Xin Shen

Currently, I am a first second third year Ph.D. student at the Uppsala University under the supervision of Ece Calikus and Christian Rohner.

I graduated with a B.Eng. degree in Computer Science and Technology from Ningbo University, China, in 2022, advised by Chengbin Peng.

Network analysis has always fascinated me, particularly clustering and link prediction, and I aspire to become a creative scholar in this field. I am passionate about developing algorithms that address various applications of graph analysis, including social and brain networks. My current journey of exploration revolves around the intersection of graph analysis, machine learning, and statistics, which requires a deep understanding of these fields.


[Curriculum Vitae] (updated Sep., 2025)
[Email: xin.shen@it.uu.se]

News
  • [March 2023]: My work with Matteo Magnani and Christian Rohner was accepted for a regular talk in a parallel session at NetSci 2023.
  • [March 2024]: My work with Matteo Magnani and Christian Rohner was accepted for a oral presentation in a parallel session at Sunbelt Conference 2024.
  • [July 2025]: The manuscript "Probabilistic social networks" was accepted at "Encyclopedia of Social Network Analysis and Mining".
Research Activities
Ongoing Project: Algorithms and Benchmarking in Uncertain Networks
Uncertainty is an inherent property when modeling a system as a network because of its randomness, inaccuracy of measurements, or interpretation. Uncertainty is modelled by associating each edge with a probability of existence, forming a probabilistic network. The thesis currently contains two parts. The first part includes the paper 'On the accurate computation of expected modularity in probabilistic networks': many techniques are developed in deterministic or weighted networks, but most don't consider the edge probability distribution. In this paper, we generate an efficient algorithm called Fast/Fourier Possible-World Partitioning (FPWP), which aims to compute the probability distribution of modularity and its expected value on uncertain networks. The second part contains the paper 'Benchmarking Clustering Methods for Uncertain Networks': during the observation of application of clustering algorithms on realistic networks, we identify two critical gaps in understanding how uncertainty influences clustering outcomes: (i) How do existing approaches for uncertain networks perform in comparative and specific contexts? (ii) Can clustering methods for weighted networks be used in uncertain networks? We designed specific research questions based on those gaps and formulated them in our benchmark paper to discuss those questions systematically.

Teaching Activities
    • 1DL931 Independent Project in Sociotechnical Systems Engineering - IT Systems 2023.01-2023.06, TA
    • 1DL360 Data Mining I 2023.08 – 2024.01, TA
    • 1DL201 Programkonstruktion och datastrukturer 2023.08 – 2024.01, TA