Pith. sign in

REVIEW 4 major objections 6 minor 2 cited by

The Value of AI-Generated Metadata for UGC Platforms: Evidence from a Large-scale Field Experiment

T0 review · 4 major / 6 minor · reviewed 2026-08-11 · deepseek-v4-flash

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

Pith's one-line read This paper claims that AI-drafted video titles add value by fixing missing metadata: access raised valid watches 1.6% and watch time 0.9%, adoption raised them 7.1% and 4.1%, while already-titled videos lost viewership unless creators…

desk verdict Solid ITT from a huge field experiment, but the headline LATE is not cleanly identified because the paper's own inspiration effect violates the exclusion restriction. read the letter →

arxiv 2412.18337 v1 pith:WAXTGROE submitted 2024-12-24 econ.GN cs.AIcs.HCq-fin.EC

Xinyi Zhang , Chenshuo Sun , Renyu Zhang , Khim-Yong Goh This is my paper · ORCID
classification econ.GNcs.AIcs.HCq-fin.EC
keywords generativeAIvideometadatauser-generatedcontentshort-videoplatformfieldexperimentrecommendersystemshuman-AIco-creationsparsity
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

Generative AI is usually studied as content viewers see, such as ad copy or product descriptions. This paper asks whether AI can create value in metadata that users almost never notice: video titles that feed a recommendation engine. In a randomized field experiment with about two million short-video producers, giving uploaders access to an AI-drafted title raised valid watches (full or sufficiently long plays) by 1.6% and watch time by 0.9%; among producers who adopted the AI title, the increases were 7.1% and 4.1%. The paper attributes most of the gain to filling in missing metadata rather than to any change viewers see. It then argues that AI titles are not generally better than human titles, and that the real upside lies in human-AI co-creation, where producers revise the draft into a more specific, lexically rich title.

What carries the argument

The load-bearing object is the video title as structured metadata in a two-stage recommender system: candidate generation and ranking use title text, together with user profiles and engagement, to decide which videos to surface. The paper's empirical machinery is the randomized offer of an AI-generated title in the posting interface, used both as an intention-to-treat contrast and as an instrument for actual adoption. For the already-titled comparison, the paper relies on propensity-score matching (pairing treated and control videos on observed pre-treatment covariates). Separately, a proprietary log of 93,618,096 recommendation sessions supplies predicted engagement probabilities, letting the paper compare areas under the ROC curve for treated versus control videos as a measure of matching accuracy.

What would settle it

A follow-up experiment could A/B-test identical videos whose titles are randomly assigned to be fully AI-generated, fully human-written, or a lightly revised AI draft, with no producer choice involved; the valid-watch difference between the first two arms would directly test the claim that AI titles underperform human titles on already-titled videos.

Watch

Extended reading notes

Core claim

On a user-generated-content platform, the value of AI-generated metadata is real but conditional. Using random assignment of about two million producers to access to AI-generated titles, the paper estimates that offering the tool increased the share of videos with a title by 41.4% (and tags by 72.4%), that this raised valid watches by 1.6% and watch duration by 0.9% on an intention-to-treat basis, and that adoption of an AI title raised the same outcomes by 7.1% and 4.1%. The effect is stronger among low-skilled producers and hedonic content, exactly the segments with sparse metadata. An analysis of 93.6 million recommendation sessions shows higher recommender AUCs for engagement for treatment videos, supporting the proposed mechanism: the added title text improves user-video matching rather than directly attracting viewers. The paper's second main claim is that, among videos that would have a human title anyway, adopting the AI title reduces valid watches by 37.9% and watch time by 32.6%, while substantial human revision of the AI draft flips the sign, with less similar titles performing better and showing higher lexical density, variation, and entropy.

Load-bearing premise

The paper's causal reading of the AI-versus-human-title results assumes that the 51.56% of treatment-group titled videos that received no AI title because of algorithmic issues are missing at random after matching; if failure correlates with hard-to-describe or lower-quality videos, the estimated decline and co-creation gains would partly reflect selection.

Editorial extensions

If this is right

  • A platform can increase content consumption without changing what viewers see, purely by reducing metadata sparsity on the supply side.
  • The gains are largest for low-skilled producers and hedonic videos, so targeting the tool at those segments would capture most of the benefit.
  • Fully automated AI titles should not be treated as a replacement for human titles; for videos that already have a title, auto-adoption can reduce viewership.
  • Designing the creator flow to encourage revision of AI drafts, rather than one-click adoption, can convert an average negative effect into a positive one.
  • Improved matching accuracy should generalize across recommendation channels and to downstream engagement metrics such as likes, shares, and follows.

Reading between the lines

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

  • If the paper's mechanism is sparsity relief, the observed 1.6% and 7.1% gains are a one-time catch-up effect: as AI titles become widespread, the marginal benefit of the tool should shrink toward the negative already-titled case.
  • The co-creation result suggests a testable design principle for other UGC platforms: the economically relevant output of an AI title tool may be the human revision it inspires, not the title itself, so interventions should measure revision rates rather than adoption rates.
  • The same logic should apply to other sparse metadata fields, such as product descriptions, hashtags, or image captions, but the magnitude will depend on how heavily the platform's recommender weights each field.
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.

Referee Report

4 major / 6 minor

Summary. This paper reports a large-scale randomized field experiment on a large short-video platform in Asia, in which roughly one million content producers were randomly assigned to receive access to AI-generated video titles on the posting page. The authors estimate that access to AI-generated titles increases valid watches by 1.6% and watch duration by 0.9% (intention-to-treat), and that adoption of AI-generated titles increases these outcomes by 7.1% and 4.1% (local average treatment effect estimated by instrumental variables). They attribute the effect to reduced metadata sparsity (a 41.4% increase in the likelihood of having a title) and to improved user-video matching accuracy, as measured by recommender-system AUC. Section 5 further reports that among videos that already had human titles, access to AI-generated titles is associated with a 37.9% decline in valid watches, but that videos whose producers substantially revised the AI title outperform human-titled controls, supporting a human-AI co-creation narrative.

Significance. If the headline results hold, this is a valuable contribution: it provides rare large-scale experimental evidence on the economic value of AI-generated metadata, a type of AIGC that does not directly engage viewers and whose impact operates through the recommender system rather than through user-facing content quality. The strength of the paper is the clean randomization: the ITT estimates in Table 4 are based on treatment assignment and external viewership outcomes, so they are internally valid. The paper also includes multiple robustness checks, alternative adoption thresholds (Table 28), alternative outcome measures (Tables 23-26), and a mechanism analysis using the platform's own recommendation predictions. However, as detailed below, the LATE and the quality/co-creation comparisons rely on stronger assumptions that are not fully supported by the evidence presented.

major comments (4)
  1. [Section 4.2, Eqs. (2)-(3) and Section 5.2] The instrumental-variable strategy for the LATE requires that treatment assignment affects viewership only through exact adoption of the AI-generated title. The paper's own survey evidence in Section 5.2 documents an 'inspiration effect' in which treated producers who do not exactly adopt the AI title still use it as a creative catalyst for writing better titles. Such producers are directly affected by treatment, violating the exclusion restriction. With a first-stage adoption rate of only 23.4% (Table 2 and Section 4.2), non-adopters constitute about 76.6% of the treatment group, so even a modest direct effect on non-adopters is amplified by roughly a factor of 3.3 in the Wald ratio. The ITT results are not affected, but the abstract's claim that 'when producers adopted these titles' viewership increased by 7.1% and 4.1% is not causally identified as stated. The authors should either re-estimate the LATE under weaker assumptions (e.g., bounds that allow for direct effects), use the within-treatment variation from algorithmic title-generation failures (Online Appendix B) to test for direct effects on non-adopters, or explicitly relegate the LATE to a secondary, descriptive role.
  2. [Online Appendix B and Section 5.1 (Tables 12-14)] The matched-sample analysis drops 2,226,922 treated titled videos (51.56% of the titled treatment group) because AI-generated titles were not produced due to 'algorithmic issues.' The PSM procedure can only balance observed covariates; it cannot address selection on unobservables if title-generation failure is correlated with video complexity, producer effort, or underlying video quality. The post-matching balance tests (Table 17) show very high p-values (e.g., 0.96, 0.99), which is unusual and may reflect over-matching or reduced effective sample size; more importantly, the balance statistics do not speak to unobservable drivers of the outcome. The estimated 37.9% decline in Table 12 and the co-creation gains in Table 14 would be biased if the excluded videos differ systematically from the included ones. The authors should provide evidence on the excluded videos' observable characteristics and, if possible, their viewership outcomes (e.g., by linking algorithmic failure to video features), or present bounding exercises that relax the missing-at-random assumption.
  3. [Section 5.1, Table 12] The coefficient on Treat in the matched titled-video sample is an intention-to-treat effect among videos that had titles, not the effect of adopting AI-generated titles. In the treatment group, the titled videos include exact adopters, partial revisers, and producers who wrote their own titles after being inspired by the AI suggestion. The paper's interpretation that 'adopting the AI-generated title decreased its viewership' (abstract and Section 5.1) overstates what Table 12 identifies. To support the quality-comparison claim, the authors would need to isolate adoption within the titled subsample (e.g., by instrumenting exact adoption with treatment and estimating a LATE for titled videos) or at least explicitly frame the result as the average effect of access on already-titled videos.
  4. [Section 4.3, Table 11] The AUC comparison uses the platform's recommender-system predictions, and video titles are an input to that recommender (Section 3.1). Because the treatment increases the likelihood that a title exists, the observed improvement in prediction accuracy may partly reflect the recommender having access to more input features rather than a genuine improvement in matching quality. The authors should clarify whether the AUC is computed on out-of-sample predictions from a fixed model or from a retrained model, and should discuss the extent to which the AUC gain is mechanical. The ITT viewership results stand regardless, but the mechanistic claim of 'improved user-video matching accuracy' is not fully separated from the recommender's own dependence on the title feature.
minor comments (6)
  1. [Section 1] There is a typo in 'larege-scale experimental evidence' in the final paragraph of the introduction; it should read 'large-scale.'
  2. [Section 3.2 and Section 3.3] The experiment is described as running from July 20 to August 21, 2023, but AI-generated titles were only stored between August 8 and August 21 because of technical issues. This discrepancy should be acknowledged and discussed, as it means the pre-treatment period and the start of the treatment period are not contiguous.
  3. [Table 1] The variable Treatij is defined with both producer and video subscripts, but treatment is assigned at the producer level. Please clarify the unit of treatment and use a consistent notation (e.g., Treati).
  4. [Section 5.1, Eq. (4)] The statement that the main effect of Similarity 'would be absorbed by the interaction term (Treati × Similarityij) due to collinearity' is not strictly correct. In ordinary least squares, an interaction term does not absorb a main effect unless the main effect has zero variance in the control group. Please explain the coding of Similarity for control videos (e.g., whether it is set to zero or undefined) and justify the omission of its main effect.
  5. [Section 7 (Conclusion)] The conclusion states that the viewership-boosting effect 'was amplified for utilitarian videos and those produced by low-skilled creators,' but the results in Table 6 show a negative interaction for utilitarian-content videos (Treat × Utilitarian = -0.032 for valid watches), meaning the effect is smaller, not larger, for utilitarian videos. This is internally inconsistent and should be corrected.
  6. [Online Appendix B, Table 17 and Table 18] Some post-matching standardized biases remain relatively high (e.g., Experience with %Bias of 7.7 in Table 17 and 25.6 in Table 18, and LowSkill with 9.7 in Table 18). The authors state that the mean differences are no longer statistically significant, but with over a million observations per group, statistical significance is a weak criterion; reporting whether these biases are within conventional thresholds (e.g., <10%) and discussing the implications for the matched-sample estimates would be more informative.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity found: the randomized ITT and IV-based LATE rest on external outcome data, and the recommender-AUC mechanism is a secondary empirical check rather than a derivation of the main results.

full rationale

The main derivation chain is not circular. Equation (1) compares randomized treatment and control groups on platform-logged viewership outcomes (valid watches and watch duration), so the 1.6% and 0.9% ITT estimates are external outcome comparisons, not fitted inputs renamed as predictions. The LATE in Eqs. (2)-(3) is the Wald ratio of the ITT to the first-stage adoption rate; it is an algebraic rescaling of the randomized contrast, not a parameter fitted to the outcome and then relabeled. The skeptic's concern that the 'inspiration effect' violates the IV exclusion restriction is an identification threat to the causal interpretation of the LATE, but it is not circularity: the estimate is computed from the data rather than assumed. The AUC mechanism analysis uses the platform's own recommender predictions, which consume titles as inputs, so the AUC comparison is partly mechanical in the sense that the intervention changes the recommender's input; however, it is evaluated against actual user behaviors for randomized groups and does not feed back into the main consumption estimates, so it is a secondary mechanism check rather than a load-bearing derivation. The co-creation analysis uses cosine similarity as a covariate and applies propensity-score matching; this is a standard selection-correction design, not a fitted parameter that defines the outcome. Self-citations (e.g., Zeng et al. 2023 for the two-week outcome window and Ye et al. 2023 in the literature review) are methodological precedents or literature placement and are not load-bearing. No uniqueness theorem, ansatz-smuggling citation, or renaming of a known result was found. Overall, the paper's central claims stand on the randomized experiment and external outcome data.

Assumptions & free parameters 4 free parameters · 6 assumptions · 0 invented entities

The central ITT and LATE estimates rely on standard randomization assumptions plus a few domain assumptions about title visibility and exclusion restriction. The co-creation and quality-comparison analyses add an unverifiable missing-at-random assumption after dropping 51.56% of treated titled videos. No free parameters are fitted to produce the headline effect sizes.

free parameters (4)
  • Title adoption threshold (exact match)
    Adoptij is defined as an exact match; robustness checks use 95%, 90%, 85%, and 80% similarity thresholds. This choice affects LATE magnitude but not sign.
  • LowSkill median split at 420 followers
    Median split chosen to define low-skilled producers; robustness checks with 30th to 70th percentiles yield consistent results.
  • Textual similarity threshold of 20% for low-similarity PSM
    Threshold chosen in Section 5.1 to define heavily revised titles; not fit to maximize statistical significance.
  • Two-week cumulative viewership window
    Outcomes accumulated over the first two weeks after posting; a modeling choice that affects the measured effect sizes.
assumptions (6)
  • domain assumption Producers were randomly assigned to treatment without contamination from external GAI tools.
    Section 3.2 states random assignment and that external GAI tools for video titles were unlikely in 2023; this supports the causal interpretation of ITT.
  • domain assumption Viewers do not notice titles, so consumption changes are not driven by direct user engagement with displayed titles.
    Section 3.2 and Figure 1: titles are small and bottom-left; the design deliberately makes metadata invisible to viewers.
  • domain assumption Exclusion restriction: treatment assignment affects viewership only through AI title access or adoption, not through other pathways.
    Section 4.2 IV analysis: first-stage F is huge, but exclusion is assumed; authors test that production behavior does not change in Section 6.
  • ad hoc to paper After PSM, treated titled videos without AI-generated titles (51.56% removed) are missing at random conditional on controls.
    Online Appendix B removes 2,226,922 videos due to algorithmic issues; the validity of the matched comparison depends on this ignorability assumption.
  • domain assumption The November 2023 AUC dataset reflects the same titles as during the experiment.
    Section 4.3: predicted engagement data were not archived during the experiment, so November data are used; titles are stated to be unchanged.
  • domain assumption AUC of the platform recommender is a valid measure of user-video matching accuracy.
    Section 4.3 uses AUC for like, share, and follow as a proxy for matching; viewership outcomes are earlier in the journey and not directly measured.

how reviews work

0 comments
Cite this review

Pith. "Pith review of The Value of AI-Generated Metadata for UGC Platforms: Evidence from a Large-scale Field Experiment." pith.science (2026). https://pith.science/paper/WAXTGROE

@misc{pith2026241218337,
  author       = {Pith},
  title        = {Pith review of: The Value of AI-Generated Metadata for UGC Platforms: Evidence from a Large-scale Field Experiment},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/WAXTGROE}},
  note         = {Machine review of arXiv:2412.18337}
}
read the original abstract

AI-generated content (AIGC), such as advertisement copy, product descriptions, and social media posts, is becoming ubiquitous in business practices. However, the value of AI-generated metadata, such as titles, remains unclear on user-generated content (UGC) platforms. To address this gap, we conducted a large-scale field experiment on a leading short-video platform in Asia to provide about 1 million users access to AI-generated titles for their uploaded videos. Our findings show that the provision of AI-generated titles significantly boosted content consumption, increasing valid watches by 1.6% and watch duration by 0.9%. When producers adopted these titles, these increases jumped to 7.1% and 4.1%, respectively. This viewership-boost effect was largely attributed to the use of this generative AI (GAI) tool increasing the likelihood of videos having a title by 41.4%. The effect was more pronounced for groups more affected by metadata sparsity. Mechanism analysis revealed that AI-generated metadata improved user-video matching accuracy in the platform's recommender system. Interestingly, for a video for which the producer would have posted a title anyway, adopting the AI-generated title decreased its viewership on average, implying that AI-generated titles may be of lower quality than human-generated ones. However, when producers chose to co-create with GAI and significantly revised the AI-generated titles, the videos outperformed their counterparts with either fully AI-generated or human-generated titles, showcasing the benefits of human-AI co-creation. This study highlights the value of AI-generated metadata and human-AI metadata co-creation in enhancing user-content matching and content consumption for UGC platforms.

Discussion (0). Continue with ORCID to comment.

Forward citations

Cited by 2 Pith papers

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

  1. AIGVE-MACS: Unified Multi-Aspect Commenting and Scoring Model for AI-Generated Video Evaluation

    cs.CV 2025-07 conditional novelty 5.0 of 10

    A finetuned vision-language model jointly predicts nine aspect scores and written comments for AI-generated videos, with a new benchmark and claims of state-of-the-art alignment with human judgment.

  2. Red Teaming AI Policy: A Taxonomy of Avoision and the EU AI Act

    cs.CY 2025-06 accept novelty 5.0 of 10

    A taxonomy of avoision under the EU AI Act, with strategies to escape scope, exploit exemptions, and manipulate risk or operator categories.

Reference graph

Works this paper leans on

60 extracted references · 57 canonical work pages · cited by 2 Pith papers

  1. [1]

    Adomavicius, Gediminas, Zan Huang, Alexander Tuzhilin. 2008. Personalization and recommender systems. State-of-the-Art Decision-Making Tools in the Information-Intensive Age\/ . INFORMS, 55--107

    work page 2008

  2. [2]

    Agrawal, Saurabh, John Trenkle, Jaya Kawale. 2023. Beyond labels: Leveraging deep learning and llms for content metadata. Proceedings of the 17th ACM Conference on Recommender Systems\/ . 1--1

    work page 2023

  3. [3]

    collaborating

    Anthony, Callen, Beth A. Bechky, Anne-Laure Fayard. 2023. “collaborating” with ai: Taking a system view to explore the future of work. Organization Science\/ 34 (5) 1672--1694

    work page 2023

  4. [4]

    Bai, Bing, Tat Chan, Dennis Zhang, Fuqiang Zhang, Yujie Chen, Haoyuan Hu. 2022 a . The value of logistic flexibility in e-commerce. Available at SSRN 4206229\/

    work page 2022

  5. [5]

    Bai, Bing, Hengchen Dai, Dennis J Zhang, Fuqiang Zhang, Haoyuan Hu. 2022 b . The impacts of algorithmic work assignment on fairness perceptions and productivity: Evidence from field experiments. Manufacturing & Service Operations Management\/ 24 (6) 3060--3078

    work page 2022

  6. [6]

    Bi, Xuan, Mochen Yang, Gediminas Adomavicius. 2024. Consumer acquisition for recommender systems: A theoretical framework and empirical evaluations. Information Systems Research\/ 35 (1) 339--362

    work page 2024

  7. [7]

    Boyac , Tamer, Caner Canyakmaz, Francis de V\' e ricourt. 2024. Human and machine: The impact of machine input on decision making under cognitive limitations. Management Science\/ 70 (2) 1258--1275

    work page 2024

  8. [8]

    Buell, Ryan W, Tami Kim, Chia-Jung Tsay. 2017. Creating reciprocal value through operational transparency. Management Science\/ 63 (6) 1673--1695

    work page 2017

Show all 60 references
  1. [9]

    Burtch, Gordon, Qinglai He, Yili Hong, Dokyun Lee. 2022. How do peer awards motivate creative content? experimental evidence from reddit. Management Science\/ 68 (5) 3488--3506

    2022

  2. [10]

    Burtch, Gordon, Yili Hong, Ravi Bapna, Vladas Griskevicius. 2018. Stimulating online reviews by combining financial incentives and social norms. Management Science\/ 64 (5) 2065--2082

    2018

  3. [11]

    Burtch, Gordon, Dokyun Lee, Zhichen Chen. 2023. Generative ai for ugc and online community engagement. Available at SSRN 4521754\/

    2023

  4. [12]

    Capraro, Valerio, Austin Lentsch, Daron Acemoglu, Selin Akgun, Aisel Akhmedova, Ennio Bilancini, Jean-Fran c ois Bonnefon, Pablo Bra \ n as-Garza, Luigi Butera, Karen M Douglas, et al. 2024. The impact of generative artificial intelligence on socioeconomic inequalities and pol...

    2024

  5. [13]

    Chen, Jiawei, Luo He, Hongyan Liu, Yinghui Yang, Xuan Bi. 2024. Background music recommendation on short video sharing platforms. Information Systems Research\/

    2024

  6. [14]

    Chen, Zenan, Jason Chan. 2023. Large language model in creative work: The role of collaboration modality and user expertise. Management Science\/

    2023

  7. [15]

    Cheng, Zhaoqi, Dokyun Lee, Prasanna Tambe. 2022. Innovae: Generative ai for mapping patents and firm innovation. Available at SSRN 3868599\/

    2022

  8. [16]

    Choi, Tsan-Ming, Stein W Wallace, Yulan Wang. 2018. Big data analytics in operations management. Production and Operations Management\/ 27 (10) 1868--1883

    2018

  9. [17]

    Clyde, Nicholas, Dennis Zhang, Bing Bai. 2024. The impact of ridesharing platforms on healthcare access. Available at SSRN 4968892\/

    2024

  10. [18]

    Cole, Rebel A, Tatyana Sokolyk. 2018. Debt financing, survival, and growth of start-up firms. Journal of Corporate Finance\/ 50 609--625

    2018

  11. [19]

    Cui, Ruomeng, Jun Li, Dennis J. Zhang. 2020 a . Reducing discrimination with reviews in the sharing economy: Evidence from field experiments on airbnb. Management Science\/ 66 (3) 1071--1094

    2020

  12. [20]

    Cui, Ruomeng, Meng Li, Qiang Li. 2020 b . Value of high-quality logistics: Evidence from a clash between sf express and alibaba. Management Science\/ 66 (9) 3879--3902

    2020

  13. [21]

    Cui, Ruomeng, Meng Li, Shichen Zhang. 2022. Ai and procurement. Manufacturing & Service Operations Management\/ 24 (2) 691--706

    2022

  14. [22]

    Davidson, James, Benjamin Liebald, Junning Liu, Palash Nandy, Taylor Van Vleet, Ullas Gargi, Sujoy Gupta, Yu He, Mike Lambert, Blake Livingston, et al. 2010. The youtube video recommendation system. Proceedings of the fourth ACM conference on Recommender systems\/ . 293--296

    2010

  15. [23]

    Dukes, Anthony, Qihong Liu. 2024. The consumption of advertising in the digital age: Attention and ad content. Management Science\/ 70 (4) 2086--2106

    2024

  16. [24]

    Ellis, Scott C., Shashank Rao, Dheeraj Raju, Thomas J. Goldsby. 2018. Rfid tag performance: Linking the laboratory to the field through unsupervised learning. Production and Operations Management\/ 27 (10) 1834--1848

    2018

  17. [25]

    Fang, Zhen, Ming Fan, Apurva Jain. 2023. Content proliferation and narrowcasting in the age of streaming media. Production and Operations Management\/ 32 (10) 3295--3310

    2023

  18. [26]

    Filippas, Apostolos, Srikanth Jagabathula, Arun Sundararajan. 2023. The limits of centralized pricing in online marketplaces and the value of user control. Management Science\/ 69 (12) 7202--7216

    2023

  19. [27]

    Goes, Paulo B, Mingfeng Lin, Ching-man Au Yeung. 2014. ``popularity effect" in user-generated content: Evidence from online product reviews. Information Systems Research\/ 25 (2) 222--238

    2014

  20. [28]

    Granulo, Armin, Christoph Fuchs, Stefano Puntoni. 2021. Preference for human (vs. robotic) labor is stronger in symbolic consumption contexts. Journal of Consumer Psychology\/ 31 (1) 72--80

    2021

  21. [29]

    He, Qinglai, Yili Hong, TS Raghu. 2021. The effects of machine-powered platform governance: An empirical study of content moderation. Available at SSRN 3767680\/

    2021

  22. [30]

    Hoiles, William, Anup Aprem, Vikram Krishnamurthy. 2017. Engagement and popularity dynamics of youtube videos and sensitivity to meta-data. IEEE Transactions on Knowledge and Data Engineering\/ 29 (7) 1426--1437

    2017

  23. [31]

    Huang, Ni, Gordon Burtch, Bin Gu, Yili Hong, Chen Liang, Kanliang Wang, Dongpu Fu, Bo Yang. 2019. Motivating user-generated content with performance feedback: Evidence from randomized field experiments. Management Science\/ 65 (1) 327--345

    2019

  24. [32]

    Huang, Ni, Probal Mojumder, Tianshu Sun, Jinchi Lv, Joseph M Golden. 2021. Not registered? please sign up first: A randomized field experiment on the ex ante registration request. Information Systems Research\/ 32 (3) 914--931

    2021

  25. [33]

    Kesavan, Saravanan, Tarun Kushwaha. 2020. Field experiment on the profit implications of merchants’ discretionary power to override data-driven decision-making tools. Management Science\/ 66 (11) 5182--5190

    2020

  26. [34]

    Kuang, Lini, Ni Huang, Yili Hong, Zhijun Yan. 2019. Spillover effects of financial incentives on non-incentivized user engagement: Evidence from an online knowledge exchange platform. Journal of Management Information Systems\/ 36 (1) 289--320

    2019

  27. [35]

    Liang, Paul Pu, Chiyu Wu, Louis-Philippe Morency, Ruslan Salakhutdinov. 2021. Towards understanding and mitigating social biases in language models. International Conference on Machine Learning\/ . PMLR, 6565--6576

    2021

  28. [36]

    generate

    Liu, Jin, Xingchen Xu, Xi Nan, Yongjun Li, Yong Tan. 2024. "generate" the future of work through ai: Empirical evidence from online labor markets. ://arxiv.org/abs/2308.05201

    2024 arXiv

  29. [37]

    Longoni, Chiara, Andrea Bonezzi, Carey K Morewedge. 2019. Resistance to medical artificial intelligence. Journal of Consumer Research\/ 46 (4) 629--650

    2019

  30. [38]

    Lysyakov, Mikhail, Siva Viswanathan. 2023. Threatened by ai: Analyzing users’ responses to the introduction of ai in a crowd-sourcing platform. Information Systems Research\/ 34 (3) 1191--1210

    2023

  31. [39]

    Malik, Haroon, Zifeng Tian. 2017. A framework for collecting youtube meta-data. Procedia Computer Science\/ 113 194--201

    2017

  32. [40]

    Narayanan, Sriram, Sridhar Balasubramanian, Jayashankar M Swaminathan. 2009. A matter of balance: Specialization, task variety, and individual learning in a software maintenance environment. Management science\/ 55 (11) 1861--1876

    2009

  33. [41]

    Nattamai Kannan, Karthik Babu, Eric Overby, Sridhar Narasimhan. 2024. Can improvements to mobile internet service help reduce digital inequality? an empirical analysis of education and overall data consumption. Management Science\/

    2024

  34. [42]

    Panniello, Umberto, Michele Gorgoglione, Alexander Tuzhilin. 2016. Research note—in carss we trust: How context-aware recommendations affect customers’ trust and other business performance measures of recommender systems. Information Systems Research\/ 27 (1) 182--196

    2016

  35. [43]

    Peng, Jiaxu, Jungpil Hahn, Ke-Wei Huang. 2023. Handling missing values in information systems research: A review of methods and assumptions. Information Systems Research\/ 34 (1) 5--26

    2023

  36. [44]

    Qiao, Dandan, Shun-Yang Lee, Andrew B Whinston, Qiang Wei. 2020. Financial incentives dampen altruism in online prosocial contributions: A study of online reviews. Information Systems Research\/ 31 (4) 1361--1375

    2020

  37. [45]

    Reisenbichler, Martin, Thomas Reutterer, David A Schweidel, Daniel Dan. 2022. Frontiers: Supporting content marketing with natural language generation. Marketing Science\/ 41 (3) 441--452

    2022

  38. [46]

    Saar-Tsechansky, Maytal, Prem Melville, Foster Provost. 2009. Active feature-value acquisition. Management Science\/ 55 (4) 664--684

    2009

  39. [47]

    Su, Yi, Qili Wang, Liangfei Qiu, Runyu Chen. 2024. Unveiling the effects of introducing ai-generated summaries in e-commerce. Available at SSRN 4872205\/

    2024

  40. [48]

    Zhang, Haoyuan Hu, Jan A

    Sun, Jiankun, Dennis J. Zhang, Haoyuan Hu, Jan A. Van Mieghem. 2022. Predicting human discretion to adjust algorithmic prescription: A large-scale field experiment in warehouse operations. Management Science\/ 68 (2) 846--865. doi:10.1287/mnsc.2021.3990

    2022

  41. [49]

    Sun, Tianshu, Lanfei Shi, Siva Viswanathan, Elena Zheleva. 2019. Motivating effective mobile app adoptions: Evidence from a large-scale randomized field experiment. Information Systems Research\/ 30 (2) 523--539

    2019

  42. [50]

    Susarla, Anjana, Ram Gopal, Jason Bennett Thatcher, Suprateek Sarker. 2023. The janus effect of generative ai: Charting the path for responsible conduct of scholarly activities in information systems. Information Systems Research\/ 34 (2) 399--408

    2023

  43. [51]

    Wang, Wen, Siqi Pei, Tianshu Sun. 2023 a . Unraveling generative ai from a human intelligence perspective: A battery of experiments. Available at SSRN 4543351\/

    2023

  44. [52]

    Wang, Zhihan (Helen), Jun Li, Di (Andrew) Wu. 2023 b . Mind the gap: Gender disparity in online learning platform interactions. Manufacturing & Service Operations Management\/ 25 (6) 2122--2141

    2023

  45. [53]

    Wasko, Molly McLure, Samer Faraj. 2005. Why should i share? examining social capital and knowledge contribution in electronic networks of practice. MIS Quarterly\/ 35--57

    2005

  46. [54]

    Wei, Wei, Xubin Ren, Jiabin Tang, Qinyong Wang, Lixin Su, Suqi Cheng, Junfeng Wang, Dawei Yin, Chao Huang. 2024. Llmrec: Large language models with graph augmentation for recommendation. Proceedings of the 17th ACM International Conference on Web Search and Data Mining\/ . 806--815

    2024

  47. [55]

    Ye, Zikun, Dennis J Zhang, Heng Zhang, Renyu Zhang, Xin Chen, Zhiwei Xu. 2023. Cold start to improve market thickness on online advertising platforms: Data-driven algorithms and field experiments. Management Science\/ 69 (7) 3838--3860

    2023

  48. [56]

    Zhang, Heng Zhang, Renyu Zhang, Zhiwei Xu, Zuo-Jun Max Shen

    Zeng, Zhiyu, Hengchen Dai, Dennis J. Zhang, Heng Zhang, Renyu Zhang, Zhiwei Xu, Zuo-Jun Max Shen. 2023. The impact of social nudges on user-generated content for social network platforms. Management Science\/ 69 (9) 5189--5208

    2023

  49. [57]

    Zhang, Xingyue, James A Dearden, Yuliang Yao. 2022. Let them stay or let them go? online retailer pricing strategy for managing stockouts. Production and Operations Management\/ 31 (11) 4173--4190

    2022

  50. [58]

    Zhou, Eric, Dokyun Lee. 2024. Generative artificial intelligence, human creativity, and art. PNAS Nexus\/ 3 (3) 052

    2024

  51. [59]

    , " * write output.state after.block = add.period write newline

    ENTRY address author booktitle chapter doi edition editor eid howpublished institution journal key month note number organization pages publisher school series title type url volume year label extra.label sort.label short.list INTEGERS output.state before.all mid.sentence afte...

  52. [60]

    write newline

    " write newline "" before.all 'output.state := FUNCTION n.dashify 't := "" t empty not t #1 #1 substring "-" = t #1 #2 substring "--" = not "--" * t #2 global.max substring 't := t #1 #1 substring "-" = "-" * t #2 global.max substring 't := while if t #1 #1 substring * t #2 gl...

Pith tools

Reviewed August 11, 2026 · model on record in the stance chip above.