Vpalg: Paper-publication Prediction with Graph Neural Networks Renchu Guan ∗ College of Computer Science and Technology, Jilin University China guanrenchu@jlu.edu.cn Yonghao Liu ∗ College of Computer Science and Technology, Jilin University China yonghao20@mails.jlu.edu.cn Xiaoyue Feng ∗† College of Computer Science and Technology, Jilin University China fengxy@jlu.edu.cn Ximing Li ∗† College of Computer Science and Technology, Jilin University China liximing86@gmail.com ABSTRACT Paper-publication venue prediction aims to predict candidate publi- cation venues that efectively suit given submissions. This technol- ogy is developing rapidly with the popularity of machine learning models. However, most previous methods ignore the structure infor- mation of papers, while modeling them with graphs can naturally solve this drawback. Meanwhile, they either use hand-crafted or bag-of-word features to represent the papers, ignoring the ones that involve high-level semantics. Moreover, existing methods assume that the venue where a paper is published as a correct venue for the data annotation, which is unrealistic. One paper can be relevant to many venues. In this paper, we attempt to address these prob- lems above and develop a novel prediction model, namely Venue Prediction with Abstract-Level Graph (Vpalg), which can serve as an efective decision-making tool for venue selections. Specifcally, to achieve more discriminative paper abstract representations, we construct each abstract as a semantic graph and perform a dual at- tention message passing neural network for representation learning. Then, the proposed model can be trained over the learned abstract representations with their labels and generalized via self-training. Empirically, we employ the PubMed dataset and further collect two new datasets from the top journals and conferences in computer science. Experimental results indicate the superior performance of Vpalg, consistently outperforming the existing baseline methods. ∗ Key Laboratory of Symbolic Computation and Knowledge Engineering of Ministry of Education, China † Corresponding author Permission to make digital or hard copies of all or part of this work for personal or classroom use is granted without fee provided that copies are not made or distributed for proft or commercial advantage and that copies bear this notice and the full citation on the frst page. Copyrights for components of this work owned by others than ACM must be honored. Abstracting with credit is permitted. To copy otherwise, or republish, to post on servers or to redistribute to lists, requires prior specifc permission and/or a fee. Request permissions from permissions@acm.org. CIKM ’21, November 1–5, 2021, Virtual Event, QLD, Australia © 2021 Association for Computing Machinery. ACM ISBN 978-1-4503-8446-9/21/11. . . $15.00 https://doi.org/10.1145/3459637.3482490 CCS CONCEPTS • Computing methodologies → Machine learning algorithms; • Information systems → Data mining. KEYWORDS graph neural networks, text mining, paper-publication prediction ACM Reference Format: Renchu Guan, Yonghao Liu, Xiaoyue Feng, and Ximing Li. 2021. Vpalg: Paper-publication Prediction with Graph Neural Networks. In Proceedings of the 30th ACM International Conference on Information and Knowledge Management (CIKM ’21), November 1–5, 2021, Virtual Event, QLD, Australia. ACM, New York, NY, USA, 10 pages. https://doi.org/10.1145/3459637.3482490 1 INTRODUCTION Paper-publication venue prediction refers to the automatic process that predicts appropriate candidate publication venues for given paper submissions. This task is of great practical signifcance for researchers. First, it’s time-consuming to choose a suitable venue to publish papers due to a rapid surge in the number of scholarly venues concomitant with exponential growth. An inappropriate choice of venue may result in delayed publication of the paper [4]. Second, some researchers often limit themselves to known venues where they already own publications, which reduces the option of other appropriate venues [40]. Third, it’s very overwhelming for junior researchers to publish research work in an appropriate venue, especially in the era of interdisciplinarity [22]. The outcome of venue prediction can guide them to make reasonable decisions for the desired venues. And it can also help editors determine whether a paper is appropriate for the venue. In addition, for publishers (such as Springer and Elsevier), a better publication venue prediction system can promote scholars to publish work under their venues. A typical approach is to cast venue prediction problem into a classifcation or regression task [2], focusing on extracting use- ful feature information from papers to improve prediction per- formance. They leverage a series of published papers and regard their corresponding publications as labels to build a predictor via using supervised machine learning models. More recently, a line of deep learning-based models that can automatically learn word representations from papers have been designed to solve this task Full Paper Track CIKM ’21, November 1–5, 2021, Virtual Event, Australia 617