pith. machine review for the scientific record. sign in

arxiv: 2605.14207 · v1 · submitted 2026-05-14 · 💻 cs.HC

Recognition: no theorem link

What Should Explanations Contain? A Human-Centered Explanation Content Model for Local, Post-Hoc Explanations

Helmut Degen
Authors on Pith no claims yet

Pith reviewed 2026-05-15 02:47 UTC · model grok-4.3

classification 💻 cs.HC
keywords explanation contentlocal explanationspost-hoc explanationshuman-centered XAIqualitative content analysisindustrial AIuser studiescontent model
0
0 comments X

The pith

A fourteen-code model derived from industrial user studies specifies what content local post-hoc explanations should contain.

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The paper derives a structured model of explanation content by analyzing 325 meaning units from six user studies across building technology, manufacturing, AI development, and hospital cybersecurity. An inductive phase first produced twelve codes, after which two additional codes were added based on existing XAI system architectures to reach a final fourteen-code structure. These codes are grouped into rule-based, causal, epistemic (actual), and epistemic (similar) categories. Expert review confirmed content adequacy across relevance, boundary clarity, and understandability, while independent coding of a subsample produced high agreement scores. The resulting model supplies a concrete basis for deciding which categories of information explanations must include when supporting users of industrial AI systems.

Core claim

Through hybrid inductive-deductive qualitative content analysis of user data, the work establishes a fourteen-code explanation content model organized into four groups—rule-based, causal, epistemic (actual), and epistemic (similar)—with twelve codes directly grounded in the corpus and two added as theoretical extensions for completeness. An eleven-member expert panel rated all codes as adequate (I-CVI ≥ 0.82) with strong scale-level agreement on relevance, clarity, and understandability. Independent coding of a 25% stratified subsample by two researchers yielded Krippendorff’s α = 0.920 and Cohen’s κ = 0.920, confirming both content adequacy and coding reproducibility for the model.

What carries the argument

The fourteen-code explanation content model, which groups categories of information into rule-based, causal, and two epistemic types to capture what users need from local post-hoc explanations.

If this is right

  • Explanations can be elicited, specified, and evaluated using the fourteen codes as a shared vocabulary.
  • System designers gain a reproducible checklist for deciding which content categories to include in local explanations.
  • The four-group organization allows comparison of explanation approaches across different AI architectures.
  • The two theoretical codes highlight content that existing systems can already generate but that did not appear in the studied user data.

Where Pith is reading between the lines

These are editorial extensions of the paper, not claims the author makes directly.

  • The model could serve as a starting template for standardized explanation interfaces in new industrial domains.
  • Behavioral tests with actual end-users would show whether using the codes improves task performance or trust calibration.
  • Integration into XAI toolkits would let developers automatically generate explanations that cover the required content groups.
  • The distinction between actual and similar epistemic codes suggests separate design patterns for factual versus analogical explanations.

Load-bearing premise

The six user studies and 325 meaning units capture the explanation needs that will appear for end-users across the full range of industrial AI applications.

What would settle it

A new user study in an unexamined industrial domain that reveals frequent requests for explanation content types absent from the fourteen codes.

Figures

Figures reproduced from arXiv: 2605.14207 by Helmut Degen.

Figure 1
Figure 1. Figure 1: Analytic procedure The analysis proceeded in four phases (see [PITH_FULL_IMAGE:figures/full_fig_p014_1.png] view at source ↗
Figure 2
Figure 2. Figure 2: Explanation content model for local, post-hoc explanations (post-reliability, [PITH_FULL_IMAGE:figures/full_fig_p027_2.png] view at source ↗
read the original abstract

Which categories of explanation content are relevant for users of industrial AI systems, and how can those categories be organized for local, post-hoc explanations? To address these questions, a hybrid inductive-deductive qualitative content analysis was applied to 325 meaning units drawn from six user studies in building technology, manufacturing, AI software development, and hospital cybersecurity. The inductive phase produced an initial twelve-code structure. A theory-informed coverage assessment and expert review then added two further codes, Rule base and What-if backward, that were not instantiated in the corpus but correspond to system architectures documented in the XAI literature. The resulting fourteen-code model is organized into four groups: rule-based, causal, epistemic (actual), and epistemic (similar), with twelve codes grounded in the corpus and two as theoretical extensions. An eleven-member expert panel supported the content adequacy of all codes (I-CVI $\geq$ 0.82; scale-level agreement of 0.93 for relevance, 0.92 for boundary clarity, and 0.94 for understandability). A stratified subsample of 82 units (25\% of the corpus), coded independently by two researchers using the finalized codebook, yielded Krippendorff's $\alpha = 0.920$ and Cohen's $\kappa = 0.920$. The paper therefore establishes content adequacy and coding reproducibility for a content-level explanation model intended to support elicitation, specification, and later evaluation of explanation content in industrial AI systems. Behavioral validation of downstream effects remains future work.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, simulated authors' rebuttal, and a circularity audit. Tearing a paper down is the easy half of reading it; the pith above is the substance, this is the friction.

Referee Report

2 major / 2 minor

Summary. The paper proposes a fourteen-code model for the content of local, post-hoc explanations in industrial AI systems. Derived via hybrid inductive-deductive qualitative content analysis of 325 meaning units from six user studies in building technology, manufacturing, AI software development, and hospital cybersecurity, the model organizes codes into four groups (rule-based, causal, epistemic (actual), epistemic (similar)). Twelve codes are grounded in the corpus; two (Rule base, What-if backward) are added as theoretical extensions from XAI literature. Content adequacy is supported by an eleven-expert panel (I-CVI ≥ 0.82; scale-level agreements 0.93 relevance, 0.92 boundary clarity, 0.94 understandability) and high inter-coder reliability (Krippendorff's α = 0.920, Cohen's κ = 0.920 on a 25% stratified subsample). The work claims to establish content adequacy and coding reproducibility to support elicitation, specification, and evaluation of explanations, with behavioral validation noted as future work.

Significance. If the adequacy claim holds, the model offers a structured, human-centered framework for specifying explanation content in industrial XAI, moving beyond ad-hoc designs toward reproducible elicitation and evaluation. Strengths include the transparent hybrid method, direct grounding in user-derived meaning units, expert validation with strong CVI metrics, and explicit reproducibility evidence via high α/κ values. This could inform practical XAI deployment in the studied domains while highlighting the need for downstream behavioral tests.

major comments (2)
  1. [Abstract] Abstract: The central claim that the fourteen-code model supports elicitation and specification 'in industrial AI systems' is load-bearing for the paper's scope, yet rests on six studies limited to four domains (building technology, manufacturing, AI software, hospital cybersecurity). The representativeness assumption for broader industrial contexts is not directly tested with end-users in deployment, which risks overgeneralizing the adequacy findings despite the paper's correct note on future behavioral validation.
  2. [Expert panel validation] Expert panel validation section: While I-CVI ≥ 0.82 and scale-level agreements are reported for all codes, the eleven-member panel's judgments on content adequacy (relevance, boundary clarity, understandability) may diverge from actual end-user perceptions in the original studies; the paper should clarify how panel expertise maps to the user populations sampled.
minor comments (2)
  1. [Methods] Clarify in the methods how the two theory-derived codes were integrated into the codebook without corpus instantiation, including any adjustments to boundary definitions during expert review.
  2. [Results] The abstract and results report Krippendorff's α and Cohen's κ both as 0.920 on the 82-unit subsample; confirm whether these are independent calculations or if one is derived from the other for transparency.

Simulated Author's Rebuttal

2 responses · 0 unresolved

We thank the referee for the constructive feedback and positive assessment of the work. We address the major comments point by point below and have revised the manuscript accordingly.

read point-by-point responses

  1. Referee: [Abstract] Abstract: The central claim that the fourteen-code model supports elicitation and specification 'in industrial AI systems' is load-bearing for the paper's scope, yet rests on six studies limited to four domains (building technology, manufacturing, AI software, hospital cybersecurity). The representativeness assumption for broader industrial contexts is not directly tested with end-users in deployment, which risks overgeneralizing the adequacy findings despite the paper's correct note on future behavioral validation.

    Authors: We agree that the scope requires careful qualification. The studies cover four distinct industrial domains, but we have revised the abstract and introduction to explicitly state that the model is derived from user studies in building technology, manufacturing, AI software development, and hospital cybersecurity. We now describe it as a framework intended to support elicitation and specification in industrial AI systems, while adding a sentence noting that broader generalizability requires further validation across additional domains and deployment settings. The limitations section has also been expanded to discuss this point. revision: yes

  2. Referee: [Expert panel validation] Expert panel validation section: While I-CVI ≥ 0.82 and scale-level agreements are reported for all codes, the eleven-member panel's judgments on content adequacy (relevance, boundary clarity, understandability) may diverge from actual end-user perceptions in the original studies; the paper should clarify how panel expertise maps to the user populations sampled.

    Authors: We thank the referee for highlighting this clarification need. The eleven experts were selected for their combined expertise in XAI, HCI, and the specific domains represented in the studies (e.g., manufacturing engineers and cybersecurity specialists). We have added a dedicated paragraph in the expert validation section that details panel member backgrounds and explicitly maps their domain experience to the user populations sampled in the six studies, thereby strengthening the link between panel judgments and the original user data. revision: yes

Circularity Check

0 steps flagged

No significant circularity in the model construction process

full rationale

The paper constructs its fourteen-code explanation content model through a hybrid inductive-deductive qualitative content analysis of 325 meaning units drawn from six independent user studies across distinct domains. The inductive phase directly yields twelve codes from the corpus data, with two additional codes added via coverage assessment against external XAI literature rather than self-citation or prior author work. Expert panel validation (I-CVI and scale agreements) and inter-coder reliability metrics (Krippendorff's α and Cohen's κ on a subsample) provide independent checks. No load-bearing step reduces to self-definition, fitted inputs renamed as predictions, uniqueness theorems from the same authors, or renaming of known results; the chain remains grounded in empirical data and external theory without circular reduction.

Axiom & Free-Parameter Ledger

0 free parameters · 1 axioms · 0 invented entities

The central claim rests on standard assumptions of qualitative content analysis rather than new parameters or entities.

axioms (1)
  • domain assumption Qualitative content analysis of user statements can produce a generalizable content model for AI explanations
    Invoked in the hybrid inductive-deductive analysis of the 325 meaning units and expert validation.

pith-pipeline@v0.9.0 · 5573 in / 1256 out tokens · 36059 ms · 2026-05-15T02:47:02.701492+00:00 · methodology

discussion (0)

Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.

Reference graph

Works this paper leans on

158 extracted references · 158 canonical work pages · 2 internal anchors

  1. [1]

    Case-based reasoning: Foundational issues, methodological variations, and system approaches

    Aamodt, A., Plaza, E., 1994. Case-based reasoning: Foundational issues, methodological variations, and system approaches. AI Commu- nications 7, 39–59

    work page 1994

  2. [2]

    Abdul, A., von der Weth, C., Kankanhalli, M., Lim, B.Y., 2020. COGAM: Measuring and Moderating Cognitive Load in Machine Learning Model Explanations, in: Proceedings of the 2020 CHI Con- ference on Human Factors in Computing Systems (CHI ’20), Asso- ciation for Computing Machinery, New York, NY, USA. pp. 1–14. doi:10.1145/3313831.3376615

    work page doi:10.1145/3313831.3376615 2020

  3. [3]

    Adhikari, A., Tax, D.M.J., Satta, R., Faeth, M., 2019. LEAFAGE: Example-Based and Feature-Importance-Based Explanations for Black-Box ML Models, in: Proceedings of the 2019 IEEE International Conference on Fuzzy Systems (FUZZ-IEEE), IEEE, Piscataway, NJ, USA. pp. 1–7. doi:10.1109/FUZZ-IEEE.2019.8858846

    work page doi:10.1109/fuzz-ieee.2019.8858846 2019

  4. [4]

    UP- GRADE: Universal Patching and Remediation for Autonomous De- fense.https://arpa-h.gov/explore-funding/programs/upgrade

    Advanced Research Projects Agency for Health (ARPA-H), 2024. UP- GRADE: Universal Patching and Remediation for Autonomous De- fense.https://arpa-h.gov/explore-funding/programs/upgrade. Accessed: 2026-04-05

    work page 2024

  5. [5]

    Aechtner, J., Cabrera, L., Katwal, D., Onghena, P., Valenzuela, D.P., Wilbik, A., 2022. Comparing User Perception of Explanations Devel- oped with XAI Methods, in: Proceedings of the 2022 IEEE Interna- tional Conference on Fuzzy Systems (FUZZ-IEEE), IEEE, Piscataway, NJ, USA. pp. 1–7. doi:10.1109/FUZZ-IEEE55066.2022.9882743

    work page doi:10.1109/fuzz-ieee55066.2022.9882743 2022

  6. [6]

    Contestable AI by Design: Towards a Framework

    Alfrink, K., Keller, I., Kortuem, G., Doorn, N., 2023. Contestable AI by Design: Towards a Framework. Minds and Machines 33, 613–639. doi:10.1007/s11023-022-09611-z

    work page doi:10.1007/s11023-022-09611-z 2023

  7. [7]

    Does Explainable Artificial Intelligence Improve Human 37 Decision-Making? Proceedings of the AAAI Conference on Artificial Intelligence 35, 6618–6626

    Alufaisan, Y., Marusich, L.R., Bakdash, J.Z., Zhou, Y., Kantarcioglu, M., 2021. Does Explainable Artificial Intelligence Improve Human 37 Decision-Making? Proceedings of the AAAI Conference on Artificial Intelligence 35, 6618–6626. doi:10.1609/aaai.v35i8.16819

    work page doi:10.1609/aaai.v35i8.16819 2021

  8. [8]

    9 From Human-Level AI Tales to AI Leveling Human Scales pone.0231627

    Anjara, S.G., Janik, A., Dunford-Stenger, A., Mc Kenzie, K., Collazo- Lorduy, A., Torrente, M., Costabello, L., Provencio, M., 2023. Ex- amining Explainable Clinical Decision Support Systems with Think Aloud Protocols. PLoS One 18, e0291443. doi:10.1371/journal. pone.0291443

    work page doi:10.1371/journal 2023

  9. [9]

    SoK: Pragmatic Assess- ment of Machine Learning for Network Intrusion Detection, in: Pro- ceedings of the IEEE European Symposium on Security and Privacy, IEEE, Delft, Netherlands

    Apruzzese, G., Laskov, P., Schneider, J., 2023. SoK: Pragmatic Assess- ment of Machine Learning for Network Intrusion Detection, in: Pro- ceedings of the IEEE European Symposium on Security and Privacy, IEEE, Delft, Netherlands. pp. 1–19. doi:10.1109/EuroSP57164.2023. 00016

    work page doi:10.1109/eurosp57164.2023 2023

  10. [10]

    Investigating Explanations in Conditional and Highly Automated Driving: The Effects of Situation Awareness and Modality

    Avetisyan, L., Ayoub, J., Zhou, F., 2022. Investigating Explanations in Conditional and Highly Automated Driving: The Effects of Situation Awareness and Modality. Transportation Research Part F: Traffic Psy- chologyandBehaviour89, 456–466. doi:10.1016/j.trf.2022.07.010

    work page doi:10.1016/j.trf.2022.07.010 2022

  11. [11]

    From human factors to human actors: The role of psychology and human-computer interaction studies in system design, in: Greenbaum, J., Kyng, M

    Bannon, L.J., 1991. From human factors to human actors: The role of psychology and human-computer interaction studies in system design, in: Greenbaum, J., Kyng, M. (Eds.), Design at Work: Cooperative Design of Computer Systems. Lawrence Erlbaum Associates, Hillsdale, NJ, pp. 25–44. URL:https://doi.org/10.1201/9781315800349-4, doi:10.1201/9781315800349-4

    work page doi:10.1201/9781315800349-4 1991

  12. [12]

    Bansal, G., Wu, T., Zhou, J., Fok, R., Nushi, B., Kamar, E., Ribeiro, M.T., Weld, D., 2021. Does the Whole Exceed Its Parts? The Effect of AI Explanations on Complementary Team Performance, in: Pro- ceedings of the 2021 CHI Conference on Human Factors in Computing Systems, Association for Computing Machinery, New York, NY, USA. pp. 81:1–81:16. doi:10.1145...

    work page doi:10.1145/3411764.3445717 2021

  13. [13]

    The Role of Domain Ex- pertise in Trusting and Following Explainable AI Decision Support Systems

    Bayer, S., Gimpel, H., Markgraf, M., 2022. The Role of Domain Ex- pertise in Trusting and Following Explainable AI Decision Support Systems. Journal of Decision Systems 32, 110–138. doi:10.1080/ 12460125.2021.1958505. 38

    work page arXiv 2022

  14. [14]

    Ben David, D., Resheff, Y.S., Tron, T., 2021. Explainable AI and Adoption of Financial Algorithmic Advisors: An Experimental Study, in: Proceedings of the 2021 AAAI/ACM Conference on AI, Ethics, and Society (AIES ’21), Association for Computing Machinery, New York, NY, USA. pp. 390–400. doi:10.1145/3461702.3462565

    work page doi:10.1145/3461702.3462565 2021

  15. [15]

    Bertrand, A., Eagan, J.R., Maxwell, W., 2023. Questioning the Abil- ity of Feature-Based Explanations to Empower Non-Experts in Robo- Advised Financial Decision-Making, in: Proceedings of the 2023 ACM Conference on Fairness, Accountability, and Transparency (FAccT ’23), Association for Computing Machinery, New York, NY, USA. pp. 943–

    work page 2023

  16. [16]

    doi:10.1145/3593013.3594053

    work page doi:10.1145/3593013.3594053

  17. [17]

    Bhattacharya, A., Ooge, J., Stiglic, G., Verbert, K., 2023. Direc- tive Explanations for Monitoring the Risk of Diabetes Onset: In- troducing Directive Data-Centric Explanations and Combinations to Support What-If Explorations, in: Proceedings of the 28th Interna- tional Conference on Intelligent User Interfaces (IUI ’23), Associa- tion for Computing Mach...

    work page doi:10.1145/3581641.3584075 2023

  18. [18]

    What Is Wrong with Social Theory? American Sociological Review 19, 3–10

    Blumer, H., 1954. What Is Wrong with Social Theory? American Sociological Review 19, 3–10. doi:10.2307/2088165

    work page doi:10.2307/2088165 1954

  19. [19]

    User Trust and Understanding of Explainable AI: Exploring Algorithm Visualisations and User Biases, in: Kurosu, M

    Branley-Bell, D., Whitworth, R., Coventry, L., 2020. User Trust and Understanding of Explainable AI: Exploring Algorithm Visualisations and User Biases, in: Kurosu, M. (Ed.), Human-Computer Interaction: Human Values and Quality of Life, Springer International Publishing, Cham, Switzerland. pp. 382–399. doi:10.1007/978-3-030-49065-2\ _27

    work page doi:10.1007/978-3-030-49065-2 2020

  20. [20]

    Using Thematic Analysis in Psychol- ogy

    Braun, V., Clarke, V., 2006. Using Thematic Analysis in Psychol- ogy. Qualitative Research in Psychology 3, 77–101. doi:10.1191/ 1478088706qp063oa

    work page 2006

  21. [21]

    Thematic Analysis

    Braun, V., Clarke, V., 2019. Thematic Analysis. Counselling and Psychotherapy Research 19, 77–83. doi:10.1002/capr.12362

    work page doi:10.1002/capr.12362 2019

  22. [22]

    Assessing Perceived Trust and Satisfaction with Multiple Explanation Techniques in XAI- 39 Enhanced Learning Analytics

    Brdnik, S., Podgorelec, V., Šumak, B., 2023. Assessing Perceived Trust and Satisfaction with Multiple Explanation Techniques in XAI- 39 Enhanced Learning Analytics. Electronics 12, 2594. doi:10.3390/ electronics12122594

    work page 2023

  23. [23]

    Perception and the Representative Design of Psychological Experiments

    Brunswik, E., 1956. Perception and the Representative Design of Psychological Experiments. 2nd ed., University of Califor- nia Press, Berkeley, CA. URL:https://archive.org/details/ perceptionrepres0000brun

    work page 1956

  24. [24]

    Buçinca, Z., Lin, P., Gajos, K.Z., Glassman, E.L., 2020. Proxy Tasks and Subjective Measures Can Be Misleading in Evaluating Explainable AI Systems, in: Proceedings of the 25th International Conference on Intelligent User Interfaces, Association for Computing Machinery, New York, NY, USA. pp. 454–464. doi:10.1145/3377325.3377498

    work page doi:10.1145/3377325.3377498 2020

  25. [25]

    Bunde, E., 2021. AI-Assisted and Explainable Hate Speech De- tection for Social Media Moderators: A Design Science Approach, in: Proceedings of the 54th Hawaii International Conference on Sys- tem Sciences (HICSS), University of Hawai‘i at M¯ anoa. pp. 1–10. doi:10.24251/HICSS.2021.293

    work page doi:10.24251/hicss.2021.293 2021

  26. [26]

    Counterfactuals in Explainable Artificial In- telligence (XAI): Evidence from Human Reasoning

    Byrne, R.M.J., 2019. Counterfactuals in Explainable Artificial In- telligence (XAI): Evidence from Human Reasoning. Proceedings of the International Joint Conference on Artificial Intelligence , 6276– 6282doi:10.24963/ijcai.2019/876

    work page doi:10.24963/ijcai.2019/876 2019

  27. [27]

    Cachada, A., Barbosa, J., Leitão, P., Alves Teixeira, J., Teixeira, J.P., Moreira, A.H., Moreira, M., 2019. Maintenance 4.0: Intelli- gent and Predictive Maintenance System Architecture, in: Proceed- ings of the IEEE International Conference on Industrial Informat- ics, IEEE, Helsinki, Finland. pp. 173–178. doi:10.1109/INDIN41052. 2019.8972134

    work page doi:10.1109/indin41052 2019

  28. [28]

    Hello AI

    Cai, C.J., Winter, S., Steiner, D., Wilcox, L., Terry, M., 2019. "Hello AI": Uncovering the Onboarding Needs of Medical Practitioners for Human-AI Collaborative Decision-Making, in: Proceedings of the ACM Conference on Human Factors in Computing Systems, ACM, Glasgow, Scotland. pp. 1–24. doi:10.1145/3290605.3300333

    work page doi:10.1145/3290605.3300333 2019

  29. [29]

    Cau, F.M., Hauptmann, H., Spano, L.D., Tintarev, N., 2023. Support- ing High-Uncertainty Decisions Through AI and Logic-Style Explana- 40 tions, in: Proceedings of the 28th International Conference on Intelli- gent User Interfaces (IUI ’23), Association for Computing Machinery, New York, NY, USA. pp. 251–263. doi:10.1145/3581641.3584080

    work page doi:10.1145/3581641.3584080 2023

  30. [30]

    Explanation On- tology: A General-Purpose, Semantic Representation for Supporting User-Centered Explanations

    Chari, S., Seneviratne, O., Ghalwash, M., Shirai, S., Gruen, D.M., Meyer, P., Chakraborty, P., McGuinness, D.L., 2024. Explanation On- tology: A General-Purpose, Semantic Representation for Supporting User-Centered Explanations. Semantic Web doi:10.3233/SW-233282

    work page doi:10.3233/sw-233282 2024

  31. [31]

    A Model of Explanations for User-Centered AI, in: Allemang, D., Dumontier, M., Kellogg, G., Maleshkova, M., Pesquita, C., Stadler, C., Cruz, I.F

    Chari, S., Seneviratne, O., Gruen, D.M., Foreman, M.A., Das, A.K., McGuinness, D.L., 2020. A Model of Explanations for User-Centered AI, in: Allemang, D., Dumontier, M., Kellogg, G., Maleshkova, M., Pesquita, C., Stadler, C., Cruz, I.F. (Eds.), The Semantic Web – ISWC 2020. Springer, Cham, Switzerland, pp. 244–259. doi:10.1007/ 978-3-030-62466-8\_15

    work page 2020

  32. [32]

    Constructing Grounded Theory

    Charmaz, K., 2014. Constructing Grounded Theory. 2 ed., SAGE Publications, London, UK

    work page 2014

  33. [33]

    Shneiderman

    Chien, S.Y., Yang, C.J., Yu, F., 2022. XFlag: Explainable Fake News Detection Model on Social Media. International Journal of Hu- man–Computer Interaction 38, 1808–1827. doi:10.1080/10447318. 2022.2062113

    work page doi:10.1080/10447318 2022

  34. [34]

    Chromik, M., Butz, A., 2021. Human-XAI Interaction: A Review and Design Principles for Explanation User Interfaces, in: Proceedings of theACMInternationalConferenceonIntelligentUserInterfaces, ACM, College Station, TX, USA. pp. 1–13. doi:10.1145/3397481.3450686

    work page doi:10.1145/3397481.3450686 2021

  35. [35]

    Human-centered XAI: Developing design patterns for explanations of clinical decision support systems

    Chromik, M., Eiband, M., Buchner, F., Krüger, A., Butz, A., 2021. Human-centered XAI: Developing design patterns for explanations of clinical decision support systems. International Journal of Human- Computer Studies 154, 102684. doi:10.1016/j.ijhcs.2021.102684

    work page doi:10.1016/j.ijhcs.2021.102684 2021

  36. [36]

    A Coefficient of Agreement for Nominal Scales

    Cohen, J., 1960. A Coefficient of Agreement for Nominal Scales. Ed- ucational and Psychological Measurement 20, 37–46. doi:10.1177/ 001316446002000104

    work page 1960

  37. [37]

    Toward Person- alized XAI: A Case Study in Intelligent Tutoring Systems

    Conati, C., Barral, O., Putnam, V., Rieger, L., 2021. Toward Person- alized XAI: A Case Study in Intelligent Tutoring Systems. Artificial Intelligence 298, 103503. doi:10.1016/j.artint.2021.103503. 41

    work page doi:10.1016/j.artint.2021.103503 2021

  38. [38]

    Using Ontologies to Enhance Human Understandability of Global Post-hoc Explanations of Black-box Models

    Confalonieri, R., Weyde, T., Besold, T.R., Moscoso del Prado Martín, F., 2021. Using Ontologies to Enhance Human Understandability of Global Post-hoc Explanations of Black-box Models. Artificial Intelli- gence 296, 103471. doi:10.1016/j.artint.2021.103471

    work page doi:10.1016/j.artint.2021.103471 2021

  39. [39]

    The Effects of Explanations in Automated Essay Scoring Systems on Student Trust and Motivation

    Conijn, R., Kahr, P., Snijders, C., 2023. The Effects of Explanations in Automated Essay Scoring Systems on Student Trust and Motivation. Journal of Learning Analytics 10, 37–53. doi:10.18608/jla.2023. 7801

    work page doi:10.18608/jla.2023 2023

  40. [40]

    Explainable Activity Recognition for Smart Home Systems

    Das, D., Nishimura, Y., Vivek, R.P., Takeda, N., Fish, S.T., Plötz, T., Chernova, S., 2023. Explainable Activity Recognition for Smart Home Systems. ACM Transactions on Interactive Intelligent Systems 13, 7:1–7:39. doi:10.1145/3561533

    work page doi:10.1145/3561533 2023

  41. [41]

    Degen, H., Budnik, C., 2024. How to Explain it to System Testers? A Qualitative User Study About Understandability, Validatability, Pre- dictability, and Trustworthiness, in: Artificial Intelligence in HCI: 5th International Conference, AI-HCI 2024, Held as Part of the 26th HCI International Conference, HCII 2024, Washington, DC, USA, June 29 – July 4, 20...

    work page doi:10.1007/978-3-031-60606-9 2024

  42. [42]

    Degen, H., Budnik, C., Chitre, K., Lintereur, A., 2021. How to Explain it to Facility Managers? A Qualitative, Industrial User Research Study for Explainability, in: Stephanidis, C., Kurosu, M., Chen, J.Y.C., Fragomeni, G., Streitz, N., Konomi, S., Degen, H., Ntoa, S. (Eds.), HCI International 2021 - Late Breaking Papers: Multimodality, eX- tended Reality...

    work page doi:10.1007/978-3-030-90963-5 2021

  43. [43]

    Degen, H., Budnik, C., Conte, G., Lintereur, A., Weber, S., 2022. How to Explain it to Energy Engineers? A qualitative user study about trustworthiness, understandability, and actionability, in: Stephanidis, C., Kurosu, M., Chen, J.Y.C., Fragomeni, G., Streitz, N., Konomi, S., Degen, H., Ntoa, S. (Eds.), HCI International 2022 - Late Breaking Papers: Mult...

    work page doi:10.1007/978-3-031-21707-4 2022

  44. [44]

    Degen, H., Budnik, C., Gross, R., Rothering, M., 2023. How To Ex- plain It To a Model Manager? A Qualitative User Study About Un- derstandability, Trustworthiness, Actionability, and Action Efficacy, in: Artificial Intelligence in HCI: 4th International Conference, AI-HCI 2023, Held as Part of the 25th HCI International Conference, HCII 2023, Copenhagen, ...

    work page doi:10.1007/978-3-031-35891-3 2023

  45. [45]

    How to Explain It to Data Sci- entists?, in: Degen, H., Ntoa, S

    Degen, H., Min, Z., Nagaraja, P., 2025a. How to Explain It to Data Sci- entists?, in: Degen, H., Ntoa, S. (Eds.), Artificial Intelligence in HCI: 6th International Conference, AI-HCI 2025, Held as Part of the 27th HCI International Conference, HCII 2025, Gothenburg, Sweden, June 22 – 27, 2025, Proceedings, Part I, Springer Nature Switzerland AG, Cham, Swi...

    work page doi:10.1007/978-3-031-93412-4 2025

  46. [46]

    How to explain it to data sci- entists? A mixed-methods user study about explainable AI, using men- tal models for explanations.https://arxiv.org/abs/2502.16083

    Degen, H., Min, Z., Nagaraja, P., 2025b. How to explain it to data sci- entists? A mixed-methods user study about explainable AI, using men- tal models for explanations.https://arxiv.org/abs/2502.16083. arXiv:2502.16083

    work page arXiv

  47. [47]

    User-Centric Explainability in Fintech Applications, in: Stephanidis, C., Antona, M., Ntoa, S

    Deo, S., Sontakke, N., 2021. User-Centric Explainability in Fintech Applications, in: Stephanidis, C., Antona, M., Ntoa, S. (Eds.), HCI International 2021 – Posters, Springer International Publishing, Cham, Switzerland. pp. 481–488. doi:10.1007/978-3-030-78642-7\_64

    work page doi:10.1007/978-3-030-78642-7 2021

  48. [48]

    Towards a Rigorous Science of Inter- pretable Machine Learning, in: Proceedings of the ICML Workshop on Human Interpretability in Machine Learning, Sydney, Australia

    Doshi-Velez, F., Kim, B., 2017. Towards a Rigorous Science of Inter- pretable Machine Learning, in: Proceedings of the ICML Workshop on Human Interpretability in Machine Learning, Sydney, Australia

    work page 2017

  49. [49]

    Where the Action Is: The Foundations of Embodied Interaction

    Dourish, P., 2001. Where the Action Is: The Foundations of Embodied Interaction. MIT Press, Cambridge, MA. URL:https://doi.org/10. 7551/mitpress/7221.001.0001, doi:10.7551/mitpress/7221.001. 0001. 43

    work page doi:10.7551/mitpress/7221.001 2001

  50. [50]

    Human-centered explainable ai: Towards a reflective sociotechnical approach, in: Stephanidis, C., Kurosu, M., Degen, H., Reinerman-Jones, L

    Ehsan, U., Riedl, M.O., 2020. Human-centered explainable ai: Towards a reflective sociotechnical approach, in: Stephanidis, C., Kurosu, M., Degen, H., Reinerman-Jones, L. (Eds.), HCI International 2020 – Late Breaking Papers: Multimodality and Intelligence, Springer, Cham. pp. 449–466. URL:https://doi.org/10.1007/978-3-030-60117-1\ _33, doi:10.1007/978-3-...

    work page doi:10.1007/978-3-030-60117-1 2020

  51. [51]

    Chart- ing the sociotechnical gap in explainable ai: A framework to address the gap in xai

    Ehsan, U., Saha, K., De Choudhury, M., Riedl, M.O., 2023. Chart- ing the sociotechnical gap in explainable ai: A framework to address the gap in xai. Proceedings of the ACM on Human-Computer Inter- action 7, 1–32. URL:https://doi.org/10.1145/3579467, doi:10. 1145/3579467

    work page doi:10.1145/3579467 2023

  52. [52]

    Toward a Theory of Situation Awareness in Dynamic Systems

    Endsley, M.R., 1995. Toward a Theory of Situation Awareness in Dynamic Systems. Human Factors 37, 32–64. doi:10.1518/ 001872095779049543

    work page 1995

  53. [53]

    Towards XAI in the SOC – A User- Centric Study of Explainable Alerts with SHAP and LIME, in: 2022 IEEE International Conference on Big Data (Big Data), IEEE, Osaka, Japan

    Eriksson, H.S., Grøv, G., 2022. Towards XAI in the SOC – A User- Centric Study of Explainable Alerts with SHAP and LIME, in: 2022 IEEE International Conference on Big Data (Big Data), IEEE, Osaka, Japan. pp. 2595–2600. doi:10.1109/BigData55660.2022.10020248

    work page doi:10.1109/bigdata55660.2022.10020248 2022

  54. [54]

    The Effect of Explana- tions on Trust in an Assistance System for Public Transport Users and the Role of the Propensity to Trust, in: Proceedings of Mensch und Computer 2021

    Faulhaber, A.K., Ni, I., Schmidt, L., 2021. The Effect of Explana- tions on Trust in an Assistance System for Public Transport Users and the Role of the Propensity to Trust, in: Proceedings of Mensch und Computer 2021. Association for Computing Machinery, New York, NY, USA. MuC ’21, pp. 303–310. doi:10.1145/3473856.3473886

    work page doi:10.1145/3473856.3473886 2021

  55. [55]

    An Explainable Artificial In- telligence Software Tool for Weight Management Experts (PRIMO): Mixed Methods Study

    Fernandes, G.J., Choi, A., Schauer, J.M., Pfammatter, A.F., Spring, B.J., Darwiche, A., Alshurafa, N.I., 2023. An Explainable Artificial In- telligence Software Tool for Weight Management Experts (PRIMO): Mixed Methods Study. Journal of Medical Internet Research 25, e42047. doi:10.2196/42047

    work page doi:10.2196/42047 2023

  56. [56]

    Förster, M., Hühn, P., Klier, M., Kluge, K., 2021. Capturing Users’ Re- ality: A Novel Approach to Generate Coherent Counterfactual Expla- nations, in: Proceedings of the 54th Hawaii International Conference on System Sciences (HICSS 2021), University of Hawai’i at M¯ anoa, Wailea, HI, USA. pp. 1274–1283. doi:10.24251/hicss.2021.155. 44

    work page doi:10.24251/hicss.2021.155 2021

  57. [57]

    GPT2SP: A Transformer-Based Agile Story Point Estimation Approach

    Fu, M., Tantithamthavorn, C., 2023. GPT2SP: A Transformer-Based Agile Story Point Estimation Approach. IEEE Transactions on Soft- ware Engineering 49, 611–625. doi:10.1109/TSE.2022.3158252

    work page doi:10.1109/tse.2022.3158252 2023

  58. [58]

    Structure-mapping: A theoretical frame- work for analogy

    Gentner, D., 1983. Structure-mapping: A theoretical frame- work for analogy. Cognitive Science 7, 155–170. doi:10.1207/ s15516709cog0702\_3

    work page 1983

  59. [59]

    Vera Liao, Yunfeng Zhang, Rachel Bellamy, and Klaus Mueller

    Ghai, B., Liao, Q.V., Zhang, Y., Bellamy, R., Mueller, K., 2021. Explainable Active Learning (XAL): Toward AI Explanations as In- terfaces for Machine Teachers. Proceedings of the ACM on Human- Computer Interaction 4, 235:1–235:28. doi:10.1145/3432934

    work page doi:10.1145/3432934 2021

  60. [60]

    Local Rule-Based Explanations of Black Box Decision Systems

    Guidotti, R., Monreale, A., Ruggieri, S., Pedreschi, D., Turini, F., Giannotti, F., 2018a. Local Rule-Based Explanations of Black Box Decision Systems. arXiv abs/1805.10820. doi:10.48550/arXiv.1805. 10820

    work page internal anchor Pith review Pith/arXiv arXiv doi:10.48550/arxiv.1805

  61. [61]

    ACM Comput

    Guidotti, R., Monreale, A., Ruggieri, S., Turini, F., Giannotti, F., Pedreschi, D., 2018b. A Survey of Methods for Explaining Black Box Models. ACM Computing Surveys 51, 1–42. doi:10.1145/3236009

    work page doi:10.1145/3236009

  62. [62]

    Guo, L., Daly, E.M., Alkan, O., Mattetti, M., Cornec, O., Knij- nenburg, B., 2022. Building Trust in Interactive Machine Learn- ing via User Contributed Interpretable Rules, in: Proceedings of the 27th International Conference on Intelligent User Interfaces, Associ- ation for Computing Machinery, New York, NY, USA. pp. 537–548. doi:10.1145/3490099.3511111

    work page doi:10.1145/3490099.3511111 2022

  63. [63]

    Causes and Explanations: A Structural- Model Approach

    Halpern, J.Y., Pearl, J., 2005. Causes and Explanations: A Structural- Model Approach. Part II: Explanations. British Journal for the Phi- losophy of Science 56, 889–911. doi:10.1093/bjps/axi148

    work page doi:10.1093/bjps/axi148 2005

  64. [64]

    Development of NASA-TLX (task load index): Results of empirical and theoretical research, in: Han- cock, P.A., Meshkati, N

    Hart, S.G., Staveland, L.E., 1988. Development of NASA-TLX (task load index): Results of empirical and theoretical research, in: Han- cock, P.A., Meshkati, N. (Eds.), Human Mental Workload. North- Holland, Elsevier, Amsterdam. volume 52 ofAdvances in Psychol- ogy, pp. 139–183. URL:https://doi.org/10.1016/S0166-4115(08) 62386-9, doi:10.1016/S0166-4115(08)6...

    work page doi:10.1016/s0166-4115(08 1988

  65. [65]

    Explaining Recommen- dations through Conversations: Dialog Model and the Effects of Inter- faceTypeandDegreeofInteractivity

    Hernández-Bocanegra, D.C., Ziegler, J., 2023. Explaining Recommen- dations through Conversations: Dialog Model and the Effects of Inter- faceTypeandDegreeofInteractivity. ACMTransactionsonInteractive Intelligent Systems 13, 1–47. doi:10.1145/3579541

    work page doi:10.1145/3579541 2023

  66. [66]

    Metrics for Explainable AI: Challenges and Prospects

    Hoffman, R.R., Mueller, S.T., Klein, G., Litman, J., 2018. Met- rics for Explainable AI: Challenges and Prospects. Technical Report arXiv:1812.04608. arXiv

    work page internal anchor Pith review Pith/arXiv arXiv 2018

  67. [67]

    Hüllermeier, W

    Hüllermeier, E., Waegeman, W., 2021. Aleatoric and epistemic uncer- tainty in machine learning: an introduction to concepts and methods. Machine Learning 110, 457–506. doi:10.1007/s10994-021-05946-3

    work page doi:10.1007/s10994-021-05946-3 2021

  68. [68]

    Ibrahim, L., Ghassemi, M.M., Alhanai, T., 2023. Do Explanations Im- prove the Quality of AI-assisted Human Decisions? An Algorithm-in- the-Loop Analysis of Factual & Counterfactual Explanations, in: Pro- ceedings of the 2023 International Conference on Autonomous Agents and Multiagent Systems, International Foundation for Autonomous Agents and Multiagent ...

    work page 2023

  69. [69]

    Emerging Trends in AI-Based Stock Mar- ket Prediction: A Comprehensive and Systematic Review

    Jain, R., Vanzara, R., 2023. Emerging Trends in AI-Based Stock Mar- ket Prediction: A Comprehensive and Systematic Review. Engineering Proceedings 56, 254. doi:10.3390/ASEC2023-15965

    work page doi:10.3390/asec2023-15965 2023

  70. [70]

    Toward Interpretable Machine Learning: Constructing Polynomial Models Based on Feature Interac- tion Trees, in: Kashima, H., Ide, T., Peng, W

    Jang, J., Kim, M., Bui, T., Li, W., 2023. Toward Interpretable Machine Learning: Constructing Polynomial Models Based on Feature Interac- tion Trees, in: Kashima, H., Ide, T., Peng, W. (Eds.), Advances in Knowledge Discovery and Data Mining, Springer Nature Switzerland, Cham. pp. 159–170. doi:10.1007/978-3-031-33377-4\_13

    work page doi:10.1007/978-3-031-33377-4 2023

  71. [71]

    Who needs explanation and when? Juggling explainable AI and user epistemic uncertainty

    Jiang, J., Kahai, S., Yang, M., 2022. Who needs explanation and when? Juggling explainable AI and user epistemic uncertainty. In- ternational Journal of Human-Computer Studies 165, 102839. doi:10. 1016/j.ijhcs.2022.102839

    work page arXiv 2022

  72. [72]

    EUCA: The End-User-Centered Explainable AI Framework

    Jin, W., Fan, J., Gromala, D., Pasquier, P., Hamarneh, G., 2021. EUCA: The End-User-Centered Explainable AI Framework. Techni- cal Report arXiv:2102.02437. arXiv

    work page arXiv 2021

  73. [73]

    Explaining the Unexplainable: Role 46 of XAI for Flight Take-Off Time Delay Prediction, in: Maglogiannis, I., Iliadis, L., MacIntyre, J., Dominguez, M

    Jmoona, W., Ahmed, M.U., Islam, M.R., Barua, S., Begum, S., Fer- reira, A., Cavagnetto, N., 2023. Explaining the Unexplainable: Role 46 of XAI for Flight Take-Off Time Delay Prediction, in: Maglogiannis, I., Iliadis, L., MacIntyre, J., Dominguez, M. (Eds.), Artificial Intel- ligence Applications and Innovations, Springer Nature Switzerland, Cham, Switzerl...

    work page 2023

  74. [74]

    Examining Correlation Between Trust and Trans- parency with Explainable Artificial Intelligence, in: Arai, K

    Kartikeya, A., 2022. Examining Correlation Between Trust and Trans- parency with Explainable Artificial Intelligence, in: Arai, K. (Ed.), Intelligent Computing, Springer International Publishing, Cham, Switzerland. pp. 353–358. doi:10.1007/978-3-031-10464-0\_23

    work page doi:10.1007/978-3-031-10464-0 2022

  75. [75]

    HealthXAI: Collaborative and Explainable AI for Supporting Early Diagnosis of Cognitive Decline

    Khodabandehloo, E., Riboni, D., Alimohammadi, A., 2021. HealthXAI: Collaborative and Explainable AI for Supporting Early Diagnosis of Cognitive Decline. Future Generation Computer Systems 116, 168–189. doi:10.1016/j.future.2020.10.030

    work page doi:10.1016/j.future.2020.10.030 2021

  76. [76]

    Khosravi, H., Buckingham Shum, S., Chen, G., Conati, C., Tsai, Y.S., Kay, J., Knight, S., Martinez-Maldonado, R., Sadiq, S., Gašević, D.,

    work page

  77. [77]

    Computers and Education: Artificial Intelligence 3, 100074

    Explainable Artificial Intelligence in Education. Computers and Education: Artificial Intelligence 3, 100074. doi:10.1016/j.caeai. 2022.100074

    work page doi:10.1016/j.caeai 2022

  78. [78]

    HowShouldtheResultsofArtificialIntelligenceBeExplainedtoUsers? Research on Consumer Preferences in User-Centered Explainable Ar- tificial Intelligence

    Kim, D., Song, Y., Kim, S., Lee, S., Wu, Y., Shin, J., Lee, D., 2023. HowShouldtheResultsofArtificialIntelligenceBeExplainedtoUsers? Research on Consumer Preferences in User-Centered Explainable Ar- tificial Intelligence. Technological Forecasting and Social Change 188, 122343. doi:10.1016/j.techfore.2023.122343

    work page doi:10.1016/j.techfore.2023.122343 2023

  79. [79]

    Human-Centered Evaluation of Explainable AI Applications: A Systematic Review

    Kim, J., Maathuis, H., Sent, D., 2024. Human-Centered Evaluation of Explainable AI Applications: A Systematic Review. Frontiers in Artificial Intelligence 7, 1456486. doi:10.3389/frai.2024.1456486

    work page doi:10.3389/frai.2024.1456486 2024

  80. [80]

    Reliability in Content Analysis: Some Com- mon Misconceptions and Recommendations

    Krippendorff, K., 2004. Reliability in Content Analysis: Some Com- mon Misconceptions and Recommendations. Human Communication Research 30, 411–433. doi:10.1111/j.1468-2958.2004.tb00738.x

    work page doi:10.1111/j.1468-2958.2004.tb00738.x 2004

Showing first 80 references.