| { | |
| "paper_id": "2005", | |
| "header": { | |
| "generated_with": "S2ORC 1.0.0", | |
| "date_generated": "2023-01-19T15:50:20.412581Z" | |
| }, | |
| "title": "Adaptation of an Automotive Dialogue System to Users' Expertise", | |
| "authors": [ | |
| { | |
| "first": "Liza", | |
| "middle": [], | |
| "last": "Hassel", | |
| "suffix": "", | |
| "affiliation": { | |
| "laboratory": "", | |
| "institution": "Ludwig Maximilian University Munich", | |
| "location": {} | |
| }, | |
| "email": "[email protected]" | |
| }, | |
| { | |
| "first": "Eli", | |
| "middle": [], | |
| "last": "Hagen", | |
| "suffix": "", | |
| "affiliation": { | |
| "laboratory": "", | |
| "institution": "Ludwig Maximilian University Munich", | |
| "location": {} | |
| }, | |
| "email": "[email protected]" | |
| }, | |
| { | |
| "first": "Innovationszentrum", | |
| "middle": [], | |
| "last": "Bmw", | |
| "suffix": "", | |
| "affiliation": { | |
| "laboratory": "", | |
| "institution": "Ludwig Maximilian University Munich", | |
| "location": {} | |
| }, | |
| "email": "" | |
| }, | |
| { | |
| "first": "A", | |
| "middle": [ | |
| "G" | |
| ], | |
| "last": "Munich", | |
| "suffix": "", | |
| "affiliation": { | |
| "laboratory": "", | |
| "institution": "Ludwig Maximilian University Munich", | |
| "location": {} | |
| }, | |
| "email": "" | |
| } | |
| ], | |
| "year": "", | |
| "venue": null, | |
| "identifiers": {}, | |
| "abstract": "Spoken dialogue systems (SDSs) can be used to operate devices, e.g. in the automotive environment. People using these systems usually have different levels of experience. However, most systems do not take this into account. In this paper we present a method to build a dialogue system in an automotive environment that adapts to the user's experience with the system. We implemented the adaptation in a prototype and carried out exhaustive tests. Our usability tests show that adaptation increases both user performance and user satisfaction.", | |
| "pdf_parse": { | |
| "paper_id": "2005", | |
| "_pdf_hash": "", | |
| "abstract": [ | |
| { | |
| "text": "Spoken dialogue systems (SDSs) can be used to operate devices, e.g. in the automotive environment. People using these systems usually have different levels of experience. However, most systems do not take this into account. In this paper we present a method to build a dialogue system in an automotive environment that adapts to the user's experience with the system. We implemented the adaptation in a prototype and carried out exhaustive tests. Our usability tests show that adaptation increases both user performance and user satisfaction.", | |
| "cite_spans": [], | |
| "ref_spans": [], | |
| "eq_spans": [], | |
| "section": "Abstract", | |
| "sec_num": null | |
| } | |
| ], | |
| "body_text": [ | |
| { | |
| "text": "Since the first in-car SDSs appeared in the late nineties (Heisterkamp (2001) ; Haller (2003) ), a lot of research has been carried out to make these systems adaptable: Cnossen et al. (2004) and Piechulla et al. (2003) investigated the task demand on user attention; Mourant et al. (2001) did research on the influence of age on user behaviour; Rogers et al. (2000) investigated the adaptation of in-car information services to user preferences; the main research focus of Akyol et al. (2001) and Libuda (2001) was on clarification dialogues and how to deal with errors. In these studies emphasis was on other factors than the user's knowledge about the system, thus we chose that as our basis for adapting the dialogue system's behaviour.", | |
| "cite_spans": [ | |
| { | |
| "start": 58, | |
| "end": 77, | |
| "text": "(Heisterkamp (2001)", | |
| "ref_id": "BIBREF11" | |
| }, | |
| { | |
| "start": 80, | |
| "end": 93, | |
| "text": "Haller (2003)", | |
| "ref_id": "BIBREF8" | |
| }, | |
| { | |
| "start": 169, | |
| "end": 190, | |
| "text": "Cnossen et al. (2004)", | |
| "ref_id": "BIBREF4" | |
| }, | |
| { | |
| "start": 195, | |
| "end": 218, | |
| "text": "Piechulla et al. (2003)", | |
| "ref_id": "BIBREF16" | |
| }, | |
| { | |
| "start": 267, | |
| "end": 288, | |
| "text": "Mourant et al. (2001)", | |
| "ref_id": "BIBREF14" | |
| }, | |
| { | |
| "start": 345, | |
| "end": 365, | |
| "text": "Rogers et al. (2000)", | |
| "ref_id": "BIBREF18" | |
| }, | |
| { | |
| "start": 473, | |
| "end": 492, | |
| "text": "Akyol et al. (2001)", | |
| "ref_id": null | |
| }, | |
| { | |
| "start": 497, | |
| "end": 510, | |
| "text": "Libuda (2001)", | |
| "ref_id": "BIBREF13" | |
| } | |
| ], | |
| "ref_spans": [], | |
| "eq_spans": [], | |
| "section": "Introduction", | |
| "sec_num": "1" | |
| }, | |
| { | |
| "text": "A dialogue system that behaves in the same way for all users, in spite of their different degree of experience, will, for neither of them, be a truly usable interface (Wu, 2000) . Users who have little or no experience with a dialogue system will have different priorities, expectations and needs than experienced ones. Because novices know little about a system, they will need a kind of tutorial to learn how it works. A further help for these users should consist of detailed confirmation of the actions triggered by their voice commands. Experts will want to accomplish the desired tasks quickly and straightforward. Such users need little guidance and the confirmation prompts should be short or could be left out completely. A solution to this dilemma is often the development of systems that are very easy for novices, but not very effective thereafter (Landauer, 1997) . Ideally, the system should be able to recognize the degree of expertise of the user and adapt to it (Nielsen, 1993) . Only in such a way it can provide at all times and for all users a maximum performance and user satisfaction.", | |
| "cite_spans": [ | |
| { | |
| "start": 167, | |
| "end": 177, | |
| "text": "(Wu, 2000)", | |
| "ref_id": "BIBREF19" | |
| }, | |
| { | |
| "start": 860, | |
| "end": 876, | |
| "text": "(Landauer, 1997)", | |
| "ref_id": "BIBREF12" | |
| }, | |
| { | |
| "start": 979, | |
| "end": 994, | |
| "text": "(Nielsen, 1993)", | |
| "ref_id": "BIBREF15" | |
| } | |
| ], | |
| "ref_spans": [], | |
| "eq_spans": [], | |
| "section": "Introduction", | |
| "sec_num": "1" | |
| }, | |
| { | |
| "text": "Based on these observations, we developed a system that adapts its prompts as users become more experienced. The adaptation is done at the textual level, i.e. exclusively the information content of the system prompts is adapted, but neither dialogue strategy nor functional range.", | |
| "cite_spans": [], | |
| "ref_spans": [], | |
| "eq_spans": [], | |
| "section": "Introduction", | |
| "sec_num": "1" | |
| }, | |
| { | |
| "text": "Our speech interface was implemented as part of BMW's iDrive system (Haller, 2003) . In addition to speech, iDrive has a manual-visual interface with a central input device in the centre console (controller, see figure 1) and a central display in the centre column (see figure 2) . When users operate the controller (turn left and right, push in four directions and press down), they receive visual feedback on the display. Over the speech channel, users can operate tasks in the areas entertainment, navigation, communication and climate control. Users activate the ASR with a push-to-talk (PTT) button on the steering wheel or in the middle console near the controller. The dialogue style is command and control as illustrated in tables 1 and 3 (p. 4). Users Figure 2 : Display Control can ask for help (general characteristics of the SDS) or for options (in the current dialogue state available voice commands). In the prototype we have implemented additional. They are discussed in Hassel and Hagen (2005) The iDrive SDS is currently configured for ca. 3000 words and phrases. For our experiments, we used the German version. In Hassel and Hagen (2005) we describe the experiments in detail. For further information about iDrive see Hagen et al. (2004) .", | |
| "cite_spans": [ | |
| { | |
| "start": 68, | |
| "end": 82, | |
| "text": "(Haller, 2003)", | |
| "ref_id": "BIBREF8" | |
| }, | |
| { | |
| "start": 270, | |
| "end": 279, | |
| "text": "figure 2)", | |
| "ref_id": null | |
| }, | |
| { | |
| "start": 986, | |
| "end": 1009, | |
| "text": "Hassel and Hagen (2005)", | |
| "ref_id": "BIBREF10" | |
| }, | |
| { | |
| "start": 1133, | |
| "end": 1156, | |
| "text": "Hassel and Hagen (2005)", | |
| "ref_id": "BIBREF10" | |
| }, | |
| { | |
| "start": 1237, | |
| "end": 1256, | |
| "text": "Hagen et al. (2004)", | |
| "ref_id": "BIBREF7" | |
| } | |
| ], | |
| "ref_spans": [ | |
| { | |
| "start": 761, | |
| "end": 769, | |
| "text": "Figure 2", | |
| "ref_id": null | |
| } | |
| ], | |
| "eq_spans": [], | |
| "section": "Introduction", | |
| "sec_num": "1" | |
| }, | |
| { | |
| "text": "In section 2, we introduce our method for classifying users. We then explained the proposed adaptation (section 3) and report the effects on usability (section 4).", | |
| "cite_spans": [], | |
| "ref_spans": [], | |
| "eq_spans": [], | |
| "section": "Introduction", | |
| "sec_num": "1" | |
| }, | |
| { | |
| "text": "The overall goal for our work is to increase user satisfaction and efficiency with the speech interface. One vehicle to achieve this goal is to account for the user's experience with the SDSs. We assume that experts need and want less guidance than novices. In a SDS, this aspect can easily be conveyed at the textual level, as table 1 shows. Currently, we are using only the two categories, novice and expert, but the concept easily allows for the inclusion of intermediate ones. Future work will address the optimal number of categories needed.", | |
| "cite_spans": [], | |
| "ref_spans": [], | |
| "eq_spans": [], | |
| "section": "Classification of the Users", | |
| "sec_num": "2" | |
| }, | |
| { | |
| "text": "Adaptation should happen unnoticed in the background without interfering with the problem solving task (and the driving!) (Rogers et al., 2000) . We developed a concept that unobtrusively analyses the user's behaviour while interacting with the system. After every interaction, the system assigns users a category in a range between novice and expert. This categorisation accounts for every task separately, i.e. users can be experts at telephone tasks (e.g. dialing a telephone number) but novices at navigation tasks (e.g. changing the map). In this way, the system can account not only for different global levels of experience, but also for different patterns of use. For the calculation of the user category for a given task, we calculate a user model in terms of a vector U M using the following parameters:", | |
| "cite_spans": [ | |
| { | |
| "start": 122, | |
| "end": 143, | |
| "text": "(Rogers et al., 2000)", | |
| "ref_id": "BIBREF18" | |
| } | |
| ], | |
| "ref_spans": [], | |
| "eq_spans": [], | |
| "section": "Adaptation Concept", | |
| "sec_num": "2.1" | |
| }, | |
| { | |
| "text": "Parameter Meaning # help requests, h", | |
| "cite_spans": [], | |
| "ref_spans": [], | |
| "eq_spans": [], | |
| "section": "Adaptation Concept", | |
| "sec_num": "2.1" | |
| }, | |
| { | |
| "text": "Users asked for general information about the system # option requests, o", | |
| "cite_spans": [], | |
| "ref_spans": [], | |
| "eq_spans": [], | |
| "section": "Adaptation Concept", | |
| "sec_num": "2.1" | |
| }, | |
| { | |
| "text": "Users asked for the currently available voice commands # timeouts, t", | |
| "cite_spans": [], | |
| "ref_spans": [], | |
| "eq_spans": [], | |
| "section": "Adaptation Concept", | |
| "sec_num": "2.1" | |
| }, | |
| { | |
| "text": "The system did not get any acoustic signal # ASR-failure, e", | |
| "cite_spans": [], | |
| "ref_spans": [], | |
| "eq_spans": [], | |
| "section": "Adaptation Concept", | |
| "sec_num": "2.1" | |
| }, | |
| { | |
| "text": "The system could not understand the users input, e.g. OOV words or unintelligible speech Relative use freq., Hr = Figure 3 : Calculation of the User Category A user is categorised for the actual task. When he/she calls this task again, the system responds with the appropriate prompt -for novice or expert. Table 1 shows three examples for system responses for experts and novices.", | |
| "cite_spans": [], | |
| "ref_spans": [ | |
| { | |
| "start": 114, | |
| "end": 122, | |
| "text": "Figure 3", | |
| "ref_id": null | |
| }, | |
| { | |
| "start": 307, | |
| "end": 314, | |
| "text": "Table 1", | |
| "ref_id": "TABREF1" | |
| } | |
| ], | |
| "eq_spans": [], | |
| "section": "Adaptation Concept", | |
| "sec_num": "2.1" | |
| }, | |
| { | |
| "text": "The comparison function takes three values as input: \u2206 U M (see above), s, and chance behaviour. The threshold value s accounts for people's ability to learn by analogy (section 2.2) and for the fact that people tend to forget new acquired knowledge quickly (section 2.3). A heurism makes sure that wrong category assignments due entirely to chance behaviour are avoided (section 2.4).", | |
| "cite_spans": [], | |
| "ref_spans": [], | |
| "eq_spans": [], | |
| "section": "Adaptation Concept", | |
| "sec_num": "2.1" | |
| }, | |
| { | |
| "text": "Our domain consists of many similar tasks (e.g., the command \"play\" in the CD and DVD menu), thus we include peoples's ability to learn by analogy into our modification of s. We assume that when users know how a particular task works, they will learn the use of a similar task faster and in an easier way than the use of a completely different task. Thus, we have grouped the system tasks in families, according to their similarity. Users are classified as novice or expert for each task in a family. When a user calls a task of a family, the category assignment depends on the status of the other tasks in this family, i.e. if a user is already classified as experts for other tasks in that family, then s is reduced. If users are classified as experts for more than 50% of the tasks of that family, they are then classified as experts for all tasks in that family.", | |
| "cite_spans": [], | |
| "ref_spans": [], | |
| "eq_spans": [], | |
| "section": "Task Families", | |
| "sec_num": "2.2" | |
| }, | |
| { | |
| "text": "People tend to forget newly acquired knowledge rapidly, unless they revise it frequently, until it becomes part of long term memory (Edelmann, 1996) . The influence of these psychological insights on the learning behaviour of the users is incorporated into the adaptation concept by adapting the threshold value s to the elapsed time between interactions. The more time elapses since the last interaction the faster users are assigned novice status, i.e. s is reduced.", | |
| "cite_spans": [ | |
| { | |
| "start": 132, | |
| "end": 148, | |
| "text": "(Edelmann, 1996)", | |
| "ref_id": "BIBREF6" | |
| } | |
| ], | |
| "ref_spans": [], | |
| "eq_spans": [], | |
| "section": "Elapsed Time since the Last Interaction", | |
| "sec_num": "2.3" | |
| }, | |
| { | |
| "text": "The real value of this factor could not be evaluated since our tests only lasted 45 min. We need long term evaluations to elicit the time after which we can assume users to have forgotten the voice commands. The alternating forgetting and learning periods give rise to a forgetting behaviour that can be described with a differential equation. The reason is that the intervals during which people forget a certain percentage of the commands become longer over the time because the acquired knowl-edge becomes gradually part of the long term memory. One should bear in mind these reflections when evaluating the elapsed time factor.", | |
| "cite_spans": [], | |
| "ref_spans": [], | |
| "eq_spans": [], | |
| "section": "Elapsed Time since the Last Interaction", | |
| "sec_num": "2.3" | |
| }, | |
| { | |
| "text": "We included a heurism in order to account for user behaviour produced by chance: After a prediction of the user model, look for positive/negative verification for this prediction in the next interaction (Rich, 1979) . We fulfil this claim as follows: The comparison function has to yield three times the same result for a certain task before users are assigned another category. That is, if users have novice status for the task t, they have to behave three times as experts before the system assigns them expert status for t. Doing so, the probability that the system categorizes users in a category only by chance is minimized.", | |
| "cite_spans": [ | |
| { | |
| "start": 203, | |
| "end": 215, | |
| "text": "(Rich, 1979)", | |
| "ref_id": "BIBREF17" | |
| } | |
| ], | |
| "ref_spans": [], | |
| "eq_spans": [], | |
| "section": "Chance Behaviour", | |
| "sec_num": "2.4" | |
| }, | |
| { | |
| "text": "Although we could not colletc empirical data for the use of a threefold cycle, the results of our tests indicate that 3 is a good choice. 41% of the reference test subjects (T S A ) agreed with the statement \"lists with options are too long\", but only 10% of the prototype users (T S B ) did. T S A had to ask explicitly for options; T S B (with novice status) were told the options wihtout asking for them. Bearing in mind this difference and the answers mentioned above, we conclude that, at the beginning, the repetition of the options is necessary and that the chosen three times is not too much.", | |
| "cite_spans": [], | |
| "ref_spans": [], | |
| "eq_spans": [], | |
| "section": "Chance Behaviour", | |
| "sec_num": "2.4" | |
| }, | |
| { | |
| "text": "The difference between system prompts for novices and for experts is mainly their explicitness, e.g. while for novices the SDS mentions the available voice commands without waiting for users to ask, experts have to explicitly ask for them (see table 3 ). Long and informative prompts would be in the long run annoying to frequent users.", | |
| "cite_spans": [], | |
| "ref_spans": [ | |
| { | |
| "start": 244, | |
| "end": 251, | |
| "text": "table 3", | |
| "ref_id": "TABREF6" | |
| } | |
| ], | |
| "eq_spans": [], | |
| "section": "Adaptation of the System Prompts", | |
| "sec_num": "3" | |
| }, | |
| { | |
| "text": "The different system utterances were analysed with respect to the information they convey to users and assigned a DAMSL-category ; ). Depending on the semantics they transmit, the information can be presented in different ways (cf. table 3). To this end, we rely on the notion of Grice's conversational implicatures, and the basic principles for their calculation: cooperation principle and conversational maxims (Clark, 1997) .", | |
| "cite_spans": [ | |
| { | |
| "start": 413, | |
| "end": 426, | |
| "text": "(Clark, 1997)", | |
| "ref_id": "BIBREF3" | |
| } | |
| ], | |
| "ref_spans": [], | |
| "eq_spans": [], | |
| "section": "Adaptation of the System Prompts", | |
| "sec_num": "3" | |
| }, | |
| { | |
| "text": "Contributions conveying conventional meaning do not necessarily need linguistic signals. For experts, Openings and Closings can be performed by tones, and a Signal-understanding confirming an action requirement like \"play CD\" can indirectly be accomplished by playing the CD. Contrarily, Signal-understanding confirming a dialogue state transition, e.g. \"entertainment\", needs linguistic signals to express the confirmation. For novices, these utterances can be completed with the avail- An assertion cannot be completely replaced by nonlinguistic signals. This kind of prompt can be expressed in a less verbose manner or, at the most, be inferred from another prompt type through implicature. A example for the use of implicature in expert prompts is the combination of Assert and Action-directive, e.g. users can deduce from the directive \"Insert a CD\" the assertion \"'CD slot is empty\"'.", | |
| "cite_spans": [], | |
| "ref_spans": [], | |
| "eq_spans": [], | |
| "section": "Adaptation of the System Prompts", | |
| "sec_num": "3" | |
| }, | |
| { | |
| "text": "Signal-non-understandings could also be replaced by a non-linguistic signal. However, a beep may not fulfil the maxim of quantity (= make your contribution as informative as is required), since a tone alone may not be enough to indicate users what to do next. Besides, it may not fulfil the maxim of manner (= avoid obscurity of expression and ambiguity) because it might be difficult for the driver to discern between Signal-non-understanding and Signal-understanding tones. Therefore, we decided to express non-understanding using utterances like an Actiondirective, e.g. asking users to repeat the last utterance. Table 3 summarizes these examples. The system prompts type is given in the first column. Prompt types set in brackets were left out for expert prompts. The type of the user's utterances is set in brackets.", | |
| "cite_spans": [], | |
| "ref_spans": [ | |
| { | |
| "start": 617, | |
| "end": 624, | |
| "text": "Table 3", | |
| "ref_id": "TABREF6" | |
| } | |
| ], | |
| "eq_spans": [], | |
| "section": "Adaptation of the System Prompts", | |
| "sec_num": "3" | |
| }, | |
| { | |
| "text": "In this paper we described the classification of users between beginner and expert, and the adaptation of the system prompts to the calculated user expertise. The adaptation was assessed in a real driving situation in two test series (reference system and prototype). The evaluation showed that adaptation contributes significantly to en-hance the usability of the SDS.", | |
| "cite_spans": [], | |
| "ref_spans": [], | |
| "eq_spans": [], | |
| "section": "Conclusion", | |
| "sec_num": "4" | |
| }, | |
| { | |
| "text": "The comparison of prototype and reference system showed that adaptation contributed to improve the usability. Users of the prototype needed both less time and turns to complete the tasks than users of the reference system. T S A needed on average 62.1 sec to complete a task, and T S B 47.0 sec. T S A needed on average 8.7 turns to complete a task, and T S B 6.9 turns. Furthermore, 77% of the T S B declared that options should be prompted after every system utterance, at least at the beginning, but only 27% of the T S A agreed with that. While users that did not try adaptation were sceptic about it, the ones that tested adaptation wanted to have it afterwards. User satisfaction and task success provided further evidence in favour of adaptation. User satisfaction was higher for the prototype, and while T S B could complete 94% of the tasks, T S A could complete only 81% of them. Moreover, T S B requested options only 1 5 of the times T S A did. In general, users found the enumeration of the available options a good means to learn the systems, but in the long run the enumeration would be tedious. Therefore and because they knew they could asked for options and help, T S B approved of adaptation.", | |
| "cite_spans": [], | |
| "ref_spans": [], | |
| "eq_spans": [], | |
| "section": "Conclusion", | |
| "sec_num": "4" | |
| }, | |
| { | |
| "text": "In automotive environment it is important that users keep their eyes on the road. 68% of the T S B did (almost) not look at the display, in contrast to only 45% of the T S A . This can be interpreted as a sign that T S B knew what to say without having to look at the display, and that they realized how comfortable it is to have the available voice commands read, at least until they are known.", | |
| "cite_spans": [], | |
| "ref_spans": [], | |
| "eq_spans": [], | |
| "section": "Conclusion", | |
| "sec_num": "4" | |
| }, | |
| { | |
| "text": "According to Nielsen (1993) , systems designed for novices should be easy to learn, i.e. the learning curve should be very steep at the beginning. Our test showed that T S B reached soon the asymptote of the curve be-cause they learned very fast how to use our prototype. The system prompts for novices served their purpose. We could confirme that the initial part of the learning curve for the prototype's users corresponds to the recommended shape.", | |
| "cite_spans": [ | |
| { | |
| "start": 13, | |
| "end": 27, | |
| "text": "Nielsen (1993)", | |
| "ref_id": "BIBREF15" | |
| } | |
| ], | |
| "ref_spans": [], | |
| "eq_spans": [], | |
| "section": "Conclusion", | |
| "sec_num": "4" | |
| }, | |
| { | |
| "text": "Systems designed for experts are hard to learn but highly efficient, i.e. the learning curve is even at the beginning (Nielsen, 1993) . T S A learned by trial and error that they can speak the tasks they want to activate directly, leaving out the nodes between, that is, the reference system fullfils this claim. The next question is if our prototype would also fulfill the requirements stated by Nielsen (1993) for experts. The prompts of system B for experts turn quite the same as those of system A, this improves the efficiency. Furthermore, the prototype offers users a \"suggestion\" feature (Hassel and Hagen, 2005) to learn better ways of completing a task. Long term experiments still have to show if system B displays a typical expert learning curve over the time. Besides, further tests have to be done to confirm these assumptions also for elderly people.", | |
| "cite_spans": [ | |
| { | |
| "start": 118, | |
| "end": 133, | |
| "text": "(Nielsen, 1993)", | |
| "ref_id": "BIBREF15" | |
| }, | |
| { | |
| "start": 397, | |
| "end": 411, | |
| "text": "Nielsen (1993)", | |
| "ref_id": "BIBREF15" | |
| }, | |
| { | |
| "start": 596, | |
| "end": 620, | |
| "text": "(Hassel and Hagen, 2005)", | |
| "ref_id": "BIBREF10" | |
| } | |
| ], | |
| "ref_spans": [], | |
| "eq_spans": [], | |
| "section": "Conclusion", | |
| "sec_num": "4" | |
| } | |
| ], | |
| "back_matter": [ | |
| { | |
| "text": "We thank Klaus Schulz (LMU, Munich) for helpful discussions clarifying our ideas and for comments on earlier drafts. We also thank Stefan P\u00f6hn (Berner & Mattner) for the programming, helping to make our, often chaotic, ideas concrete. Thanks to Alexander Huber (BMW AG) for his continuing encouraging support.", | |
| "cite_spans": [], | |
| "ref_spans": [], | |
| "eq_spans": [], | |
| "section": "Aknowledgements", | |
| "sec_num": null | |
| } | |
| ], | |
| "bib_entries": { | |
| "BIBREF2": { | |
| "ref_id": "b2", | |
| "title": "Draft of DAMSL: Dialog Act Markup in Several Layers", | |
| "authors": [ | |
| { | |
| "first": "J", | |
| "middle": [ | |
| "F" | |
| ], | |
| "last": "Allen", | |
| "suffix": "" | |
| }, | |
| { | |
| "first": "M", | |
| "middle": [ | |
| "G" | |
| ], | |
| "last": "Core", | |
| "suffix": "" | |
| } | |
| ], | |
| "year": 1997, | |
| "venue": "", | |
| "volume": "", | |
| "issue": "", | |
| "pages": "", | |
| "other_ids": {}, | |
| "num": null, | |
| "urls": [], | |
| "raw_text": "J. F. Allen and M. G. Core. 1997. Draft of DAMSL: Dialog Act Markup in Several Layers. http://www.cs.rochester.edu/research/cisd/resources/damsl/.", | |
| "links": null | |
| }, | |
| "BIBREF3": { | |
| "ref_id": "b3", | |
| "title": "Using Language. Cambridge University Press", | |
| "authors": [ | |
| { | |
| "first": "H", | |
| "middle": [ | |
| "H" | |
| ], | |
| "last": "Clark", | |
| "suffix": "" | |
| } | |
| ], | |
| "year": 1997, | |
| "venue": "", | |
| "volume": "", | |
| "issue": "", | |
| "pages": "", | |
| "other_ids": {}, | |
| "num": null, | |
| "urls": [], | |
| "raw_text": "H. H. Clark. 1997. Using Language. Cambridge Univer- sity Press, Cambridge, New York, Melbourne.", | |
| "links": null | |
| }, | |
| "BIBREF4": { | |
| "ref_id": "b4", | |
| "title": "Adaptive Strategy Changes as a Function of Task Demands: A Study of Car Drivers", | |
| "authors": [ | |
| { | |
| "first": "F", | |
| "middle": [], | |
| "last": "Cnossen", | |
| "suffix": "" | |
| }, | |
| { | |
| "first": "T", | |
| "middle": [], | |
| "last": "Meijman", | |
| "suffix": "" | |
| }, | |
| { | |
| "first": "T", | |
| "middle": [], | |
| "last": "Rothengatter", | |
| "suffix": "" | |
| } | |
| ], | |
| "year": 2004, | |
| "venue": "Ergonomics", | |
| "volume": "47", | |
| "issue": "2", | |
| "pages": "218--236", | |
| "other_ids": {}, | |
| "num": null, | |
| "urls": [], | |
| "raw_text": "F. Cnossen, T. Meijman, and T. Rothengatter. 2004. Adaptive Strategy Changes as a Function of Task Demands: A Study of Car Drivers. Ergonomics, 47(2):218-236.", | |
| "links": null | |
| }, | |
| "BIBREF5": { | |
| "ref_id": "b5", | |
| "title": "Coding Dialogs with the DAMSL Annotation Scheme", | |
| "authors": [ | |
| { | |
| "first": "M", | |
| "middle": [ | |
| "G" | |
| ], | |
| "last": "Core", | |
| "suffix": "" | |
| }, | |
| { | |
| "first": "J", | |
| "middle": [ | |
| "F" | |
| ], | |
| "last": "Allen", | |
| "suffix": "" | |
| } | |
| ], | |
| "year": 1997, | |
| "venue": "AAAI Fall 1997 Symposium on Communicative Action in Humans and Machines", | |
| "volume": "", | |
| "issue": "", | |
| "pages": "28--35", | |
| "other_ids": {}, | |
| "num": null, | |
| "urls": [], | |
| "raw_text": "M. G. Core and J. F. Allen. 1997. Coding Di- alogs with the DAMSL Annotation Scheme. In AAAI Fall 1997 Symposium on Communicative Action in Humans and Machines, pages 28-35. American As- sociation for Artificial Intelligence (AAAI). URL: http://citeseer.nj.nec.com/core97coding.html.", | |
| "links": null | |
| }, | |
| "BIBREF6": { | |
| "ref_id": "b6", | |
| "title": "Lernpsychologie. Psychologie Verlagsunion", | |
| "authors": [ | |
| { | |
| "first": "W", | |
| "middle": [], | |
| "last": "Edelmann", | |
| "suffix": "" | |
| } | |
| ], | |
| "year": 1996, | |
| "venue": "", | |
| "volume": "", | |
| "issue": "", | |
| "pages": "", | |
| "other_ids": {}, | |
| "num": null, | |
| "urls": [], | |
| "raw_text": "W. Edelmann. 1996. Lernpsychologie. Psychologie Ver- lagsunion, Weinheim, 5 edition.", | |
| "links": null | |
| }, | |
| "BIBREF7": { | |
| "ref_id": "b7", | |
| "title": "Spracheingabe im neuen BMW 6er. Sonderheft ATZ/MTZ (Der neue BMW 6er)", | |
| "authors": [ | |
| { | |
| "first": "E", | |
| "middle": [], | |
| "last": "Hagen", | |
| "suffix": "" | |
| }, | |
| { | |
| "first": "T", | |
| "middle": [], | |
| "last": "Said", | |
| "suffix": "" | |
| }, | |
| { | |
| "first": "J", | |
| "middle": [], | |
| "last": "Eckert", | |
| "suffix": "" | |
| } | |
| ], | |
| "year": 2004, | |
| "venue": "", | |
| "volume": "", | |
| "issue": "", | |
| "pages": "134--139", | |
| "other_ids": {}, | |
| "num": null, | |
| "urls": [], | |
| "raw_text": "E. Hagen, T. Said, and J. Eckert. 2004. Spracheingabe im neuen BMW 6er. Sonderheft ATZ/MTZ (Der neue BMW 6er), May:134-139.", | |
| "links": null | |
| }, | |
| "BIBREF8": { | |
| "ref_id": "b8", | |
| "title": "The Display and Control Concept iDrive -Quick Access to All Driving and Comfort Functions", | |
| "authors": [ | |
| { | |
| "first": "R", | |
| "middle": [], | |
| "last": "Haller", | |
| "suffix": "" | |
| } | |
| ], | |
| "year": 2003, | |
| "venue": "", | |
| "volume": "", | |
| "issue": "", | |
| "pages": "", | |
| "other_ids": {}, | |
| "num": null, | |
| "urls": [], | |
| "raw_text": "R. Haller. 2003. The Display and Control Concept iDrive -Quick Access to All Driving and Comfort Functions.", | |
| "links": null | |
| }, | |
| "BIBREF9": { | |
| "ref_id": "b9", | |
| "title": "The New BMW 5-Series)", | |
| "authors": [ | |
| { | |
| "first": "Atz/Mtz", | |
| "middle": [], | |
| "last": "Extra", | |
| "suffix": "" | |
| } | |
| ], | |
| "year": null, | |
| "venue": "", | |
| "volume": "", | |
| "issue": "", | |
| "pages": "51--53", | |
| "other_ids": {}, | |
| "num": null, | |
| "urls": [], | |
| "raw_text": "ATZ/MTZ Extra (The New BMW 5-Series), August:51- 53.", | |
| "links": null | |
| }, | |
| "BIBREF10": { | |
| "ref_id": "b10", | |
| "title": "Evaluation of a Dialogue System in an Automotive Environment", | |
| "authors": [ | |
| { | |
| "first": "L", | |
| "middle": [], | |
| "last": "Hassel", | |
| "suffix": "" | |
| }, | |
| { | |
| "first": "E", | |
| "middle": [], | |
| "last": "Hagen", | |
| "suffix": "" | |
| } | |
| ], | |
| "year": 2005, | |
| "venue": "6th SIGdial Workshop on Discourse and Dialogue", | |
| "volume": "", | |
| "issue": "", | |
| "pages": "", | |
| "other_ids": {}, | |
| "num": null, | |
| "urls": [], | |
| "raw_text": "L. Hassel and E. Hagen. 2005. Evaluation of a Dialogue System in an Automotive Environment. In 6th SIGdial Workshop on Discourse and Dialogue, Lisbon, Portu- gal, 2-3 September 2005. Draft Version.", | |
| "links": null | |
| }, | |
| "BIBREF11": { | |
| "ref_id": "b11", | |
| "title": "Linguatronic -Product-Level Speech System for Mercedes-Benz Cars", | |
| "authors": [ | |
| { | |
| "first": "P", | |
| "middle": [], | |
| "last": "Heisterkamp", | |
| "suffix": "" | |
| } | |
| ], | |
| "year": 2001, | |
| "venue": "Proceedings of the 1st International Conference on Human Language Technology Research (HLT-01)", | |
| "volume": "", | |
| "issue": "", | |
| "pages": "", | |
| "other_ids": {}, | |
| "num": null, | |
| "urls": [], | |
| "raw_text": "P. Heisterkamp. 2001. Linguatronic -Product-Level Speech System for Mercedes-Benz Cars. In Proceed- ings of the 1st International Conference on Human Language Technology Research (HLT-01), San Diego, CA, USA.", | |
| "links": null | |
| }, | |
| "BIBREF12": { | |
| "ref_id": "b12", | |
| "title": "Behavioral Research Methods in Human-Computer Interaction", | |
| "authors": [ | |
| { | |
| "first": "T", | |
| "middle": [ | |
| "K" | |
| ], | |
| "last": "Landauer", | |
| "suffix": "" | |
| } | |
| ], | |
| "year": 1997, | |
| "venue": "Handbook of Human-Computer Interactionsecond", | |
| "volume": "", | |
| "issue": "", | |
| "pages": "203--227", | |
| "other_ids": {}, | |
| "num": null, | |
| "urls": [], | |
| "raw_text": "T. K. Landauer. 1997. Behavioral Research Methods in Human-Computer Interaction. In Martin A. He- lander, Thomas K. Landauer, and Prassad V. Prabhu, editors, Handbook of Human-Computer Interaction - second, completely revised edition, pages 203-227. North-Holland, Amsterdam, Lausanne, New York u.a.", | |
| "links": null | |
| }, | |
| "BIBREF13": { | |
| "ref_id": "b13", | |
| "title": "Improving Clarification Dialogs in Speech Command Systems with the Help of User Modeling: A Conceptualization for an In-Car User Interface", | |
| "authors": [ | |
| { | |
| "first": "L", | |
| "middle": [], | |
| "last": "Libuda", | |
| "suffix": "" | |
| } | |
| ], | |
| "year": 2001, | |
| "venue": "Online-Proceedings des 9. GI-Workshops: ABIS-Adaptivit\u00e4t und Benutzermodellierung in interaktiven Softwaresystemen. GI-Fachgruppe: Adaptivit\u00e4t und Benutzermodellierung in Interaktiven Softwaresystemen (ABIS)", | |
| "volume": "", | |
| "issue": "", | |
| "pages": "", | |
| "other_ids": {}, | |
| "num": null, | |
| "urls": [], | |
| "raw_text": "L. Libuda. 2001. Improving Clarification Dialogs in Speech Command Systems with the Help of User Modeling: A Conceptualization for an In-Car User In- terface. In Online-Proceedings des 9. GI-Workshops: ABIS-Adaptivit\u00e4t und Benutzermodellierung in in- teraktiven Softwaresystemen. GI-Fachgruppe: Adap- tivit\u00e4t und Benutzermodellierung in Interaktiven Soft- waresystemen (ABIS).", | |
| "links": null | |
| }, | |
| "BIBREF14": { | |
| "ref_id": "b14", | |
| "title": "Divided Attention Ability of Young and Older Drivers", | |
| "authors": [ | |
| { | |
| "first": "R", | |
| "middle": [ | |
| "R" | |
| ], | |
| "last": "Mourant", | |
| "suffix": "" | |
| }, | |
| { | |
| "first": "F.-J", | |
| "middle": [], | |
| "last": "Tsai", | |
| "suffix": "" | |
| }, | |
| { | |
| "first": "T", | |
| "middle": [], | |
| "last": "Al-Shihabi", | |
| "suffix": "" | |
| }, | |
| { | |
| "first": "B", | |
| "middle": [ | |
| "K" | |
| ], | |
| "last": "Jaeger", | |
| "suffix": "" | |
| } | |
| ], | |
| "year": 2001, | |
| "venue": "Proceedings of the 80th Annual Meeting of the Transportation Research Board", | |
| "volume": "", | |
| "issue": "", | |
| "pages": "", | |
| "other_ids": {}, | |
| "num": null, | |
| "urls": [], | |
| "raw_text": "R. R. Mourant, F.-J. Tsai, T. Al-Shihabi, and B. K. Jaeger. 2001. Divided Attention Ability of Young and Older Drivers. Proceedings of the 80th Annual Meeting of the Transportation Research Board.", | |
| "links": null | |
| }, | |
| "BIBREF15": { | |
| "ref_id": "b15", | |
| "title": "Usability Engineering", | |
| "authors": [ | |
| { | |
| "first": "J", | |
| "middle": [], | |
| "last": "Nielsen", | |
| "suffix": "" | |
| } | |
| ], | |
| "year": 1993, | |
| "venue": "", | |
| "volume": "", | |
| "issue": "", | |
| "pages": "", | |
| "other_ids": {}, | |
| "num": null, | |
| "urls": [], | |
| "raw_text": "J. Nielsen. 1993. Usability Engineering. Academic Press Professional, Boston u. a.", | |
| "links": null | |
| }, | |
| "BIBREF16": { | |
| "ref_id": "b16", | |
| "title": "Reducing Drivers' Mental Workload by Means of an Adaptive Man-Machine Interface", | |
| "authors": [ | |
| { | |
| "first": "W", | |
| "middle": [], | |
| "last": "Piechulla", | |
| "suffix": "" | |
| }, | |
| { | |
| "first": "C", | |
| "middle": [], | |
| "last": "Mayserb", | |
| "suffix": "" | |
| }, | |
| { | |
| "first": "H", | |
| "middle": [], | |
| "last": "Gehrke", | |
| "suffix": "" | |
| }, | |
| { | |
| "first": "W", | |
| "middle": [], | |
| "last": "K\u00f6nig", | |
| "suffix": "" | |
| } | |
| ], | |
| "year": 2003, | |
| "venue": "Transportation Research Part F: Traffic Psychology and Behaviour", | |
| "volume": "6", | |
| "issue": "4", | |
| "pages": "233--248", | |
| "other_ids": {}, | |
| "num": null, | |
| "urls": [], | |
| "raw_text": "W. Piechulla, C. Mayserb, H. Gehrke, and W. K\u00f6nig. 2003. Reducing Drivers' Mental Workload by Means of an Adaptive Man-Machine Interface. Transporta- tion Research Part F: Traffic Psychology and Behav- iour, 6(4):233-248.", | |
| "links": null | |
| }, | |
| "BIBREF17": { | |
| "ref_id": "b17", | |
| "title": "User Modeling via Stereotypes", | |
| "authors": [ | |
| { | |
| "first": "E", | |
| "middle": [], | |
| "last": "Rich", | |
| "suffix": "" | |
| } | |
| ], | |
| "year": 1979, | |
| "venue": "Cognitive Science", | |
| "volume": "3", | |
| "issue": "", | |
| "pages": "329--354", | |
| "other_ids": {}, | |
| "num": null, | |
| "urls": [], | |
| "raw_text": "E. Rich. 1979. User Modeling via Stereotypes. Cogni- tive Science, 3:329-354.", | |
| "links": null | |
| }, | |
| "BIBREF18": { | |
| "ref_id": "b18", | |
| "title": "Adaptive User Interfaces for Automotive Environments", | |
| "authors": [ | |
| { | |
| "first": "S", | |
| "middle": [], | |
| "last": "Rogers", | |
| "suffix": "" | |
| }, | |
| { | |
| "first": "C.-N", | |
| "middle": [], | |
| "last": "Fiechter", | |
| "suffix": "" | |
| }, | |
| { | |
| "first": "C", | |
| "middle": [], | |
| "last": "Thompson", | |
| "suffix": "" | |
| } | |
| ], | |
| "year": 2000, | |
| "venue": "Proceedings of the IEEE Intelligent Vehicles (IV) Symposium", | |
| "volume": "", | |
| "issue": "", | |
| "pages": "662--667", | |
| "other_ids": {}, | |
| "num": null, | |
| "urls": [], | |
| "raw_text": "S. Rogers, C.-N. Fiechter, and C. Thompson. 2000. Adaptive User Interfaces for Automotive Environ- ments. In Proceedings of the IEEE Intelligent Vehicles (IV) Symposium 2000, Detroit (USA), pages 662-667.", | |
| "links": null | |
| }, | |
| "BIBREF19": { | |
| "ref_id": "b19", | |
| "title": "Accomodating both Experts and Novices in One Interface", | |
| "authors": [ | |
| { | |
| "first": "Jing", | |
| "middle": [], | |
| "last": "Wu", | |
| "suffix": "" | |
| } | |
| ], | |
| "year": 2000, | |
| "venue": "", | |
| "volume": "", | |
| "issue": "", | |
| "pages": "", | |
| "other_ids": {}, | |
| "num": null, | |
| "urls": [], | |
| "raw_text": "Jing Wu. 2000. Accomodating both Experts and Novices in One Interface. Universal Usability Guide. Depart- ment of Computer Science, University of Maryland, http://www.otal.umd.edu/UUGuide/.", | |
| "links": null | |
| } | |
| }, | |
| "ref_entries": { | |
| "FIGREF0": { | |
| "num": null, | |
| "type_str": "figure", | |
| "text": "Controller and PTT-button", | |
| "uris": null | |
| }, | |
| "FIGREF1": { | |
| "num": null, | |
| "type_str": "figure", | |
| "text": "|task t| |interactions| How often users activated task t in relation to all other performed tasks Mean response time, Tm How long users needed to answer, also known as onset time Confidence measure, K How reliable the speech recognition was The user model U M is updated after every interaction, and it is the basis for calculating the user's current position on the expertise scale. The process is illustrated in figure 3. U M is multiplied with a weight vector U M ref representing the importance of each component of U M . The scalar product \u2206 U M is compared to a threshold value s. If \u2206 U M > s, users behave as novices, if \u2206 U M \u2264 s, they behave as experts.Expert \uf09f\uf09f\uf09f NoviceComparison function\u2022 Elapsed time since the last interaction (forgetting curve) \u2022 Function family (learning by analogy) \u2022 How often was the user asigned the same category? (\"Rich-Heurism\")", | |
| "uris": null | |
| }, | |
| "TABREF1": { | |
| "type_str": "table", | |
| "num": null, | |
| "html": null, | |
| "content": "<table/>", | |
| "text": "Novice and Expert Prompts" | |
| }, | |
| "TABREF3": { | |
| "type_str": "table", | |
| "num": null, | |
| "html": null, | |
| "content": "<table><tr><td>Speech Command</td><td>Family</td></tr><tr><td>play</td><td>setMusic</td></tr><tr><td>scan</td><td>setMusic</td></tr><tr><td>repeat</td><td>setMusic</td></tr><tr><td>random</td><td>setMusic</td></tr></table>", | |
| "text": "shows an example for a possible task family, setMusic." | |
| }, | |
| "TABREF4": { | |
| "type_str": "table", | |
| "num": null, | |
| "html": null, | |
| "content": "<table/>", | |
| "text": "" | |
| }, | |
| "TABREF6": { | |
| "type_str": "table", | |
| "num": null, | |
| "html": null, | |
| "content": "<table/>", | |
| "text": "Classification of the System Utterances: Examples able voice commands to help the user to carry on with the dialogue." | |
| } | |
| } | |
| } | |
| } |