Wolfgang Maier
Research Subjects
Wolfgang Maier has a total of 9 co-authors. Here are the top co-authors by number of shared publications:
- Stefan Ultes
Faculty of Information Systems and Applied Computer Sciences, University of Bamberg, Germany
2 shared publications - Wolfgang Minker
Ulm University, Ulm, Germany
2 shared publications - Hans-Ulrich Wittchen
Department of Psychiatry and Psychotherapy, University Hospital, Ludwig-Maximilians-University Munich, Munich, Germany
1 shared publication K. Barkow
Department of Psychiatry, University of Bonn, Germany
1 shared publication- Liu Ye
Science and Technology on Aerospace Flight Dynamic Laboratory
1 shared publication M. Gänsicke
From the Department of Psychiatry, University of Bonn and Max-Planck Institute of Psychiatry, Munich, Germany; and World Health Organization, Geneva, Switzerland
1 shared publication
Latest Publications
- Open Access
- 19 July 2023
- Published by Association for Computing Machinery (ACM)
- p. 1-9
Unified Conversational Models with System-Initiated Transitions between Chit-Chat and Task-Oriented Dialogues
Spoken dialogue systems (SDSs) have been separately developed under two different categories, task-oriented and chit-chat. The former focuses on achieving functional goals and the latter aims at creating engaging social conversations without special goals. Creating a unified conversational model that can engage in both chit-chat and task-oriented dialogue is a promising research topic in recent years. However, the potential “initiative” that occurs when there is a change between dialogue modes in one dialogue has rarely been explored. In this work, we investigate two kinds of dialogue scenarios, one starts from chit-chat implicitly involving task-related topics and finally switching to task-oriented requests; the other starts from task-oriented interaction and eventually changes to casual chat after all requested information is provided. We contribute two efficient prompt models which can proactively generate a transition sentence to trigger system-initiated transitions in a unified dialogue model. One is a discrete prompt model trained with two discrete tokens, the other one is a continuous prompt model using continuous prompt embeddings automatically generated by a classifier. We furthermore show that the continuous prompt model can also be used to guide the proactive transitions between particular domains in a multi-domain task-oriented setting.
- 19 July 2023
- Published by Association for Computing Machinery (ACM)
- p. 1-9
Open AccessUnified Conversational Models with System-Initiated Transitions between Chit-Chat and Task-Oriented Dialogues
Spoken dialogue systems (SDSs) have been separately developed under two different categories, task-oriented and chit-chat. The former focuses on achieving functional goals and the latter aims at creating engaging social conversations without special goals. Creating a unified conversational model that can engage in both chit-chat and task-oriented dialogue is a promising research topic in recent years. However, the potential “initiative” that occurs when there is a change between dialogue modes in one dialogue has rarely been explored. In this work, we investigate two kinds of dialogue scenarios, one starts from chit-chat implicitly involving task-related topics and finally switching to task-oriented requests; the other starts from task-oriented interaction and eventually changes to casual chat after all requested information is provided. We contribute two efficient prompt models which can proactively generate a transition sentence to trigger system-initiated transitions in a unified dialogue model. One is a discrete prompt model trained with two discrete tokens, the other one is a continuous prompt model using continuous prompt embeddings automatically generated by a classifier. We furthermore show that the continuous prompt model can also be used to guide the proactive transitions between particular domains in a multi-domain task-oriented setting.
- Open Access
- 1 January 2023
- Published by Association for Computational Linguistics (ACL)
- p. 279-292
System-Initiated Transitions from Chit-Chat to Task-Oriented Dialogues with Transition Info Extractor and Transition Sentence Generator
Ye Liu, Stefan Ultes, Wolfgang Minker, Wolfgang Maier. Proceedings of the 16th International Natural Language Generation Conference. 2023.
- 1 January 2023
- Published by Association for Computational Linguistics (ACL)
- p. 279-292
Open AccessSystem-Initiated Transitions from Chit-Chat to Task-Oriented Dialogues with Transition Info Extractor and Transition Sentence Generator
Ye Liu, Stefan Ultes, Wolfgang Minker, Wolfgang Maier. Proceedings of the 16th International Natural Language Generation Conference. 2023.
- Open Access
- 1 May 2002
- Published by Cambridge University Press (CUP) in Psychological Medicine
- Vol. 32 (4) , 595-607
Risk factors for new depressive episodes in primary health care: an international prospective 12-month follow-up study
Background. Studies that examined community samples have reported several risk factors for the development of depressive episodes. The few studies that have been performed on primary care samples were mostly cross-sectional. Most samples had originated from highly developed industrial countries. This is the first study that prospectively investigates the risk factors of depressive episodes in an international primary care sample.Methods. A stratified primary care sample of initially non-depressed subjects (N = 2445) from 15 centres from all over the world was examined for the presence or absence of a depressive episode (ICD-10) at the 12 month follow-up assessment. The initial measures addressed sociodemographic variables, psychological/psychiatric problems and social disability. Logistic regression analysis was carried out to determine their relationship with the development of new depressive episodes.Results. At the 12-month follow-up, 4·4% of primary care patients met ICD-10 criteria for a depressive episode. Logistic regression analysis revealed that the recognition by the general practitioner as a psychiatric case, repeated suicidal thoughts, previous depressive episodes, the number of chronic organic diseases, poor general health, and a full or subthreshold ICD-10 disorder were related to the development of new depressive episodes.Conclusions. Psychological/psychiatric problems were found to play the most important role in the prediction of depressive episodes while sociodemographic variables were of lower importance. Differences compared with other studies might be due to our prospective design and possibly also to our culturally different sample. Applied stratification procedures, which resulted in a sample at high risk of developing depression, might be a limitation of our study.
- 1 May 2002
- Published by Cambridge University Press (CUP) in Psychological Medicine
- Vol. 32 (4) , 595-607
Open AccessRisk factors for new depressive episodes in primary health care: an international prospective 12-month follow-up study
Background. Studies that examined community samples have reported several risk factors for the development of depressive episodes. The few studies that have been performed on primary care samples were mostly cross-sectional. Most samples had originated from highly developed industrial countries. This is the first study that prospectively investigates the risk factors of depressive episodes in an international primary care sample.Methods. A stratified primary care sample of initially non-depressed subjects (N = 2445) from 15 centres from all over the world was examined for the presence or absence of a depressive episode (ICD-10) at the 12 month follow-up assessment. The initial measures addressed sociodemographic variables, psychological/psychiatric problems and social disability. Logistic regression analysis was carried out to determine their relationship with the development of new depressive episodes.Results. At the 12-month follow-up, 4·4% of primary care patients met ICD-10 criteria for a depressive episode. Logistic regression analysis revealed that the recognition by the general practitioner as a psychiatric case, repeated suicidal thoughts, previous depressive episodes, the number of chronic organic diseases, poor general health, and a full or subthreshold ICD-10 disorder were related to the development of new depressive episodes.Conclusions. Psychological/psychiatric problems were found to play the most important role in the prediction of depressive episodes while sociodemographic variables were of lower importance. Differences compared with other studies might be due to our prospective design and possibly also to our culturally different sample. Applied stratification procedures, which resulted in a sample at high risk of developing depression, might be a limitation of our study.