Dr. Manuel Sebastian Mariani



My research focuses on three main, interrelated pillars: social influence, network science, and algorithmic bias.

Social influence:

A common hypothesis in social influence theories is that the diffusion of innovation is often mediated by a relatively small group of special agents, referred to as influentials or influencers, who may accelerate or impede the large-scale adoption of an innovation. Despite our common experience that some agents -- whether individuals, teams, research institutes, firms, or pieces of information -- are more influential than others, there is not yet agreement on several interrelated questions: how do influencers exert their influence on their peers? How to best detect influencers? How to best target them for marketing or immunization purposes? What is their global impact for innovation diffusion in real-world settings? My research is driven by these research questions, and it aims to (1) design and validate improved methods for the influencers identification; (2) understand the impact of influencers for real-world diffusion processes such as product and technology adoption; (3) exploit influencers to predict the future success of recent agents. I focus on online markets where individuals buy products, physical markets where individuals visit shops, scientific and technological innovation.

Network science:

One of the main goals of network science is to detect (possibly important) structural patterns in networks. Crucially, different structural patterns can emerge as a result of different network formation mechanisms, and have important implications for systemic stability and diffusion processes. My research aims to design improved metrics and algorithms to detect structural patterns in networks, and to assess their implications. Patterns that I'm currently investigating include the community structure of networks and nestedness. Some questions that drive me are: How do temporal effects in the growth of networks affect the communities detected by popular community detection algorithms? How to improve such algorithms by including temporal information? How can we detect communities that exhibit a specific internal organization? How can we exploit observed structural patterns to predict missing connections?

Algorithmic bias:

As our society becomes increasingly digitalized, our decisions are often influenced by quantitative metrics. For example, since it is impossible to manually browse large portions of the Web, we tend to only visit the web pages that are ranked at the top by our favorite search engine. As we are routinely exposed to and influenced by quantitative metrics, it becomes critical to properly assess their biases, to correct for them, and to evaluate the potential long-term impact and useful applications of the adopted metrics. My research addresses these questions, with a special focus on information networks (e.g., paper and patent citation networks). For example, my recent works have shown that suppressing the bias by age of citation-based metrics of scientific and technological impact allows us to single out earlier and more effectively milestone papers and patents.

Research Interests

  • Social influence
  • Innovation diffusion
  • Network science
  • Algorithmic bias


M. S. Mariani, Z.-M. Ren, J. Bascompte, C. J. Tessone, Nestedness in complex networks: Observation, emergence, and implications, Physics Reports 813, 1–90 (2019).

M. S. Mariani, M. Medo, F. Lafond, Early identification of important patents: Design and validation of citation network metrics, Technological Forecasting and Social Change 146, 644–654 (2019).

M. Medo, A. Zeng, Y.-C. Zhang, M. S. Mariani, Optimal timescale for community detection in growing networks, New Journal of Physics (2019),

H. Liao, M.-K. Liu, M. S. Mariani, M. Zhou, X. Wu, Temporal similarity metrics for latent network reconstruction: The role of time-lag decay, Information Sciences, 489, 182–192 (2019).

S. Zhang, M. Medo, L. Lü, M. S. Mariani, The long-term impact of ranking algorithms in growing networks, Information Sciences, 488, 257–271 (2019). 

F. Zhou, L. Lu, M. S. Mariani, Fast influencers in complex networks, Communications in Nonlinear Science and Numerical Simulation 74, 69–83 (2019).

M. Medo, M. S. Mariani, L. Lü, Link Prediction in Bipartite Nested Networks, Entropy 20 (10), 777 (2018).

F. Iannelli, M. S. Mariani, I. Sokolov, Influencers identification in complex networks through reaction-diffusion dynamics, Physical Review E 98, 062302 (2018).

J.-H. Lin, C. J. Tessone, M. S. Mariani, "Nestedness Maximization in Complex Networks through the Fitness-Complexity Algorithm", Entropy 20 (10), 768 (2018).

A. Solé-Ribalta, C. J. Tessone, M. S. Mariani, J. Borge-Holthoefer, "Revaling In-Block Nestedness: Detection and benchmarking", Physical Review E 97, 062302 (2018).

Z.-M. Ren, M. S. Mariani, M. Medo, Y.-C. Zhang, ''Randomizing growing networks with a time-respecting null model``, Physical Review E 97, 052311 (2018).

G. Vaccario, M. Medo, N. Wider, and M. S. Mariani: “Quantifying and suppressing ranking bias in a large citation network”, Journal of Informetrics, pp. 766-782, vol. 11, (2017).

H. Liao, M. S. Mariani, M. Medo, Y.-C. Zhang, and M.-Y. Zhou: “Ranking in evolving complex networks”, Physics Reports, pp.1-54, vol.689, (2017).


"Ranking bias in networks: detection and suppression", Poster at the conference NetSci 2018, Paris, France.

"Influencers identification in complex networks through reaction-diffusion dynamics", Poster at the conference NetSci 2018, Paris, France. 

"Influencers identification in complex networks through reaction-diffusion dynamics", Invited talk at the International Conference on Frontiers of Electronic Science and Technology, Chengdu, China.

“The structure of information, economic, and social networks”, Invited talk at Zhejian University of Technology, Hangzhou, China, January 2018.
“Early identification of significant papers and patents in citation networks”, Contributed talk at NetSciX 2018, Hangzhou, China, January 2018.
“Identification of significant papers and patents in citation networks: citation count or PageRank?” Invited talk at the workshop “Complex networks: from socio-economic systems to biology and brain”, Lipari, Italy, September 2017:
“A time-respecting null model to explore the structure of growing networks”, Contributed talk at the SigmaPhi 2017 conference, Corfu, Greece, July 2017
“Early-identification of significant nodes in growing networks”, Contributed talk at the SigmaPhi 2017 conference, Corfu, Greece, July, 2017.
“The temporal dimension of ranking in growing networks”, Invited talk at Alibaba Business College, Hangzhou, China, June 2017
“The temporal dimension of complex networks and its application to ranking, null models and community detection", Invited talk at Shanghai University of Economics and Finance, Shanghai, China, April 2017.
“Understanding and improving the performance of PageRank in growing networks”, Invited talk at College of Computer Science and Software Engineering, Shenzhen University, Shenzhen, China, April 2017.