The DOI matches its Crossref record; 18 of the 20 references it cites were found in a registry or an index; 2 of the 11 sources its claims cite were checked for support.
Enabling robust and fluid spoken dialogue with cognitively impaired users
Open accessAbstract
We present the flexdiam dialogue management architecture, which was developed in a series of projects dedicated to tailoring spoken interaction to the needs of users with cognitive impairments in an everyday assistive domain, using a multimodal front-end. This hybrid DM architecture affords incremental processing of uncertain input, a flexible, mixed-initiative information grounding process that can be adapted to users' cognitive capacities and interactive idiosyncrasies, and generic mechanisms that foster transitions in the joint discourse state that are understandable and controllable by those users, in order to effect a robust interaction for users with varying capacities.
Overview
CoChat Summary
## TL;DR
The paper presents flexdiam, a hybrid dialogue management architecture for spoken interaction with cognitively impaired users that emphasizes incremental processing, uncertainty handling, and flexible mixed-initiative grounding. An initial evaluation with 44 participants found that all groups could complete the scheduling task, and the system generally achieved successful grounding in most cases.
## Key findings
- flexdiam is designed to support robust spoken dialogue by combining incremental processing, uncertainty representation, and flexible grounding tailored to users’ cognitive capacities and interaction styles.
- In the initial evaluation, the study recruited 44 participants: 19 older adults (SEN), 15 cognitively impaired adults (CIM), and 10 university controls (CTL).
- All subjects managed to enter at least the required number of appointments; negotiated entries ranged from 5 to 18, and final entries averaged 10.4 (CTL), 8.5 (SEN), and 8.9 (CIM).
- Older adults spent 15% longer on average per topic than controls, while the cognitively impaired group spent 23% longer; some CIM participants had long hesitations of up to tens of seconds.
## Methods
- The paper proposes the flexdiam dialogue management framework, built on the IPAACA middleware, with hierarchical “Issues,” incremental input handling, and explicit support for uncertainty, revisions, and mixed initiative.
- The initial evaluation used a spoken scheduling task in a fictional weekly calendar domain, with a virtual assistant and graphical calendar; participants were asked to enter at least five items.
- The evaluation involved 44 participants split into older adults, cognitively impaired adults, and university controls; analysis reported descriptive task-performance measures and qualitative observations of repair, revision, and metacommunicative behavior.
## Limitations
- The paper states that the evaluation is initial and that detailed conversational analysis had only recently started; a statistical description of language use and word error rates was still pending at the time of writing.
- The paper notes that the architecture’s future learning-based adaptation and deeper NLU were still in development, including better constraining of candidate interpretations.
- Evident from the text: the study focuses on a single constrained scheduling domain, so generalization beyond that domain is not established in the paper.
Model-written summary. Check it against the paper before relying on it.
Integrity signals
Registry
Cites
Statistics
Verification
Claims the paper makes
2 Valid · 0 Not valid · 8 Not checked · 36 Own results and interpretations
The cited sources support only part of this claim — read each source’s verdict below.
In the paper
“More recently, Wargnier et al. (2016) have evaluated a low-level attention monitoring and management module with a small sample of older adults with mild cognitive impairment; their system performed as well as with the control group.”
What the sources say
- Partly supportedcites Wargnier et al., 2016 sourceExperimental — accuracy not yet measured
Section IV (p. 5) describes a small sample that included three participants with mild cognitive impairment, three with Alzheimer’s disease, and two without cognitive impairment. Section C (p. 6) reports no significant difference in attention-estimator performance between the older participants as a whole and healthy younger adults tested previously. It does not establish equivalent performance specifically for the MCI participants versus a control group. (LLM, conf=0.88)
plain- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: unclear — It is unclear whether the paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
- Partly supported
The cited sources support only part of this claim — read each source’s verdict below.
In the paper
“The two latter teams also mentioned that interactions were unsuccessful for only their respective participant with the most overtly noticeable impairments.”
What the sources say
- Not checked — no full text of the sourcecites Yaghoubzadeh et al., 2015 source
- Partly supportedcites Wargnier et al., 2016 sourceExperimental — accuracy not yet measured
In B. Results (p. 6), the paper says the participant who could not complete the interaction had the highest cognitive impairment in its cohort. It also reports a participant with very poor hearing who could interact successfully when helped. B. Conversation Management (p. 4) says all but one participant interacted successfully in the other team's study, but does not say that participant had the most noticeable impairments. (LLM, conf=0.85)
plainstates an association- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: unclear — It is unclear whether the paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
A result the paper reports from its own data — it cites no other source for it, so there is none to check. The quote shows it is in the paper.
In the paper
“subjects were generally able to solve the set task,”
plain- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: passes — The paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
A result the paper reports from its own data — it cites no other source for it, so there is none to check. The quote shows it is in the paper.
In the paper
“The older adults spent 15% longer on average on a topic compared to controls,”
plainstates an association- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: unclear — It is unclear whether the paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
A result the paper states — it cites no source for it, so there is none to check. The quote shows it is in the paper.
In the paper
“All subjects managed to enter at least the required number of appointments into the calendar.”
plain- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: unclear — It is unclear whether the paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
14 more claims
Create a free account to see all 46 claims and every source verdict.
- It may state two separate claims as one. 14
- Probably not a substantial claim. 23
Citations
Cites
Clark, H. H., & Schaefer, E. F. (1989) · Cognitive Science · doi:10.1207/s15516709cog1302_7 · Cited by 1,699 · cited 1×
“Clark and Schaefer (1989) stated that positive evidence for understanding generally arrives in five cate-”
BackgroundGinzburg, J. (2012) · doi:10.1093/acprof:oso/9780199697922.001.0001 · Cited by 389 · cited 1×
“…leaves, see above), stored together with access time information to produce an implicit priority queue, similar to the ‘partially ordered set’ in Ginzburg (2012); however, rank is defined solely at invocation time since locally estimated utility is factored in. Invocations that trigger…”
BackgroundThe Information State Approach to Dialogue Management
Traum, D. R., & Larsson, S. (2003) · Text, Speech and Language Technology · doi:10.1007/978-94-010-0019-2_15 · Cited by 342 · cited 1×
“With the information-state-update (ISU) approach, Traum and Larsson (2003) proposed a generic mechanism for the concurrent matching of a set of update rules to the current state of a globally accessible information blackboard – in contrast to plan-based or finite state machine-based approaches.…”
MethodA randomized controlled trial of an automated exercise coach for older adults
Bickmore, T. W., Silliman, R. A., Nelson, K., Cheng, D. M., Winter, M., Henault, L., & Paasche‐Orlow, M. K. (2013) · Journal of the American Geriatrics Society · doi:10.1111/jgs.12449 · Cited by 305 · cited 1×
“Bickmore et al. (2013) analyzed month-long phases of interactions of older adults with a per-sonified exercise coaching system – it used spoken language, but user input was selected from sets of touchscreen buttons.”
The RavenClaw dialog management framework: Architecture and systems
Bohus, D., & Rudnicky, A. I. (2009) · Computer Speech & Language · doi:10.1016/j.csl.2008.10.001 · Cited by 264 · cited 1×
“Bohus and Rudnicky (2009), with RavenClaw, proposed an logic-based approach that separated the task domain model, provided in a domain spe-cific language to yield a hierarchical description of tasks and dependent subtasks, and a generic dialogue engine, configured with the task model and capable of…”
BackgroundIssue-based Dialogue Management
Larsson, S. (2002) · Cited by 232
Young, V., & Mihailidis, A. (2010) · Assistive Technology · doi:10.1080/10400435.2010.483646 · Cited by 149 · cited 1×
“There is regular comorbidity with impaired articulation which can complicate interactions (Young and Mihailidis, 2010) – although for mild cases automated speech recognition software has caught up in recent years to ensure suitable”
BackgroundIrisTK: a statechart-based toolkit for multi-party face-to-face interaction
Skantze, G., & Moubayed, S. A. (2012) · INTERNATIONAL CONFERENCE ON MULTIMODAL INTERACTION · doi:10.1145/2388676.2388698 · Cited by 119 · cited 1×
“Skantze and Moubayed (2012), with IrisTK, presented another hybrid approach that combined a generic ‘attention manager’ with a hierarchical task and dialogue model (IrisFlow) based on a generalized, extended version of Harel statecharts, which can be conveniently authored.”
BackgroundA General, Abstract Model of Incremental Dialogue Processing
Schlangen, D., & Skantze, G. (2011) · Dialogue & Discourse · doi:10.5087/dad.2011.105 · Cited by 102
Virtual Agents as Daily Assistants for Elderly or Cognitively Impaired People
Yaghoubzadeh, R., Kramer, M., Pitsch, K., & Kopp, S. (2013) · Lecture Notes in Computer Science · doi:10.1007/978-3-642-40415-3_7 · Cited by 98 · cited 3×
“ing and lead to incomplete appreciation and inadequate reflection of the contents (Yaghoubzadeh et al., 2013).”
BackgroundInfluentialSelf-citation
Cited by
Conversational Assistants for Elderly Users – The Importance of Socially Cooperative Dialogue
Kopp, S., Brandt, M., Buschmeier, H., Cyra, K., Freigang, F., Krämer, N. C., Kümmert, F., Opfermann, C., Pitsch, K., Schillingmann, L., et al. (2018) · PUB – Publications at Bielefeld University (Bielefeld University) · Cited by 51
Self-citationTokunaga, S., Tamura, K., & Otake, M. (2021) · Frontiers in Robotics and AI · doi:10.3389/frobt.2021.644964 · Cited by 28
Bevilacqua, R., Margaritini, A., Cucchieri, G., Giammarchi, C., Amabili, G., & Maranesi, E. (2023) · Studies in big data · doi:10.1007/978-981-99-3989-3_9 · Cited by 1
Confidence in uncertainty: Error cost and commitment in early speech hypotheses
Loth, S., Jettka, K., Giuliani, M., Kopp, S., & de Ruiter, J. P. (2018) · PLoS ONE · doi:10.1371/journal.pone.0201516 · Cited by 2
Self-citation
Metadata
- Authors
- Yaghoubzadeh, R., & Kopp, S.
- Affiliations
- Bielefeld University (DE)
- ORCID
- Kopp, Stefan
- Venue
- Proceedings of the 18th Annual SIGdial Meeting on Discourse and Dialogue
- Published
- 2017
- Pages
- 273-283
- OpenAlex
- W2759853919
- Cited by
- 5
- References
- 20
- Citations by year
- 2023: 1 · 2022: 1 · 2021: 1 · 2018: 2
- FWCI
- 0.25
- Type
- conference paper
- Language
- en
- Open access
- hybrid
- doi.org
- Licence
- cc-by
- Locations
- Annual Meeting of the Special Interest Group on Discourse and Dialogue (publishedVersion, cc-by)
- PUB – Publications at Bielefeld University (Bielefeld University) (submittedVersion, other-oa)
- Funders
- Deutsche Forschungsgemeinschaft
- Bundesministerium für Bildung und Forschung (FKZ 16SV7271K)
- Topics
- Physical Sciences › Computer Science › Artificial Intelligence › Speech and dialogue systems
- Health Sciences › Health Professions › Occupational Therapy › Assistive Technology in Communication and Mobility
- Social Sciences › Psychology › Social Psychology › Social Robot Interaction and HRI
- Keywords
- spoken dialogue systems, cognitive impairment, assistive technology, mixed-initiative interaction
About these checks
Registry, retraction, Cites and statistics results are computed, not judged, with one exception: a printed reference that no index lists is read once by a model before it counts as a citation error. Claims are extracted by a model and kept only when their quote appears verbatim in the text; each cited source is then checked, with a model’s help, for whether it says what the claim relies on. Those verdicts can be wrong — read the claim’s quote and the cited source before relying on one.
The claim verdicts this report counts are experimental: their accuracy is not yet measured.
A language model drew the claims from the paper on 2026-10-10, under version 3 of the claim definition (model deepseek-v4.1-flash, prompt extract-v3).
A second model labelled each claim on 2026-10-10 (model jev-1.13-20260917, definition revision gold-guideline-v0.2).
Cited sources are checked with a language model’s help (gpt-6-sol).
Statistics and arithmetic are recomputed, not judged.
Confidence bands are shown only for fields calibrated on our annotated test set. None is calibrated yet, so no band is shown.
What counts as a claim ›How we measure confidence ›How we check this ›
Are you an author, or do you see a problem with this page? Request a review
How to cite
Cite CoChat for the checks on this page, and the paper for its own findings.
- This page
- CoChat. (2026). Citation check: Enabling robust and fluid spoken dialogue with cognitively impaired users. https://app.cochat.ai/papers/10.18653/v1/w17-5533
- The paper
- Yaghoubzadeh, R., & Kopp, S. (2017). Enabling robust and fluid spoken dialogue with cognitively impaired users. Proceedings of the 18th Annual SIGdial Meeting on Discourse and Dialogue. https://doi.org/10.18653/v1/w17-5533
Data sources
Bibliographic data: Semantic Scholar, OpenAlex, Crossref.
Contains information from the Semantic Scholar Academic Graph Datasets, made available under the ODC Attribution License (ODC-BY 1.0).
Semantic Scholar’s data is described in: Kinney, R., et al. (2023). The Semantic Scholar Open Data Platform. arXiv:2301.10140.