PolyInterview: An LLM-based Platform for Immersive Mock Interview Practice with Comprehensive Multimodal Assessment

Chen Chen; Jianing Yin; Jiannong Cao; Kelly Chan; Xiaoyun Liu; Zhiyuan Wen; Zhuo Li; Zijian Wang

REVIEW 2 major objections 4 minor 1 cited by

PolyInterview: An LLM-based Platform for Immersive Mock Interview Practice with Comprehensive Multimodal Assessment

T0 review · 2 major / 4 minor · reviewed 2026-08-02 · deepseek-v4-flash

Challenge this review Re-run · record.json Download PDF Read on arXiv ↗ Review JSON

Pith's one-line read PolyInterview combines tailored question generation, answer-aware follow-ups, and evidence-linked multimodal scoring into one deployed platform.

desk verdict A well-engineered, honestly reported systems paper whose distinctive multimodal-assessment claim lacks the validity evidence it would need; referee it, but expect to ask for a calibration study. read the letter →

arxiv 2607.10310 v3 pith:4J5VYNBQ submitted 2026-07-11 cs.CL

Zhiyuan Wen , Jiannong Cao , Kelly Chan , Zijian Wang , Chen Chen , Xiaoyun Liu , Jianing Yin , Zhuo Li This is my paper · ORCID

classification cs.CL

keywords mockinterviewpracticemultimodalassessmentlargelanguagemodelsdigitalhumaninterviewerKSAframeworkSTARstructureadaptivefollow-upquestionsevidence-linkedfeedback

verification ladder T0 review T1 audit T2 compute T3 formal

The pith

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

The reading

PolyInterview aims to show that realistic mock interview practice can be delivered end-to-end by a single platform: it reads the job description and CV to plan a multi-stage question set, conducts a spoken interview through a lip-synced digital human that asks answer-aware follow-ups, and then scores content, voice, and non-verbal cues through four parallel evaluators. The scores are organized into a three-level rubric—13 behavior features, 10 aspects, two competency tracks—so every recommendation can be traced back to specific evidence in the candidate's answers. The paper reports that in the deployed system, question sets align with the matched job description rather than cross-role descriptions in 93.7% of sessions, and that ten experts rated question plans at 4.62/5 and feedback actionability at 4.70/5. The point is not a new algorithm but a working integration that makes adaptive, multimodal, evidence-linked practice accessible.

What carries the argument

The load-bearing mechanism is the three-layer assessment architecture: four parallel evaluators output 13 behavior-level features (e.g., conceptual accuracy, logic, clarity, eye contact, facial expression, posture, gesture, pronunciation, prosody, fluency), an aggregation agent maps them to 10 assessment aspects using a 70/30 primary/secondary weighting, and the aspects collapse into two competency tracks (Professional and Communication). This hierarchy is what makes the feedback traceable: every score can be followed down to the feature and modality that produced it. The question planner and answer-aware follow-up mechanism use the same KSA aspect vocabulary, so the questions and the assess

What would settle it

A blinded comparison of PolyInterview's automated scores against independent human expert ratings on the same interview recordings: if any of the three modality channels (content, voice, non-verbal) shows near-zero or negative correlation with expert ratings on the corresponding construct—for example, eye contact, pronunciation, or answer completeness—then the comprehensive multimodal assessment claim is undermined.

Watch

Extended reading notes

Core claim

The central claim is that the separate capabilities of personalized question generation, adaptive spoken dialogue, and comprehensive multimodal assessment can be fused into one coherent workflow without sacrificing traceability. PolyInterview does this through a four-stage pipeline: setup, immersive interview, multimodal assessment, and report. Each response is evaluated in parallel by four evaluators that produce 13 behavior-level feature scores; an aggregation agent maps these into 10 KSA-aligned aspects and then into Professional Competency and Communication Competency tracks. Because the mapping is explicit and each score is tied to behavioral evidence, the report can say why a score is

Load-bearing premise

The weakest load-bearing assumption is that the four automated evaluators' 13 behavior-level scores—LLM content scoring, VLM non-verbal analysis, and speech prosody/pronunciation—are accurate measures of the constructs they claim to score; the platform's distinctive multimodal assessment rests entirely on this, while the paper defers score-level validation to future work.

Editorial extensions

If this is right

If a candidate practices on PolyInterview, every follow-up question is bounded by the chosen persona and can request resolution of a contradiction, clarification of an underspecified response, or elaboration on a point of interest; session logs confirm such adaptive probing occurs in the deployed system.
The assessment report links each of the 13 behavior features to 10 aspects and two competency tracks, so a candidate can see the specific utterance, vocal cue, or gesture behind each score rather than a bare number.
Question sets conditioned on a specific job description differ systematically from cross-role sets, with matched job description alignment in 93.7% of sessions and first-rank matching in 82.4%, indicating that practice content is role-specific rather than generic.
Expert evaluation indicates strong question plan quality (4.62/5) and actionable recommendations (4.70/5), supporting the claim that the system outputs are usable by candidates for directed improvement.
The pipeline runs in real time behind HTTPS with concurrent evaluators and a session pool, suggesting this kind of practice can be delivered at scale, not just as a research prototype.

Reading between the lines

Editorial extensions of the paper, not claims the author makes directly.

If the rubric hierarchy is as robust as reported, the same feature-to-aspect-to-track mapping could be adapted to other high-stakes communication training domains, such as teaching, sales, or clinical consultation, by replacing the KSA vocabulary with the relevant competency taxonomy.
The paper's own limitations note that artifact-level expert ratings do not establish score-level criterion validity; a natural next test is to compare PolyInterview's automated scores with independent career-service ratings on the same recordings, in a blinded design.
The 93.7% lexical alignment result suggests a cheap, automatic check for role-conditioning that could serve as a live quality gate for question planning, independent of human review.
Because the report is traceable to specific evidence, a candidate could use the linked evidence to rehearse targeted behaviors; whether such targeted practice actually improves real interview outcomes remains an open, testable question that the paper does not yet answer.

Share X Bluesky LinkedIn Reddit HN

Signed reviews

No signed human review yet.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, and a circularity audit.

Desk Editor's Note

read the letter

You should know two things before reading. First, PolyInterview is real: there is a deployed platform, 1,564 sessions, 1,744 audio/video pairs, and the pipeline runs end-to-end with JD/CV-conditioned question planning, answer-aware follow-ups through a digital human, and a 13-feature→10-aspect→2-track rubric with per-score evidence traces. That integration plus deployment is the genuinely new part, and the work is honestly reported — including the internal/test accounts in the snapshot and the weak spots that surfaced in their own expert evaluation.

Second, the main soft spot is exactly what the stress-test note says: the multimodal assessment scores themselves are unvalidated. The paper’s distinctive contribution is the “comprehensive” feedback, but the LLM content scores, VLM non-verbal features, and speech prosody/pronunciation scores are never checked against human raters or outcomes. The Limitations explicitly defer this (“Artifact-level expert ratings also do not establish score-level criterion validity”), so the authors are transparent. Transparent does not mean harmless: the aggregation from features to aspects to tracks is deterministic, so if the evaluators are fluent but wrong, the polished feedback propagates the error downstream. This is the load-bearing weakness and it is not fixed in this version.

The other concerns are smaller. The 93.7% role-alignment figure is a lexical test that is partly by construction — the question set comes from the matched JD — and the cross-role contrast mitigates but does not remove the circularity. No error bars, unspecified similarity metric. The expert study is believable but small: ten experts, twelve artifacts, with follow-up diagnostic depth (3.13) and polished-response faithfulness (2.75) clearly below the other scores. The authors disclose all of this, which is to their credit.

The central argument — that this platform works as an integrated system and generates role-tailored questions and actionable reports — holds up. What does not hold up is the stronger implied claim that those automated scores measure the constructs they name. For an ed-tech tool, that is not a minor issue; it is the educational value of the product. But it is addressable.

Who should read this: people designing interview practice systems, work on multimodal conversation feedback, or studying LLM-as-judge reliability. It deserves a serious referee. I would send it out, with the clear expectation of major revision: add at minimum a human-rater calibration sample on the 13 features, or narrow the assessment claims until that evidence exists.

Referee Report

2 major / 4 minor

Summary. PolyInterview is a deployed LLM-based mock-interview platform. It takes a JD and CV, generates a five-stage question plan tailored to the role and candidate, runs a spoken interview with a lip-synced digital human and answer-aware follow-ups, and produces a three-layer assessment report: four evaluators compute 13 behavior-level features; these are aggregated into 10 KSA/Big-Five-aligned aspects and two competency tracks, with STAR-based feedback. The paper reports an all-account snapshot (101 accounts, 1,564 sessions, 7,665 questions), a lexical alignment analysis in which 93.7% of sessions' question sets are closer to their matched JD than to cross-role JDs, and a ten-expert study rating question plans 4.62/5, follow-ups 3.68/5, and feedback reports 3.74/5. The authors are explicit that the deployment metrics include internal/test activity and that the expert study rates artifacts rather than establishing criterion validity of the automated feature scores.

Significance. If the system's claims hold, this is a useful end-to-end contribution: it combines personalization, adaptive spoken interaction, and evidence-linked multimodal feedback in one publicly accessible platform, with a substantial deployment snapshot. The rubric hierarchy from features to aspects to tracks is a concrete design that others can adopt, and the authors deserve credit for reporting real deployment data, including limitations and negative expert findings (answer dependence 3.28; response faithfulness 2.75), instead of only favorable results. The main unresolved issue is the validity of the automated feature scores. The reported human study validates artifacts (question plans, follow-ups, feedback reports) at the text level; it does not establish that the 13 behavior-level scores from LLM/VLM/speech evaluators match human perception of the intended constructs. Because 'comprehensive multimodal assessment' is one of the paper's three headline contributions, this gap is not merely a future-work item. The transparency about the limitation is commendable, but a calibration study would materially change the strength of the central claim.

major comments (2)

[§2.3 and Limitations] The central contribution (contribution 2, Abstract) is 'comprehensive multimodal assessment.' This rests entirely on the accuracy of the 13 behavior-level feature scores produced by the four evaluators. The paper provides no human-rater validation, inter-rater calibration, or outcome criterion for any of these features. The Limitations explicitly state: 'Artifact-level expert ratings also do not establish score-level criterion validity, which requires blinded comparison against independent career-service ratings.' I agree; the expert study (§5) rated 12 text artifacts, not the underlying feature scores. Since the remainder of the pipeline is a deterministic aggregation (70/30 weights, gating in Table 4), any systematic error in the features propagates to aspects, tracks, and recommendations. A blinded comparison of automated feature scores against expert ratings on a sample of recorded r
[Abstract, §2.3, Table 4] The paper states that the evaluators produce 13 behavior-level features, and the Oral Expression evaluator analyzes 'pronunciation accuracy, prosody, and fluency' (§2.3). Table 4, however, lists only 12 unique features (ConcAcc, Term, Logic, Clarity, Coher, Word, Eye, Face, Posture, Gest, Pron, Prosody); fluency appears in no aspect mapping. Either fluency is a separate feature that should appear in the mapping, or the count should be corrected. This inconsistency undermines the traceability claim of §2.3/Appendix E and should be resolved.

minor comments (4)

[§4, Figure 5] 'Returning accounts' is presented as 61.4% but never defined. Clarify whether it means users with more than one session, and how the count is derived.
[Table 3] The column heading '1agree' is cryptic. Spell out the agreement criterion, e.g., 'ratings within ±1 point' or similar.
[§4] The snapshot reports 1,744 WAV and 1,744 WebM recordings but 1,425 scored responses. Explain the relationship (e.g., failures in ASR, video conversion, or scoring).
[§4, Abstract] The 93.7% matched-JD alignment is presented as evidence of role-conditioned generation, but because the matched JD is one of the generation inputs, high lexical alignment is expected by construction. The cross-role baseline and rank-first measure provide some control, and the authors do say the test 'does not replace expert judgment.' Still, the abstract's wording gives the result an evidentiary weight it does not have. Recommend framing it explicitly as a self-consistency sanity check, with the expert-rated role relevance (4.80/5 in Table 3) as the evidence that actually addresses quality.

Circularity Check

1 steps flagged · score 2.0 of 10

No significant circularity; one self-referential alignment metric is transparently limited and independently supplemented by expert ratings.

self definitional [Abstract & §4 (Role-conditioned question generation); Appendix C (Question planning)]
"Generated questions are more closely aligned with their matched job description than with cross-role job descriptions in 93.7% of sessions. ... The planner receives the target company and position, JD, CV, interviewer persona, and session duration ... and grounds each question in relevant JD or CV evidence."

The matched JD is itself an input to the question-planning step, so lexical proximity between generated questions and that same JD is partly imposed by construction rather than independently discovered. The cross-role baseline shows the generator is not generic, but it does not by itself establish role-relevance quality. The paper itself de-weights this evidence: 'This lexical test does not replace expert judgment.' This is a mild self-referential consistency check, not a load-bearing derivation.

full rationale

The paper is a systems-and-deployment report, not a derivation from first principles, so most circularity patterns do not apply. The 13 behavior-level features are produced by LLM/VLM/speech evaluators and aggregated by designer-specified 70/30 weights and gating rules; no claim is made that these scores are derived from or fitted to the human ratings, and the authors explicitly disclaim criterion validity ('Artifact-level expert ratings also do not establish score-level criterion validity...'). That is a validity gap, not circularity. The KSA/STAR/Big Five grounding cites external literature (Peterson et al. 2001; Janz 1982; McCrae & Costa 1987), and there are no load-bearing self-citations. The expert study uses ten external raters on 12 artifacts and provides independent evidence for question-plan and feedback quality. The only partially self-referential piece is the lexical role-alignment statistic, which compares generated questions to a JD that was one of their inputs; the paper itself labels this a lexical test that 'does not replace expert judgment,' and the cross-role contrast plus the expert ratings keep the central claim independent. Score 2 reflects this one minor, non-load-bearing self-referential analysis.

Assumptions & free parameters 3 free parameters · 5 assumptions · 0 invented entities

This is a systems paper: the central claims rest on design choices (KSA/STAR grounding, 70/30 weights, gating rules) and domain assumptions (lexical similarity ≈ alignment; LLM/VLM scores ≈ valid measures) rather than on derivations. The largest uncharged assumption is that the automated multimodal scores are trustworthy, which the paper explicitly defers to future criterion-validity work.

free parameters (3)

Assessment aspect weighting (70/30 primary/secondary) = 0.7 / 0.3
Hand-chosen constant in the aggregation from 13 behavior features to 10 assessment aspects (§2.3, Table 4); no data were used to tune it, but it directly determines every aspect score.
Follow-up caps per interviewer persona = Basic=1, Intermediate=2, Advanced=3
Design parameter bounding follow-up rounds (Appendix C); affects how much adaptive probing occurs.
Question-category gating rules for aspects = hand-authored mapping
Which aspects a question category can activate is hand-defined (Appendix E); gating determines which aspect scores are produced for each question.

assumptions (5)

domain assumption KSA is an appropriate competency framework for structuring interview assessment
The 10-aspect rubric and Professional Competency track are organized by KSA (Peterson et al., 2001); the validity of this choice for assessing mock interviews is assumed (§2.3, Appendix D).
domain assumption STAR is an appropriate structure for evaluating behavioral-response evidence
Behavioral responses are scored for Situation-Task-Action-Result coverage and recommendations convert missing STAR elements into advice (Janz, 1982; §2.3, Appendix D).
domain assumption Lexical similarity between question sets and JDs is a meaningful proxy for role alignment
The 93.7%/82.4% claims measure lexical correspondence only; the paper concedes 'this lexical test does not replace expert judgment' (§4).
domain assumption LLM/VLM/speech evaluator scores reflect the real behavioral constructs being scored
The 13 feature scores feed all aspect and track scores, but no human-rated ground-truth validation or inter-rater calibration is reported; criterion validity is explicitly deferred (Limitations).
domain assumption The five-stage interview structure (self-intro, behavioral, skill-QA, scenario, candidate questions) is a complete/appropriate interview model
90.9% coverage of 'five-stage question sets' is treated as workflow completeness (§4, Table 2); the completeness of the stage taxonomy itself is assumed.

how reviews work

0 comments

Cite this review

Pith. "Pith review of PolyInterview: An LLM-based Platform for Immersive Mock Interview Practice with Comprehensive Multimodal Assessment." pith.science (2026). https://pith.science/paper/4J5VYNBQ

@misc{pith2026260710310,
  author       = {Pith},
  title        = {Pith review of: PolyInterview: An LLM-based Platform for Immersive Mock Interview Practice with Comprehensive Multimodal Assessment},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/4J5VYNBQ}},
  note         = {Machine review of arXiv:2607.10310}
}

read the original abstract

Preparing for job interviews is important for securing desired positions, yet realistic practice remains difficult to access: real interviews are infrequent, expert mock coaching is costly, and self-practice offers neither adaptive dialogue nor structured assessment. Existing systems typically address only parts of this need through fixed question sequences, limited communication channels, or feedback with little supporting evidence. We present PolyInterview, an LLM-based platform for immersive mock interview practice with comprehensive multimodal assessment. PolyInterview uses the target job description and CV to generate questions tailored to the role and candidate, conducts multi-turn spoken interviews with a lip-synced digital human interviewer that asks answer-aware follow-up questions, and evaluates response content, vocal delivery, and non-verbal behavior. Four parallel evaluators produce 13 behavior-level features that are aggregated into 10 assessment aspects and two competency tracks. Guided by the KSA and STAR frameworks, the report links each score to behavioral evidence and actionable recommendations. PolyInterview is publicly accessible. Its current all-account snapshot contains 101 accounts, 1,564 interview sessions, 7,665 generated questions, and 1,422 five-stage question sets. Generated questions are more closely aligned with their matched job description than with cross-role job descriptions in 93.7% of sessions. An evaluation by ten experts found strong question plans and actionable feedback.

Figures

Figures reproduced from arXiv: 2607.10310 by the authors.

**Figure 2.** PolyInterview’s pipeline for personalized questioning, answer-aware follow-ups, multimodal assessment, [PITH_FULL_IMAGE:figures/full_fig_p004_2.png] view at source ↗

**Figure 3.** PolyInterview’s interface for personalized setup, immersive interviewing, and multilevel assessment [PITH_FULL_IMAGE:figures/full_fig_p005_3.png] view at source ↗

Figures from the paper (4 more)

**Figure 4.** Figure 4: Three-layer multimodal assessment from 13 [PITH_FULL_IMAGE:figures/full_fig_p006_4.png]

**Figure 5.** Figure 5: Workflow coverage, role alignment, and returning-account evidence. Adaptive interaction and multimodal assessment coverage. The logs contain 1,274 answered main questions, including 208 with follow-ups and 342 follow-up turns. The assessment pool contains 1,425 score…

**Figure 6.** Figure 6: Personalized question generation through [PITH_FULL_IMAGE:figures/full_fig_p009_6.png]

**Figure 7.** Figure 7: Answer-aware follow-up generation using diagnostic scoring, bounded follow-up rounds, and explicit exit conditions. can request resolution of a contradiction, clarification of an underspecified response, or elaboration on a point of interest. The interview proceeds o…

Review history (2 revisions) →

Discussion (0). Continue with ORCID to comment.

Forward citations

Cited by 1 Pith paper

Reviewed papers in the Pith corpus that reference this work. Sorted by Pith novelty score. Full citation record

PolyPresentation: A Multimodal AI Platform for Slide-Aware Iterative Presentation Practice
cs.HC 2026-08 conditional novelty 6.0 of 10

PolyPresentation, a slide-aware AI rehearsal platform, scores highest among five systems in a feedback-quality evaluation, though the comparison uses the same AI model as judge and system.

Reference graph

Works this paper leans on

22 extracted references · cited by 1 Pith paper

[1]

Psychological Bulletin , volume =

The validity and utility of selection methods in personnel psychology: Practical and theoretical implications of 85 years of research findings , author =. Psychological Bulletin , volume =. 1998 , publisher =

1998
[2]

Psychological Bulletin , volume =

Thin slices of expressive behavior as predictors of interpersonal consequences: A meta-analysis , author =. Psychological Bulletin , volume =. 1992 , publisher =

1992
[3]

Journal of Personality and Social Psychology , volume =

Validation of the five-factor model of personality across instruments and observers , author =. Journal of Personality and Social Psychology , volume =. 1987 , publisher =

1987
[4]

IEEE Transactions on Affective Computing , volume =

Automated Analysis and Prediction of Job Interview Performance , author =. IEEE Transactions on Affective Computing , volume =. 2018 , publisher =

2018
[5]

Proceedings of the AAAI Conference on Artificial Intelligence , volume =

Hemamou, L. Proceedings of the AAAI Conference on Artificial Intelligence , volume =
[6]

and Zhang, Hao and Gonzalez, Joseph E

Zheng, Lianmin and Chiang, Wei-Lin and Sheng, Ying and Zhuang, Siyuan and Wu, Zhanghao and Zhuang, Yonghao and Lin, Zi and Li, Zhuohan and Li, Dacheng and Xing, Eric P. and Zhang, Hao and Gonzalez, Joseph E. and Stoica, Ion , booktitle =. Judging
[7]

Journal of Applied Psychology , volume =

Initial comparisons of patterned behavior description interviews versus unstructured interviews , author =. Journal of Applied Psychology , volume =. 1982 , publisher =

1982
[8]

and Mumford, Michael D

Peterson, Norman G. and Mumford, Michael D. and Borman, Walter C. and Jeanneret, P. Richard and Fleishman, Edwin A. and Levin, Kerry Y. and Campion, Michael A. and Mayfield, Melinda S. and Morgeson, Frederick P. and Pearlman, Kenneth and Gowing, Marilyn K. and Lancaster, Anita R. and Silver, Marilyn B. and Dye, Donna M. , journal =. Understanding Work Usi...

2001

Show all 22 references

[9]

Proceedings of the 28th ACM International Conference on Multimedia , pages =

A lip sync expert is all you need for speech to lip generation in the wild , author =. Proceedings of the 28th ACM International Conference on Multimedia , pages =
[10]

Journal of Applied Psychology , volume =

Automated video interview personality assessments: Reliability, validity, and generalizability investigations , author =. Journal of Applied Psychology , volume =. 2022 , publisher =

2022
[11]

, title =

Hoque, Mohammed Ehsan and Courgeon, Matthieu and Martin, Jean-Claude and Mutlu, Bilge and Picard, Rosalind W. , title =. Proceedings of the 2013 ACM International Joint Conference on Pervasive and Ubiquitous Computing (UbiComp '13) , year =

2013
[12]

Anderson, Keith and Andr. The. Advances in Computer Entertainment (ACE 2013) , series =. 2013 , publisher =

2013
[13]

A Job Interview Simulation: Social Cue-Based Interaction with a Virtual Character , booktitle =

Baur, Tobias and Damian, Ionut and Gebhard, Patrick and Porayska-Pomsta, Ka. A Job Interview Simulation: Social Cue-Based Interaction with a Virtual Character , booktitle =. 2013 , pages =

2013
[14]

Dear Computer, Teach Me Manners: Testing Virtual Employment Interview Training , journal =

Langer, Markus and K. Dear Computer, Teach Me Manners: Testing Virtual Employment Interview Training , journal =. 2016 , doi =

2016
[15]

and Ginger, Emily J

Smith, Matthew J. and Ginger, Emily J. and Wright, Katherine and Wright, Michael A. and Taylor, Julie Lounds and Humm, Laura Boteler and Olsen, Dale E. and Bell, Morris D. and Fleming, Michael F. , title =. Journal of Autism and Developmental Disorders , volume =. 2014 , doi =

2014
[16]

Proceedings of the 2020 International Conference on Multimodal Interaction (ICMI '20) , publisher =

Inoue, Koji and Hara, Kohei and Lala, Divesh and Yamamoto, Kenta and Nakamura, Shizuka and Takanashi, Katsuya and Kawahara, Tatsuya , title =. Proceedings of the 2020 International Conference on Multimodal Interaction (ICMI '20) , publisher =. 2020 , pages =

2020
[17]

Proceedings of the Sixteenth ACM International Conference on Web Search and Data Mining (WSDM '23) , year =

Li, Mingzhe and Chen, Xiuying and Liao, Weiheng and Song, Yang and Zhang, Tao and Zhao, Dongyan and Yan, Rui , title =. Proceedings of the Sixteenth ACM International Conference on Web Search and Data Mining (WSDM '23) , year =
[18]

, title =

Wang, Zihao and Keyes, Nathan and Crawford, Terry and Choi, Jinho D. , title =. Information , volume =. 2023 , doi =

2023
[19]

Proceedings of the 31st ACM SIGKDD Conference on Knowledge Discovery and Data Mining V.2 (KDD '25) , year =

Sun, Hongda and Lin, Hongzhan and Yan, Haiyu and Song, Yang and Gao, Xin and Yan, Rui , title =. Proceedings of the 31st ACM SIGKDD Conference on Knowledge Discovery and Data Mining V.2 (KDD '25) , year =
[20]

2025 , note =

Nguyen, Truong Thanh Hung and Nguyen, Tran Diem Quynh and Cao, Hoang Loc and Tran, Thi Cam Thanh and Truong, Thi Cam Mai and Cao, Hung , title =. 2025 , note =

2025
[21]

and Stil, Sophia and Knutsen, Kirk McInnis and Rho, Eugenia H

Daryanto, Taufiq and Ding, Xiaohan and Wilhelm, Lance T. and Stil, Sophia and Knutsen, Kirk McInnis and Rho, Eugenia H. , title =. Proceedings of the ACM on Human-Computer Interaction , volume =. 2025 , doi =

2025
[22]

Bai, Jinze and Bai, Shuai and Yang, Shusheng and Wang, Shijie and Tan, Sinan and Wang, Peng and Lin, Junyang and Zhou, Chang and Zhou, Jingren , journal =

Pith tools

Reviewed August 2, 2026 · model on record in the stance chip above.

[1] [1]

Psychological Bulletin , volume =

The validity and utility of selection methods in personnel psychology: Practical and theoretical implications of 85 years of research findings , author =. Psychological Bulletin , volume =. 1998 , publisher =

1998

[2] [2]

Psychological Bulletin , volume =

Thin slices of expressive behavior as predictors of interpersonal consequences: A meta-analysis , author =. Psychological Bulletin , volume =. 1992 , publisher =

1992

[3] [3]

Journal of Personality and Social Psychology , volume =

Validation of the five-factor model of personality across instruments and observers , author =. Journal of Personality and Social Psychology , volume =. 1987 , publisher =

1987

[4] [4]

IEEE Transactions on Affective Computing , volume =

Automated Analysis and Prediction of Job Interview Performance , author =. IEEE Transactions on Affective Computing , volume =. 2018 , publisher =

2018

[5] [5]

Proceedings of the AAAI Conference on Artificial Intelligence , volume =

Hemamou, L. Proceedings of the AAAI Conference on Artificial Intelligence , volume =

[6] [6]

and Zhang, Hao and Gonzalez, Joseph E

Zheng, Lianmin and Chiang, Wei-Lin and Sheng, Ying and Zhuang, Siyuan and Wu, Zhanghao and Zhuang, Yonghao and Lin, Zi and Li, Zhuohan and Li, Dacheng and Xing, Eric P. and Zhang, Hao and Gonzalez, Joseph E. and Stoica, Ion , booktitle =. Judging

[7] [7]

Journal of Applied Psychology , volume =

Initial comparisons of patterned behavior description interviews versus unstructured interviews , author =. Journal of Applied Psychology , volume =. 1982 , publisher =

1982

[8] [8]

and Mumford, Michael D

Peterson, Norman G. and Mumford, Michael D. and Borman, Walter C. and Jeanneret, P. Richard and Fleishman, Edwin A. and Levin, Kerry Y. and Campion, Michael A. and Mayfield, Melinda S. and Morgeson, Frederick P. and Pearlman, Kenneth and Gowing, Marilyn K. and Lancaster, Anita R. and Silver, Marilyn B. and Dye, Donna M. , journal =. Understanding Work Usi...

2001