated to determine the preferable work-rest schedule from among the three different work-rest schedules: 30 min of work followed by 5 min break (30-5); 60 min of work followed by 10 min break (60-10); and 120 min of work without any break (120-0). The total working duration was 2 h. The errors made during the work and the subjective responses before and after the working period (2 h) were recorded and analysed. The results showed that the first two work-rest schedules were preferable to the third condition (120-0) as significantly fewer errors were made (p > or = 0.01) when a short break was provided. There was no significant difference (p > or = 0.10) in errors between (30-5) and (60-10) schemes. Fewer before and after subjective responses differed significantly for (30-5) and (60-10) schemes than for the (120-0) scheme. The overall effect of schemes on subjective responses, however, was not statistically significant at the 5% level. Since both (30-5) and (60-10) schemes were found acceptable, the (60-10) scheme is recommended since it is easy to implement, causes fewer breaks and therefore, fewer work interruptions, and leads to fewer total minutes of break.

Telephone directories: Alternatives to print

In the Spring 1992 issue of Reference Services Review, Douglas Ernest, Joan Beam, and Jennifer Monath noted that “Telephone directories have been an integral part of most public and academic libraries for nearly a century. Telephone directories represent an anomaly among library collections; known to virtually all users, they nevertheless often go unrecognized when librarians discuss reference sources.” After the break‐up of the AT&T telephone monopoly in 1984, the process of acquiring telephone directories became more difficult and expensive. Seven regional holding companies, called the “Baby Bells,” were created to provide competition for the equipment and services that were once dominated by “Ma Bell.” The regional Bell companies began to charge each other for directories; as a result, the regional telephone companies had to pass on the expense to customers. It is still common for a regional Bell to provide libraries with directories of in‐state white pages, but to charge for areas outside the state, separate yellow pages within the state, and independently published telephone directories. Telephone directories, once provided free of charge to libraries to reduce the workload of operators, are now in competition with the fee‐based service of directory assistance.

A very large vocabulary continuous speech recognition algorithm for telephone directory assistance

This paper proposes a speech recognition algorithm for large vocabulary continuous speech. The proposed algorithm is based on the hidden Markov model (HMM)-LR algorithm using a generalized predictive LR parser and phoneme HMMs. The following three techniques are applied to improve recognition performance and reduce processing time. 1 The forward and the backward likelihood are used to accurately determine the likelihood in the beam search. 2 To reduce the trellis computation in HMM speech recognition and for efficient search, only the speech frames in which the predicted phoneme seems to exist are used by the window for phoneme matching. 3 For efficient search, adjusting identical phoneme sequences are merged by checking the stack and the state of the LR parser. The algorithm was applied to a telephone directory assistance task involving more than 70, 000 subscribers. A recognition experiment for continuous word utterance was done. The sentence recognition rate was 85 percent for speaker-dependent speech recognition; the sentence recognition rate was 71 percent for speaker-independent speech recognition. The sentence understanding rate was 59 percent for speaker-dependent speech recognition with spontaneous utterances.

The effect of screen format on visual list search

Three experiments are reported which investigate the effect of different screen formatting variables on the time it takes a subject to locate a target in a list of surnames presented on a computer screen. This has direct application to large databases such as electronic directories, where an operator must search for the names and addresses of customers in order to locate telephone numbers in response to directory assistance enquiries. The goal of this research was to isolate those factors which facilitate target recognition, and to produce a set of recommendations for screen design whenever lists of information are presented. The variables manipulated included case, spacing of surnames, alternating line format, and the insertion of blank lines between categories of information. The results indicated that subjects located targets more rapidly when screens were spaced than when they were non–spaced, and when alternating line format was used compared with non–alternating line format. No difference was found ...

Employee stress and health complaints in jobs with and without electronic performance monitoring

Current applications of electronic performance monitoring based on job design theories that consider worker performance rather than stress issues are likely to generate unsatisfying and stressful jobs (Smith et al, 1986). This study examines critical job design elements that could influence worker stress responses in an electronic monitoring context. A questionaire survey of employees in telecommunications companies representative of each region in the United States examined job stress in directory assistance, service representative and clerical jobs with specific emphasis on the influence of electronic monitoring of job performance, satisfaction and employee health. Useable surveys were received from 745 employees representing seven operating companies and AT&T; a response rate of about 25%. The results of this survey indicated that employees who had their performance electronically monitored perceived their working conditions as more stressful, and reported higher levels of job boredom, psychological tension, anxiety, depression, anger, health complaints and fatigue. It is postulated that these effects may be related to changes in job design due to electronic performance monitoring.

Arm pain in the workplace. A small area analysis.

In the mid-1980s, use-related arm pain was recognized as a major issue for worker health and workplace safety. National policy targeted these "cumulative trauma disorders," "overuse syndromes," and "motion illnesses" for a "special emphasis program" by the Occupational Safety and Health Administration, a federal regulatory agency. The program begins with case recognition to identify the responsible ergonomic hazards with the goal of mandating ergonomic remedies. This report is a small area analysis of the impact of this process on the US West Communications, Inc work force. US West employs some 55,000 workers in 14 states. Commencing in the mid-1980s, workers in four of these states complained of upper extremity pain and were diagnosed as suffering from conditions encompassed by the "cumulative trauma disorders" rubric. The incidence was tenfold higher in one task category, directory assistance operators. No ergonomic descriptor can account for the four endemics of arm pain in directory assistance operators. The response of the medical communities to the plight of these injured claimants varied considerably from community to community. Denver represents one extreme where the clinical judgments led to multiple surgical interventions, generated a total direct cost of some $1.5 million, and left many permanently disabled workers in their wake. This analysis raises many reservations about the validity of the "cumulative trauma disorder" hypothesis and provides an object lesson in the potential for untoward outcomes from the premature introduction of clinical hypotheses into the arena of health policy.