pith. sign in

arxiv: 2110.12569 · v2 · submitted 2021-10-25 · 💻 cs.SI · cs.CY

Conductance and Influence-Capital: Modeling Online Social Influence

Rohit Ram , Marian-Andrei Rizoiu This is my paper

Pith reviewed 2026-05-24 13:05 UTC · model grok-4.3

classification 💻 cs.SI cs.CY
keywords social influenceTwitterCOVID-19influence modelmisinformationnetwork conductanceinfluence capitalsocial media
0
0 comments X

The pith

A model using network conductance and influence capital shows executives and media exert more online influence than health experts during COVID discussions.

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

The paper introduces a Generalized Influence Model that adds two mechanisms drawn from psychosocial ideas: how easily influence spreads through a network and how influence capital is distributed among users. This data-driven approach is tested on millions of Twitter accounts discussing COVID-19 and compared against follower count as a baseline. The model produces different influence rankings that favor certain occupations over others and links higher influence to greater spread of misinformation. It claims these rankings better reflect real influence dynamics than simple popularity metrics.

Core claim

The Generalized Influence Model incorporates conductance of the diffusion network and influence-capital distribution to quantify influence for over 21.5 million Twitter users. When applied to COVID-19 content, the model ranks executives, media, and military figures higher in influence than life scientists and healthcare professionals, while also showing that some high-influence occupations spread more misinformation.

What carries the argument

Generalized Influence Model (GIM) operationalizing conductance of the diffusion network and influence-capital distribution from Twitter data to measure influence.

If this is right

  • GIM outperforms existing state-of-the-art influence models on Twitter data.
  • GIM reduces biases present in follower-count rankings.
  • Executives, media, and military figures show higher influence than pandemic experts in COVID discussions.
  • Occupations with highest influence also spread the most misinformation according to existing datasets.
  • Expert information dissemination may be less effective in crises than assumed.

Where Pith is reading between the lines

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

  • Platforms might replace follower counts with conductance-based scores for ranking accounts in recommendation systems.
  • Crisis communication strategies could prioritize reaching high-conductance occupations rather than credentialed experts alone.
  • The same mechanisms could be tested on other platforms to check whether occupation influence patterns hold beyond Twitter.

Load-bearing premise

The two new mechanisms can be computed from Twitter data in a way that captures genuine influence without adding fresh biases or circular definitions.

What would settle it

An experiment that measures real-world adoption of information from users ranked by GIM versus follower count and finds no difference or reversal in predictive accuracy would falsify the superiority claim.

Figures

Figures reproduced from arXiv: 2110.12569 by Marian-Andrei Rizoiu, Rohit Ram.

Figure 1
Figure 1. Figure 1: Schema of the Generalized Influence Model (GIM). [PITH_FULL_IMAGE:figures/full_fig_p001_1.png] view at source ↗
Figure 2
Figure 2. Figure 2: Evaluating GIM in three steps. (a) Select worker interface design features. The blue line shows the relation between the average accuracy (y-axis) and the noise (𝜆) (averaged over 100 simulations, 𝑛 = 500 targets, 𝐵 = 30, 000). The points show ablations of design features (see [PITH_FULL_IMAGE:figures/full_fig_p007_2.png] view at source ↗
Figure 3
Figure 3. Figure 3: (a) Residuals (relative to the empirical follower ranking) for the follower count (top) and GIM (bottom) against the [PITH_FULL_IMAGE:figures/full_fig_p009_3.png] view at source ↗
read the original abstract

Human interactions are mediated by social influence. During crises like the COVID-19 pandemic, social influence determines whether life-saving information is adopted or immunization campaigns meet their targets. The literature on online social influence presents notable limitations across disciplines. Psychosocial approaches characterize the nature of influence by measuring how social factors impact these phenomena, but lack computational modeling capabilities and rely on slow, non-scalable measurement methods. Conversely, computational approaches, while data-driven, often fail to incorporate critical social factors. Our work bridges this gap through two main contributions. First, we present a data-driven Generalized Influence Model (GIM) incorporating two novel psychosocial-inspired mechanisms: the conductance of the diffusion network and the influence-capital distribution. GIM not only outperforms existing state-of-the-art approaches but also corrects the inherent biases introduced by the widely used follower count metric. Second, we empirically test long-held sociological hypotheses regarding influence, social class, and expertise by applying GIM to COVID-19 discussions. We quantify the influence and content veracity for more than 21.5 million X/Twitter users in relation to their professions. Our model suggests that executives, media, and military figures exert greater influence than pandemic-related experts such as life scientists and healthcare professionals. Worryingly, by leveraging existing COVID-19 misinformation datasets, we show that some of the most influential occupations also spread the most misinformation. These findings raise questions about the effectiveness of information dissemination by experts in situations of crisis.

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

3 major / 0 minor

Summary. The manuscript introduces a Generalized Influence Model (GIM) that augments data-driven influence modeling with two psychosocial-inspired mechanisms: conductance of the diffusion network and influence-capital distribution. It claims that GIM outperforms existing state-of-the-art methods while correcting biases inherent in the follower-count metric, and applies the model to a large Twitter dataset on COVID-19 discussions to quantify influence and content veracity across occupations, concluding that executives, media, and military figures exert greater influence than pandemic-related experts and that some high-influence occupations also propagate more misinformation.

Significance. If the conductance and influence-capital mechanisms can be shown to be non-circular with standard network measures and externally validated, the work would usefully bridge computational and psychosocial perspectives on influence. The scale of the empirical analysis (21.5 million users) and the substantive claims about occupational influence and misinformation during crises would be of interest to both computational social science and public-health communication research.

major comments (3)
  1. [Abstract] Abstract: the central claims of outperformance over SOTA and bias correction relative to follower count are asserted without any reference to the specific baselines, evaluation metrics, statistical tests, or data-exclusion criteria employed; this absence is load-bearing because the soundness of the model cannot be assessed from the provided text.
  2. [Abstract] Abstract and introduction: no explicit formulas, algorithms, or pseudocode are supplied for extracting conductance from the diffusion graph or for allocating influence-capital; without these definitions it is impossible to verify that the two mechanisms are independent of degree, PageRank, or activity counts already implicit in the Twitter diffusion data.
  3. [Abstract] Abstract: the empirical finding that executives, media, and military figures exert greater influence than life scientists and healthcare professionals, together with the misinformation correlation, rests on the untested assumption that the new mechanisms capture genuine psychosocial signal rather than re-expressing network structure; an external validation (e.g., against surveys or adoption rates) is required to support this claim.

Simulated Author's Rebuttal

3 responses · 0 unresolved

We thank the referee for the constructive and detailed feedback. We address each major comment point by point below, indicating planned revisions where appropriate.

read point-by-point responses

  1. Referee: [Abstract] Abstract: the central claims of outperformance over SOTA and bias correction relative to follower count are asserted without any reference to the specific baselines, evaluation metrics, statistical tests, or data-exclusion criteria employed; this absence is load-bearing because the soundness of the model cannot be assessed from the provided text.

    Authors: We agree that the abstract would be strengthened by greater specificity. In the revised manuscript we will expand the abstract to name the primary SOTA baselines (follower-count, PageRank, and standard diffusion models), the evaluation metrics (influence ranking precision and prediction accuracy), the statistical tests employed, and the data-exclusion criteria applied to the 21.5 million user COVID-19 Twitter dataset. revision: yes

  2. Referee: [Abstract] Abstract and introduction: no explicit formulas, algorithms, or pseudocode are supplied for extracting conductance from the diffusion graph or for allocating influence-capital; without these definitions it is impossible to verify that the two mechanisms are independent of degree, PageRank, or activity counts already implicit in the Twitter diffusion data.

    Authors: The Methods section of the full manuscript defines conductance as the normalized diffusion flow across edges and influence-capital as a per-node resource allocation derived from local network topology and activity. To address the concern directly, we will insert the core formulas together with a short pseudocode block into the Introduction of the revised version, explicitly showing that neither quantity reduces to degree, PageRank, or raw activity counts. revision: yes

  3. Referee: [Abstract] Abstract: the empirical finding that executives, media, and military figures exert greater influence than life scientists and healthcare professionals, together with the misinformation correlation, rests on the untested assumption that the new mechanisms capture genuine psychosocial signal rather than re-expressing network structure; an external validation (e.g., against surveys or adoption rates) is required to support this claim.

    Authors: We acknowledge that external validation against surveys or adoption rates would further strengthen the psychosocial interpretation. The present work is a large-scale computational study; we will add an explicit Limitations and Future Work subsection that discusses this gap and outlines feasible validation routes. We maintain that the performance improvement and bias-correction results already indicate that the two mechanisms capture signal beyond standard network measures, but we accept that this remains an assumption pending external corroboration. revision: partial

Circularity Check

0 steps flagged

No circularity: model described at abstract level with no equations or self-referential derivations shown

full rationale

The provided text consists solely of the abstract and high-level claims. No equations, parameter-fitting procedures, or derivation steps are present, so no load-bearing step can be shown to reduce to its own inputs by construction. The model is introduced as data-driven with two novel mechanisms, but without explicit formulas or citations that would allow verification of self-definition, fitted-input predictions, or imported uniqueness, the derivation chain cannot be walked and no circularity is exhibited. This is the expected honest non-finding when the source material supplies no mathematical content.

Axiom & Free-Parameter Ledger

0 free parameters · 0 axioms · 0 invented entities

Abstract-only review provides no equations, so free parameters, axioms, and invented entities cannot be enumerated; the model introduces conductance and influence-capital as new constructs whose definitions and independence from data fitting are unknown.

pith-pipeline@v0.9.0 · 5788 in / 1241 out tokens · 23502 ms · 2026-05-24T13:05:58.010845+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

73 extracted references · 73 canonical work pages

  1. [1]

    Sakshi Agarwal and Shikha Mehta. 2020. Effective influence estimation in twitter using temporal, profile, structural and interaction characteristics. Information Processing & Management 57, 6 (2020), 102321

    work page 2020

  2. [2]

    Nir Ailon. 2008. Reconciling real scores with binary comparisons: A new logistic based model for ranking. Advances in Neural Information Processing Systems (2008)

    work page 2008

  3. [3]

    Online Appendix. 2021. Appendix: Conductance and Social Capital: Modeling and Empirically Measuring Online Social Influence. https://www.dropbox.com/ s/963oc5fx4f7412p/Supplementary_Material.pdf?dl=0

    work page 2021

  4. [4]

    Ines Arous, Jie Yang, Mourad Khayati, and Philippe Cudré-Mauroux. 2020. Open- crowd: A human-ai collaborative approach for finding social influencers via open-ended answers aggregation. In Proceedings of The Web Conference 2020 . 1851–1862

    work page 2020

  5. [5]

    Solomon E Asch. 1961. Effects of group pressure upon the modification and dis- tortion of judgments. In Documents of gestalt psychology. University of California Press, 222–236

    work page 1961

  6. [6]

    Yousra Asim, Ahmad Kamran Malik, Basit Raza, and Ahmad Raza Shahid. 2019. A trust model for analysis of trust, influence and their relationship in social network communities. Telematics and Informatics 36 (2019), 94–116

    work page 2019

  7. [7]

    Eytan Bakshy, Jake M Hofman, Winter A Mason, and Duncan J Watts. 2011. Everyone’s an influencer: quantifying influence on twitter. InProceedings of the fourth ACM international conference on Web search and data mining . 65–74

    work page 2011

  8. [8]

    Adam J Berinsky, Gregory A Huber, and Gabriel S Lenz. 2012. Evaluating on- line labor markets for experimental research: Amazon. com’s Mechanical Turk. Political analysis 20, 3 (2012), 351–368

    work page 2012

  9. [9]

    Sushil Bikhchandani, David Hirshleifer, and Ivo Welch. 1992. A theory of fads, fashion, custom, and cultural change as informational cascades. Journal of political Economy 100, 5 (1992), 992–1026. Conductance and Social Capital: Modeling and Empirically Measuring Online Social Influence Conference’17, July 2017, Washington, DC, USA −0.5 0.0 0.5 0.00 0.25 ...

    work page 1992

  10. [10]

    Ralph Allan Bradley and Milton E Terry. 1952. Rank analysis of incomplete block designs: I. The method of paired comparisons. Biometrika (1952)

    work page 1952

  11. [11]

    Michael Buhrmester, Tracy Kwang, and Samuel D Gosling. 2016. Amazon’s Mechanical Turk: A new source of inexpensive, yet high-quality data? (2016)

    work page 2016

  12. [12]

    Manuela Cattelan. 2012. Models for paired comparison data: A review with emphasis on dependent data. Statist. Sci. (2012)

    work page 2012

  13. [13]

    Manuela Cattelan, Cristiano Varin, and David Firth. 2013. Dynamic Bradley– Terry modelling of sports tournaments. Journal of the Royal Statistical Society: Series C (Applied Statistics) 62, 1 (2013), 135–150

    work page 2013

  14. [14]

    Meeyoung Cha, Hamed Haddadi, Fabricio Benevenuto, and Krishna Gummadi

    work page

  15. [15]

    In Pro- ceedings of the international AAAI conference on web and social media , Vol

    Measuring user influence in twitter: The million follower fallacy. In Pro- ceedings of the international AAAI conference on web and social media , Vol. 4

    work page

  16. [16]

    Minje Choi, Luca Maria Aiello, Krisztián Zsolt Varga, and Daniele Quercia. 2020. Ten social dimensions of conversations and relationships. In Proceedings of The Web Conference 2020. 1514–1525

    work page 2020

  17. [17]

    Robert B Cialdini. 2001. Influence: Science and practice

    work page 2001

  18. [18]

    David Crandall, Dan Cosley, Daniel Huttenlocher, Jon Kleinberg, and Siddharth Suri. 2008. Feedback effects between similarity and social influence in online communities. In Knowledge Discovery and Data Mining (KDD’08). 160–168

    work page 2008

  19. [19]

    Nan Du, Le Song, Manuel Gomez-Rodriguez, and Hongyuan Zha. 2013. Scalable influence estimation in continuous-time diffusion networks. NIPS (2013)

    work page 2013

  20. [20]

    Nan Du, Le Song, Manuel Gomez-Rodriguez, and Hongyuan Zha. 2013. Scalable influence estimation in continuous-time diffusion networks. InAdvances in Neural Information Processing Systems (NIPS) , Vol. 26. NIH Public Access, 3147

    work page 2013

  21. [21]

    Erick Elejalde, Leo Ferres, and Eelco Herder. 2018. On the nature of real and perceived bias in the mainstream media. PloS one 13, 3 (2018), e0193765

    work page 2018

  22. [22]

    National Center for O*NET Development. [n.d.]. O*NET OnLine,. https://www. onetonline.org/. Accessed: 2021-06-30

    work page 2021

  23. [23]

    Chiara Francalanci and Ajaz Hussain. 2017. Influence-based Twitter browsing with NavigTweet. Information Systems 64 (2017), 119–131

    work page 2017

  24. [24]

    Terrill L Frantz, Marcelo Cataldo, and Kathleen M Carley. 2009. Robustness of centrality measures under uncertainty: Examining the role of network topology. Computational and Mathematical Organization Theory 15, 4 (2009), 303–328

    work page 2009

  25. [25]

    Rahul Goel, Sandeep Soni, Naman Goyal, John Paparrizos, Hanna Wallach, Fer- nando Diaz, and Jacob Eisenstein. 2016. The social dynamics of language change in online networks. In International conference on social informatics . Springer, 41–57

    work page 2016

  26. [26]

    Google. [n.d.]. CLD3. https://github.com/google/cld3

    work page

  27. [27]

    T Graham and TR Keller. 2020. Bushfires, bots and arson claims: Australia flung in the global disinformation spotlight. The Conversation 10 (2020)

    work page 2020

  28. [28]

    Yupeng Gu, Yizhou Sun, Yanen Li, and Yang Yang. 2018. Rare: Social rank regulated large-scale network embedding. In Proceedings of the 2018 World Wide Web Conference. 359–368

    work page 2018

  29. [29]

    Avni Gulati and Magdalini Eirinaki. 2019. With a little help from my friends (and their friends): Influence neighborhoods for social recommendations. In The World Wide Web Conference. 2778–2784

    work page 2019

  30. [30]

    Alan G Hawkes. 1971. Spectra of some self-exciting and mutually exciting point processes. Biometrika (1971)

    work page 1971

  31. [31]

    Alan G Hawkes and David Oakes. 1974. A cluster process representation of a self-exciting process. Journal of Applied Probability (1974)

    work page 1974

  32. [32]

    Charles AR Hoare. 1962. Quicksort. Comput. J. (1962)

    work page 1962

  33. [33]

    Joann Horai, Nicholas Naccari, and Elliot Fatoullah. 1974. The effects of expertise and physical attractiveness upon opinion agreement and liking. Sociometry (1974), 601–606

    work page 1974

  34. [34]

    Melissa G Keith, Louis Tay, and Peter D Harms. 2017. Systems perspective of Ama- zon Mechanical Turk for organizational research: Review and recommendations. Frontiers in psychology 8 (2017), 1359

    work page 2017

  35. [35]

    David Kempe, Jon Kleinberg, and Éva Tardos. 2003. Maximizing the spread of influence through a social network. In Proceedings of the ninth ACM SIGKDD international conference on Knowledge discovery and data mining . 137–146

    work page 2003

  36. [36]

    Joost Kruis and Denny Borsboom. 2020. Patrick Mair: Modern Psychometrics with R

    work page 2020

  37. [37]

    Richard Lenton. 2006. Using the method of paired comparisons in non-designed experiments. Ph.D. Dissertation. Griffith University

    work page 2006

  38. [38]

    Erik Lewis and George Mohler. 2011. A nonparametric EM algorithm for multi- scale Hawkes processes. Journal of Nonparametric Statistics (2011)

    work page 2011

  39. [39]

    Yuchen Li, Ju Fan, Yanhao Wang, and Kian-Lee Tan. 2018. Influence maximization on social graphs: A survey. IEEE Transactions on Knowledge and Data Engineering 30, 10 (2018), 1852–1872

    work page 2018

  40. [40]

    Lynda C Lin, Yang Qu, and Eva H Telzer. 2018. Intergroup social influence on emotion processing in the brain. Proceedings of the National Academy of Sciences 115, 42 (2018), 10630–10635

    work page 2018

  41. [41]

    Luca Luceri, Torsten Braun, and Silvia Giordano. 2019. Analyzing and inferring human real-life behavior through online social networks with social influence deep learning. Applied network science 4, 1 (2019), 1–25

    work page 2019

  42. [42]

    Luca Luceri, Alberto Vancheri, Torsten Braun, and Silvia Giordano. 2017. On the social influence in human behavior: Physical, homophily, and social communities. In International Conference on Complex Networks and their Applications . Springer, 856–868

    work page 2017

  43. [43]

    Marco Lui and Timothy Baldwin. 2012. langid. py: An off-the-shelf language identification tool. In Proceedings of the ACL 2012 system demonstrations

    work page 2012

  44. [44]

    Winter A Mason, Frederica R Conrey, and Eliot R Smith. 2007. Situating social influence processes: Dynamic, multidirectional flows of influence within social networks. Personality and social psychology review 11, 3 (2007), 279–300

    work page 2007

  45. [45]

    Lucas Maystre and Matthias Grossglauser. 2017. Just sort it! A simple and effective approach to active preference learning. In International Conference on Machine Learning

    work page 2017

  46. [46]

    Stanley Milgram and Christian Gudehus. 1978. Obedience to authority

    work page 1978

  47. [47]

    Swapnil Mishra, Marian-Andrei Rizoiu, and Lexing Xie. 2016. Feature driven and point process approaches for popularity prediction. In Proceedings of the 25th ACM international on conference on information and knowledge management . 1069–1078

    work page 2016

  48. [48]

    John Moody. 1989. Fast learning in multi-resolution hierarchies. In Advances in neural information processing systems . 29–39

    work page 1989

  49. [49]

    Subhabrata Mukherjee and Stephan Günnemann. 2019. GhostLink: Latent Net- work Inference for Influence-aware Recommendation. In The World Wide Web Conference. 1310–1320

    work page 2019

  50. [50]

    Theodore M Newcomb. 1953. An approach to the study of communicative acts. Psychological review 60, 6 (1953), 393. Conference’17, July 2017, Washington, DC, USA Rohit Ram and Marian-Andrei Rizoiu

    work page 1953

  51. [51]

    Maximilian Nickel and Matthew Le. 2021. Modeling Sparse Information Diffusion at Scale via Lazy Multivariate Hawkes Processes. In Proceedings of the Web Conference 2021. 706–717

    work page 2021

  52. [52]

    Andrzej Nowak, Robin R Vallacher, Marek Kus, and Jakub Urbaniak. 2005. The dy- namics of societal transition: Modeling nonlinear change in the Polish economic system. International Journal of Sociology 35, 1 (2005), 65–88

    work page 2005

  53. [53]

    Lawrence Page, Sergey Brin, Rajeev Motwani, and Terry Winograd. 1999. The PageRank citation ranking: Bringing order to the web. Technical Report. Stanford InfoLab

    work page 1999

  54. [54]

    Sancheng Peng, Yongmei Zhou, Lihong Cao, Shui Yu, Jianwei Niu, and Weijia Jia

    work page

  55. [55]

    Journal of Network and Computer Applications 106 (2018), 17–32

    Influence analysis in social networks: A survey. Journal of Network and Computer Applications 106 (2018), 17–32

    work page 2018

  56. [56]

    Jiezhong Qiu, Jian Tang, Hao Ma, Yuxiao Dong, Kuansan Wang, and Jie Tang

    work page

  57. [57]

    In Proceedings of the 24th ACM SIGKDD International Conference on Knowledge Discovery & Data Mining

    Deepinf: Social influence prediction with deep learning. In Proceedings of the 24th ACM SIGKDD International Conference on Knowledge Discovery & Data Mining. 2110–2119

    work page

  58. [58]

    Jeff Riddell, Alisha Brown, Ivor Kovic, and Joshua Jauregui. 2017. Who are the most influential emergency physicians on Twitter? Western Journal of Emergency Medicine 18, 2 (2017), 281

    work page 2017

  59. [59]

    Marian-Andrei Rizoiu, Timothy Graham, Rui Zhang, Yifei Zhang, Robert Ackland, and Lexing Xie. 2018. # debatenight: The role and influence of socialbots on twitter during the 1st 2016 us presidential debate. In AAAI

    work page 2018

  60. [60]

    Daniel M Romero, Wojciech Galuba, Sitaram Asur, and Bernardo A Huberman

    work page

  61. [61]

    In Joint European Conference on Machine Learning and Knowledge Discovery in Databases

    Influence and passivity in social media. In Joint European Conference on Machine Learning and Knowledge Discovery in Databases . Springer, 18–33

    work page

  62. [62]

    Burr Settles. 2009. Active learning literature survey. (2009)

    work page 2009

  63. [63]

    Nihar B Shah, Sivaraman Balakrishnan, Joseph Bradley, Abhay Parekh, Kannan Ramchandran, and Martin J Wainwright. 2016. Estimation from pairwise compar- isons: Sharp minimax bounds with topology dependence. The Journal of Machine Learning Research (2016)

    work page 2016

  64. [64]

    Arlei Silva, Sara Guimarães, Wagner Meira Jr, and Mohammed Zaki. 2013. Pro- fileRank: finding relevant content and influential users based on information diffusion. In Proceedings of the 7th Workshop on Social Network Mining and Anal- ysis. 1–9

    work page 2013

  65. [65]

    Luke Sloan, Jeffrey Morgan, Pete Burnap, and Matthew Williams. 2015. Who tweets? Deriving the demographic characteristics of age, occupation and social class from Twitter user meta-data. PloS one 10, 3 (2015), e0115545

    work page 2015

  66. [66]

    Steven T Smith, Edward K Kao, Danelle C Shah, Olga Simek, and Donald B Rubin

    work page

  67. [67]

    In 2018 IEEE Statistical Signal Processing Workshop (SSP)

    Influence estimation on social media networks using causal inference. In 2018 IEEE Statistical Signal Processing Workshop (SSP) . IEEE, 328–332

    work page 2018

  68. [68]

    Anton Tsitsulin, Davide Mottin, Panagiotis Karras, and Emmanuel Müller. 2018. Verse: Versatile graph embeddings from similarity measures. InProceedings of the 2018 World Wide Web Conference. 539–548

    work page 2018

  69. [69]

    Kristi Tsukida and Maya R Gupta. 2011. How to analyze paired comparison data. Technical Report. WASHINGTON UNIV SEATTLE DEPT OF ELECTRICAL ENGINEERING

    work page 2011

  70. [70]

    Biao Xiang, Qi Liu, Enhong Chen, Hui Xiong, Yi Zheng, and Yu Yang. 2013. Pagerank with priors: An influence propagation perspective. In Twenty-Third International Joint Conference on Artificial Intelligence

    work page 2013

  71. [71]

    Wenhui Yu and Zheng Qin. 2019. Spectrum-enhanced pairwise learning to rank. In The World Wide Web Conference. 2247–2257

    work page 2019

  72. [72]

    Ernst Zermelo. 1929. Die berechnung der turnier-ergebnisse als ein max- imumproblem der wahrscheinlichkeitsrechnung. Mathematische Zeitschrift (1929)

    work page 1929

  73. [73]

    Which user is the proxy user most likely to retweet?

    Yi Zhao, Haixu Xi, and Chengzhi Zhang. 2021. Exploring Occupation Differ- ences in Reactions to COVID-19 Pandemic on Twitter. Data and Information Management 5, 1 (2021), 110–118. Conductance and Social Capital: Modeling and Empirically Measuring Online Social Influence Conference’17, July 2017, Washington, DC, USA APPENDIX This document accompanies Condu...

    work page 2021