Linlin Jia

About Me

Linlin Jia is a postdoctoral research in computer science in Pattern Recognition Group (PRG), Institute of Computer Science, Faculty of Science of the University of Bern. Before that, he was a postdoc in the COBRA Lab, INSA Rouen, Normandie Université, France. He received his doctoral degree in 2021 in computer science from Laboratoire d’Informatique, de Traitement de l’Information et des Systèmes (LITIS) of INSA Rouen Normandie, Normandie Université, France, under the supervision of Prof. Paul Honeine and Prof. Benoit Gaüzère, with the PH.D. thesis "machine learning and patterns recognition in chemoinformatics", focusing on graph kernels and graph edit distances in machine learning, and the graph pre-image problem. Before that, he received the B.E. degree in information engineering, in 2014, and the M.E. degree in software engineering, in 2017, both from Xi’an Jiaotong University, China. For more information, please download his CV or visit his profile on LinkedIn.

Research interests

- Graph kernel within chemoinformatics.

- Graph edit distances.

- Graph Neural Networks (GNNs).

- pre-image problems on graphs.

- Graph machine learning in computational chemistry.

Projects

Novel State-of-the-Art Graph Matching Algorithms (2023-Now): This project is funded by the Swiss National Science Foundation (SNSF). In the project, I work on the graph matching problem, develop novel graph matching algorithms and explore their applications.

The project OCTOPUSSY and 2 other projects (2021-now): OCTOPUSSY (Original Computational Techniques for the Optimization of Polymers Using Sustainable SYnthesis) is a collaborative projct between the COBRA lab, the LITIS lab, Université Normandie and Centre for Sustainable, and Circular Technologies, University of Bath. I am one of the main executors in the project, applying graph machine learning methods to the optimization of polymers. I am also a main participant of 2 other projects lying in the interdiscipline of machine learning and computational chemistry.

The grant APi (2018-2021): The grant APi (Apprivoiser la Pré-image) is funded by the French national research agency (ANR) for research of the pre-image in machine learning for structured data. I worked on the pattern recognition in discrete structured spaces. I have published two journal paper [J1, J2], two workshop papers [W1, W2], and a library available online on GitHub (github.com/jajupmochi/graphkit- learn).

Research on Service-oriented Programmable Control and Scheduling Scheme for Software Defined Network (2014-2017): I applied extreme learning machine to predict network mobility and published a patent on it [P1].

Publications

See my Google Scholar page and ResearchGate page for the up-to-date list of publications.

[J0]

Linlin Jia, Xiao Ning, Benoit Gaüzère, Paul Honeine, and Kaspar Riesen. Bridging Distinct Spaces in Graph-based Machine Learning. ACPR, 2023. [pdf] [code]

[J1]

Linlin Jia, Benoit Gaüzère, and Paul Honeine. Graph Kernels Based on Linear Patterns:Theoretical and Experimental Comparisons. Expert Systems with Applications, 2021. [preprint] [code] [doi]

[J2]

Linlin Jia, Benoit Gaüzère, and Paul Honeine. graphkit-learn: A python library for graph kernels based on linear patterns. Pattern Recognition Letters, 2021. [preprint] [code] [doi]

[J3]

Linlin Jia, Vincent Tognetti, Laurent Joubert, Benoit Gaüzère and Paul Honeine. A Study on the Stability of Graph Edit Distance Heuristics. Electronics, 2022. [preprint] [doi]

[W1]

Linlin Jia, Benoit Gaüzère, and Paul Honeine. A Graph Pre-image Method Based on Graph Edit Distances. Proceedings of IAPR Joint International Workshops on Statistical techniques in Pattern Recognition (SPR 2020) and Structural and Syntactic Pattern Recognition (SSPR 2020). 2021. [preprint] [video] [slides] [doi]

[W2]

Linlin Jia, Benoit Gaüzère, Florian Yger and Paul Honeine. A Metric Learning Approach to Graph Edit Costs for Regression. Proceedings of IAPR Joint International Workshops on Statistical techniques in Pattern Recognition (SPR 2020) and Structural and Syntactic Pattern Recognition (SSPR 2020). 2021. [preprint] [video] [slides] [doi]

[P1]

Qu Hua, Zhao Jihong, Wu Jinkang, Linlin Jia, etc., A Kind of Name Data Network Mobility Switching Method Predicted Using ELM: China, CN106376041B[P]. [patent]

Libraries

For more detailed project descriptions, please visit my profile on GitHub.

graphkit-learn

A Python package on graph kernels, graph edit distances and graph pre-image problem. [homepage]

Academic Activities

2021.01: Attended S+SSPR Workshops 2020 and gave two video presentations. [video1] [video2]

2020.09: Attended 2020 online CSC-Seminar in DE. and Fr. and presented a work: A graph pre-image method based on graph edit distances. [slides] [award]

2018.08: Attended Machine Learning Summer School 2018 Madrid and presented a poster. [poster]

Awards & Fundings

2017.10 - 2021.04: Ph.D. grants from China Scholarship Council (CSC China/UT-INSA Program).

Volunteer Work

2019.12 - 2023: I was a translator and proofreader volunteer on the Khan Academy simplified Chinese translation team. I have translated / proofread more than 100,000 words. [introduction]