From 8fd2e00738ed1fd0db0cab0f54cdd53b5945f566 Mon Sep 17 00:00:00 2001 From: Junxiang Wang Date: Mon, 1 Apr 2024 12:03:50 -0400 Subject: [PATCH] Update paper.md --- paper.md | 25 ++++++++++++------------- 1 file changed, 12 insertions(+), 13 deletions(-) diff --git a/paper.md b/paper.md index 66588c6..366d694 100644 --- a/paper.md +++ b/paper.md @@ -27,7 +27,7 @@ bibliography: paper.bib # Summary -We present GraphSL, an extensible library designed for investigating the graph source localization problem. Our library facilitates the exploration of various graph diffusion models for simulating information spread and enables the evaluation of cutting-edge source localization approaches on established benchmark datasets. Ongoing development efforts are underway, and we enthusiastically invite contributions from both academic and industrial sectors to further enhance GraphSL's capabilities. +We present GraphSL, a novel library designed for investigating the graph source localization problem. Our library facilitates the exploration of various graph diffusion models for simulating information spread and enables the evaluation of cutting-edge source localization approaches on established benchmark datasets. The source code of GraphSL is made available at (https://github.com/xianggebenben/GraphSL). Bug reports and feedback can be directed to the Github issues page (https://github.com/xianggebenben/GraphSL/issues). # Statement of need @@ -53,7 +53,7 @@ graph source localization, is to infer $x$ from $Y_{T}$: ![The hierarchical structure of our GraphSL library version 0.1.\label{fig:overview}](overview.png) -The structure of our GraphSL library is depicted in \autoref{fig:overview]. Existing graph source localization methods can be categorized into two groups: Prescribed methods and Graph Neural Networks (GNN)-based methods. +The structure of our GraphSL library is depicted in \autoref{fig:overview}. Existing graph source localization methods can be categorized into two groups: Prescribed methods and Graph Neural Networks (GNN)-based methods. Prescribed methods rely on hand-crafted rules and heuristics. For instance, LPSI propagated infection in networks and labels local peaks as source nodes [@wang2017multiple]. NetSleuth employed the Minimum Description Length principle to identify the optimal set of source nodes and virus propagation ripple [@prakash2012spotting]. OJC identified a set of nodes (Jordan cover) that cover all observed infected nodes with the minimum radius [@zhu2017catch]. @@ -72,28 +72,27 @@ GNN-based methods learn rules from graph data in an end-to-end manner by capturi Table: \label{tab:statistics} The statistics of eight datasets. - Aside from methods, we also provide eight benchmark datasets to facilitate the graph SL research, whose statistics are shown in \autoref{tab:statistics}. - Memetracker and Digg provide Source-Diffusion pairs $(x,Y_{T})$, while others do not. All datasets are introduced as follows: +Aside from methods, we also provide eight benchmark datasets to facilitate the graph SL research, whose statistics are shown in \autoref{tab:statistics}. Memetracker and Digg provide Seed-Diffusion vector pairs $(x,Y_{T})$. While others do not have such pairs, they can be generated by information diffusion simulations. All datasets are introduced as follows: - 1. Karate [@lusseau2003bottlenose]. Karate contains the social ties among the members of a university karate club. +1. Karate [@lusseau2003bottlenose]: Karate depicts the social ties among members of a university karate club. - 2. Dolphins [@lusseau2003bottlenose]. Dolphins is a social network of bottlenose dolphins, where edges represent frequent associations between dolphins. +2. Dolphins [@lusseau2003bottlenose]: Dolphins represents a social network of bottlenose dolphins, with edges indicating frequent associations between dolphins. - 3. Jazz [@gleiser2003community]. Jazz is a collaboration network between Jazz musicians. Each edge represents that two musicians have played together in a band. +3. Jazz [@gleiser2003community]: Jazz illustrates a collaboration network among Jazz musicians, where edges signify instances of playing together in a band. - 4. Network Science [@newman2006finding]. Network Science is a coauthorship network of scientists working on network theory and experiment. Each edge represents two scientists who have co-authored a paper. +4. Network Science [@newman2006finding]: Network Science portrays a coauthorship network of scientists engaged in network theory and experimentation, with each edge representing co-authorship of a paper. - 5. Cora-ML [@mccallum2000automating]. Cora-ML is a portal network of computer science research papers crawled by machine learning techniques. +5. Cora-ML [@mccallum2000automating]: Cora-ML is a portal network of computer science research papers obtained through machine learning techniques. - 6. Power Grid [@watts1998collective]. Power Grid is a topology network of the Western States Power Grid of the United States. +6. Power Grid [@watts1998collective]: Power Grid delineates the topology network of the Western States Power Grid in the United States. - 7. Memetracker [@leskovec2009meme]. The Memetracker keeps track of frequently used phrases on news social media. Only a small subset of the Memetracker network is used here. +7. Memetracker [@leskovec2009meme]: Memetracker tracks frequently used phrases on news social media, and a small subset of the whole Memetracker network was retrieved here. - 8. Digg [@hogg2012social]. Digg is a reply network of the social news. Only a small subset of the Memetracker network is used here. +8. Digg [@hogg2012social]: Digg showcases a reply network within social news, and a small subset of the whole Digg network was retrieved here. # Availability and Documentation -GraphSL is available under the MIT License. The library may be cloned from the GitHub repository or via PyPI: pip install GraphSL. Documentation is provided via Read the Docs, including a quickstart introducing major functionality and a detailed API reference. Extensive unit testing is employed throughout the library. The source code of GraphSL is made available at (https://github.com/xianggebenben/GraphSL). Bug reports and feedback can be directed to the Github issues page (https://github.com/xianggebenben/GraphSL/issues). +GraphSL is available under the MIT License. The library may be cloned from the GitHub repository or via PyPI: pip install GraphSL. Documentation is provided via Read the Docs, including a quickstart introducing major functionality and a detailed API reference. Extensive unit testing is employed throughout the library. # Acknowledgements