Lecture Notes in Artificial Intelligence 13031

Subseries of Lecture Notes in Computer Science

Series Editors

Randy GoebelUniversity of Alberta, Edmonton, Canada

Yuzuru TanakaHokkaido University, Sapporo, Japan

Wolfgang WahlsterDFKI and Saarland University, Saarbrücken, Germany

Founding Editor

Jörg SiekmannDFKI and Saarland University, Saarbrücken, Germany

More information about this subseries at http://www.springer.com/series/1244

Duc Nghia Pham • Thanaruk Theeramunkong •

Guido Governatori • Fenrong Liu (Eds.)

PRICAI 2021:Trends inArtificial Intelligence18th Pacific RimInternational Conference on Artificial Intelligence, PRICAI 2021Hanoi, Vietnam, November 8–12, 2021Proceedings, Part I

123

EditorsDuc Nghia PhamMIMOS BerhadKuala Lumpur, Malaysia

Thanaruk TheeramunkongSirindhorn International Institute of Scienceand TechnologyThammasat UniversityMueang Pathum Thani, ThailandGuido Governatori

Data61CSIROBrisbane, QLD, Australia

Fenrong LiuDepartment of PhilosophyTsinghua UniversityBeijing, China

ISSN 0302-9743 ISSN 1611-3349 (electronic)Lecture Notes in Artificial IntelligenceISBN 978-3-030-89187-9 ISBN 978-3-030-89188-6 (eBook)https://doi.org/10.1007/978-3-030-89188-6

LNCS Sublibrary: SL7 – Artificial Intelligence

© Springer Nature Switzerland AG 2021, corrected publication 2022This work is subject to copyright. All rights are reserved by the Publisher, whether the whole or part of thematerial is concerned, specifically the rights of translation, reprinting, reuse of illustrations, recitation,broadcasting, reproduction on microfilms or in any other physical way, and transmission or informationstorage and retrieval, electronic adaptation, computer software, or by similar or dissimilar methodology nowknown or hereafter developed.The use of general descriptive names, registered names, trademarks, service marks, etc. in this publicationdoes not imply, even in the absence of a specific statement, that such names are exempt from the relevantprotective laws and regulations and therefore free for general use.The publisher, the authors and the editors are safe to assume that the advice and information in this book arebelieved to be true and accurate at the date of publication. Neither the publisher nor the authors or the editorsgive a warranty, expressed or implied, with respect to the material contained herein or for any errors oromissions that may have been made. The publisher remains neutral with regard to jurisdictional claims inpublished maps and institutional affiliations.

This Springer imprint is published by the registered company Springer Nature Switzerland AGThe registered company address is: Gewerbestrasse 11, 6330 Cham, Switzerland

Preface

These three-volume proceedings contain the papers presented at the 18th Pacific RimInternational Conference on Artificial Intelligence (PRICAI 2021) held virtually duringNovember 8–12, 2021, in Hanoi, Vietnam.

PRICAI, which was inaugurated in Tokyo in 1990, started out as a biennial inter-national conference concentrating on artificial intelligence (AI) theories, technologies,and applications in the areas of social and economic importance for Pacific Rimcountries. It provides a common forum for researchers and practitioners in variousbranches of AI to exchange new ideas and share experience and expertise. Since then,the conference has grown, both in participation and scope, to be a premier internationalAI event for all major Pacific Rim nations as well as countries from all around theworld. In 2018, the PRICAI Steering Committee decided to hold PRICAI on an annualbasis starting from 2019.

This year, we received an overwhelming number of 382 submissions to both theMain track (365 submissions) and the Industry special track (17 submissions). Thisnumber was impressive considering that for the first time PRICAI was being heldvirtually during a global pandemic situation. All submissions were reviewed andevaluated with the same highest quality standard through a double-blind review pro-cess. Each paper received at least two reviews, in most cases three, and in some casesup to four. During the review process, discussions among the Program Committee(PC) members in charge were carried out before recommendations were made, andwhen necessary, additional reviews were sourced. Finally, the conference and programco-chairs read the reviews and comments and made a final calibration for differencesamong individual reviewer scores in light of the overall decisions. The entire ProgramCommittee (including PC members, external reviewers, and co-chairs) expendedtremendous effort to ensure fairness and consistency in the paper selection process.Eventually, we accepted 92 regular papers and 28 short papers for oral presentation.This gives a regular paper acceptance rate of 24.08% and an overall acceptance rate of31.41%.

The technical program consisted of three tutorials and the main conference program.The three tutorials covered hot topics in AI from “Collaborative Learning and Opti-mization” and “Mechanism Design Powered by Social Interactions” to “TowardsHyperdemocary: AI-enabled Crowd Consensus Making and Its Real-World SocietalExperiments”. All regular and short papers were orally presented over four days inparallel and in topical program sessions. We were honored to have keynote presen-tations by four distinguished researchers in the field of AI whose contributions havecrossed discipline boundaries: Mohammad Bennamoun (University of WesternAustralia, Australia), Johan van Benthem (University of Amsterdam, The Netherlands;Stanford University, USA; and Tsinghua University, China), Virginia Dignum (UmeåUniversity, Sweden), and Yutaka Matsuo (University of Tokyo, Japan). We weregrateful to them for sharing their insights on their latest research with us.

The success of PRICAI 2021 would not be possible without the effort and support ofnumerous people from all over the world. First, we would like to thank the authors, PCmembers, and external reviewers for their time and efforts spent in making PRICAI2021 a successful and enjoyable conference. We are also thankful to various fellowmembers of the conference committee, without whose support and hard work PRICAI2021 could not have been successful:

– Advisory Board: Hideyuki Nakashima, Abdul Sattar, and Dickson Lukose– Industry Chair: Shiyou Qian– Local/Virtual Organizing Chairs: Sankalp Khanna and Adila Alfa Krisnadhi– Tutorial Chair: Guandong Xu– Web and Publicity Chair: Md Khaled Ben Islam– Workshop Chair: Dengji Zhao

We gratefully acknowledge the organizational support of several institutionsincluding Data61/CSIRO (Australia), Tsinghua University (China), MIMOS Berhad(Malaysia), Thammasat University (Thailand), and Griffith University (Australia).

Finally, we thank Springer, Ronan Nugent (Editorial Director, Computer ScienceProceedings), and Anna Kramer (Assistant Editor, Computer Science Proceedings) fortheir assistance in publishing the PRICAI 2021 proceedings as three volumes of itsLecture Notes in Artificial Intelligence series.

November 2021 Duc Nghia PhamThanaruk Theeramunkong

Guido GovernatoriFenrong Liu

vi Preface

Organization

PRICAI Steering Committee

Steering Committee

Quan Bai University of Tasmania, AustraliaTru Hoang Cao The University of Texas Health Science Center

at Houston, USAXin Geng Southeast University, ChinaGuido Governatori Data61, CSIRO, AustraliaTakayuki Ito Nagoya Institute of Technology, JapanByeong-Ho Kang University of Tasmania, AustraliaM. G. M. Khan University of the South Pacific, FijiSankalp Khanna Australian e-Health Research Centre, CSIRO, AustraliaDickson Lukose Monash University, AustraliaHideyuki Nakashima Sapporo City University, JapanAbhaya Nayak Macquarie University, AustraliaSeong Bae Park Kyung Hee University, South KoreaDuc Nghia Pham MIMOS Berhad, MalaysiaAbdul Sattar Griffith University, AustraliaAlok Sharma RIKEN, Japan, and University of the South Pacific, FijiThanaruk Theeramunkong Thammasat University, ThailandZhi-Hua Zhou Nanjing University, China

Honorary Members

Randy Goebel University of Alberta, CanadaTu-Bao Ho Japan Advanced Institute of Science and Technology,

JapanMitsuru Ishizuka University of Tokyo, JapanHiroshi Motoda Osaka University, JapanGeoff Webb Monash University, AustraliaAlbert Yeap Auckland University of Technology, New ZealandByoung-Tak Zhang Seoul National University, South KoreaChengqi Zhang University of Technology Sydney, Australia

Conference Organizing Committee

General Chairs

Guido Governatori Data61, CSIRO, AustraliaFenrong Liu Tsinghua University, China

Program Chairs

Duc Nghia Pham MIMOS Berhad, MalaysiaThanaruk Theeramunkong Thammasat University, Thailand

Local/Virtual Organizing Chairs

Sankalp Khanna Australian e-Health Research Centre, CSIRO, AustraliaAdila Alfa Krisnadhi University of Indonesia, Indonesia

Workshop Chair

Dengji Zhao ShanghaiTech University, China

Tutorial Chair

Guandong Xu University of Technology Sydney, Australia

Industry Chair

Shiyou Qian Shanghai Jiao Tong University, China

Web and Publicity Chair

Md Khaled Ben Islam Griffith University, Australia

Advisory Board

Hideyuki Nakashima Sapporo City University, JapanAbdul Sattar Griffith University, AustraliaDickson Lukose Monash University, Australia

Program Committee

Eriko Aiba The University of Electro-Communications, JapanPatricia Anthony Lincoln University, New ZealandChutiporn Anutariya Asian Institute of Technology, ThailandMohammad Arshi Saloot MIMOS Berhad, MalaysiaYun Bai University of Western Sydney, AustraliaChutima Beokhaimook Rangsit University, ThailandAteet Bhalla Independent Technology Consultant, IndiaChih How Bong Universiti Malaysia Sarawak, MalaysiaPoonpong Boonbrahm Walailak University, ThailandAida Brankovic Australian e-Health Research Centre, CSIRO, AustraliaXiongcai Cai University of New South Wales, AustraliaTru Cao University of Texas Health Science Center at Houston,

USAHutchatai Chanlekha Kasetsart University, ThailandSapa Chanyachatchawan National Electronics and Computer Technology Center,

ThailandSiqi Chen Tianjin University, China

viii Organization

Songcan Chen Nanjing University of Aeronautics and Astronautics,China

Wu Chen Southwest University, ChinaYingke Chen Sichuan University, ChinaWai Khuen Cheng Universiti Tunku Abdul Rahman, MalaysiaBoonthida Chiraratanasopha Yala Rajabhat University, ThailandPhatthanaphong

ChomphuwisetMahasarakham University, Thailand

Dan Corbett Optimodal Technologies, USACélia Da Costa Pereira Université Côte d’Azur, FranceJirapun Daengdej Assumption University, ThailandHoa Khanh Dam University of Wollongong, AustraliaXuan-Hong Dang IBM Watson Research, USAAbdollah Dehzangi Rutgers University, USASang Dinh Hanoi University of Science and Technology, VietnamClare Dixon University of Manchester, UKShyamala Doraisamy University Putra Malaysia, MalaysiaNuttanart Facundes King Mongkut’s University of Technology Thonburi,

ThailandEduardo Fermé Universidade da Madeira, PortugalSomchart Fugkeaw Thammasat University, ThailandKatsuhide Fujita Tokyo University of Agriculture and Technology,

JapanNaoki Fukuta Shizuoka University, JapanMarcus Gallagher University of Queensland, AustraliaDragan Gamberger Rudjer Boskovic Institute, CroatiaWei Gao Nanjing University, ChinaXiaoying Gao Victoria University of Wellington, New ZealandXin Geng Southeast University, ChinaManolis Gergatsoulis Ionian University, GreeceGuido Governatori Data61, CSIRO, AustraliaAlban Grastien Australian National University, AustraliaCharles Gretton Australian National University, AustraliaFikret Gurgen Bogazici University, TurkeyPeter Haddawy Mahidol University, ThailandChoochart Haruechaiyasak National Electronics and Computer Technology Center,

ThailandHamed Hassanzadeh Australian e-Health Research Centre, CSIRO, AustraliaTessai Hayama Nagaoka University of Technology, JapanJuhua Hu University of Washington, USAXiaodi Huang Charles Sturt University, AustraliaVan Nam Huynh Japan Advanced Institute of Science and Technology,

JapanNorisma Idris University of Malaya, MalaysiaMitsuru Ikeda Japan Advanced Institute of Science and Technology,

Japan

Organization ix

Masashi Inoue Tohoku Institute of Technology, JapanTakayuki Ito Kyoto University, JapanSanjay Jain National University of Singapore, SingaporeGuifei Jiang Nankai University, ChinaYichuan Jiang Southeast University, ChinaNattagit Jiteurtragool Digital Government Development Agency, ThailandHideaki Kanai Japan Advanced Institute of Science and Technology,

JapanRyo Kanamori Nagoya University, JapanNatsuda Kaothanthong Thammasat University, ThailandJessada Karnjana National Electronics and Computer Technology Center,

ThailandC. Maria Keet University of Cape Town, South AfricaGabriele Kern-Isberner Technische Universitaet Dortmund, GermanySankalp Khanna Australian e-Health Research Centre, CSIRO, AustraliaNichnan

KittiphattanabawonWalailak University, Thailand

Frank Klawonn Ostfalia University, GermanySébastien Konieczny CRIL-CNRS, FranceKrit Kosawat National Electronics and Computer Technology Center,

ThailandAlfred Krzywicki University of New South Wales, AustraliaKun Kuang Zhejiang University, ChinaYoung-Bin Kwon Chung-Ang University, South KoreaWeng Kin Lai Tunku Abdul Rahman University College, MalaysiaHo-Pun Lam Data61, CSIRO, AustraliaNasith Laosen Phuket Rajabhat University, ThailandVincent CS Lee Monash University, AustraliaRoberto Legaspi KDDI Research Inc., JapanGang Li Deakin University, AustraliaGuangliang Li Ocean University of China, ChinaTianrui Li Southwest Jiaotong University, ChinaChanjuan Liu Dalian University of Technology, ChinaFenrong Liu Tsinghua University, ChinaMichael Maher Reasoning Research Institute, AustraliaXinjun Mao National University of Defense Technology, ChinaEric Martin University of New South Wales, AustraliaMaria Vanina Martinez Instituto de Ciencias de la Computación, ArgentinaSanparith Marukatat National Electronics and Computer Technology Center,

ThailandMichael Mayo University of Waikato, New ZealandBrendan McCane University of Otago, New ZealandRiichiro Mizoguchi Japan Advanced Institute of Science and Technology,

JapanNor Liyana Mohd Shuib University of Malaya, MalaysiaM. A. Hakim Newton Griffith University, Australia

x Organization

Hung Duy Nguyen Thammasat University, ThailandPhi Le Nguyen Hanoi University of Science and Technology, VietnamKouzou Ohara Aoyama Gakuin University, JapanFrancesco Olivieri Griffith University, AustraliaMehmet Orgun Macquarie University, AustraliaNoriko Otani Tokyo City University, JapanMaurice Pagnucco University of New South Wales, AustraliaLaurent Perrussel IRIT - Universite de Toulouse, FranceBernhard Pfahringer University of Waikato, New ZealandDuc Nghia Pham MIMOS Berhad, MalaysiaJantima Polpinij Mahasarakham University, ThailandThadpong

PongthawornkamolKasikorn Business-Technology Group, Thailand

Yuhua Qian Shanxi University, ChinaJoel Quinqueton LIRMM, FranceTeeradaj Racharak Japan Advanced Institute of Science and Technology,

JapanFenghui Ren University of Wollongong, AustraliaMark Reynolds University of Western Australia, AustraliaJandson S. Ribeiro University of Koblenz-Landau, GermanyKazumi Saito University of Shizuoka, JapanChiaki Sakama Wakayama University, JapanKen Satoh National Institute of Informatics and Sokendai, JapanAbdul Sattar Griffith University, AustraliaNicolas Schwind National Institute of Advanced Industrial Science

and Technology, JapanNazha Selmaoui-Folcher University of New Caledonia, FranceLin Shang Nanjing University, ChinaAlok Sharma RIKEN, JapanChenwei Shi Tsinghua University, ChinaZhenwei Shi Beihang University, ChinaMikifumi Shikida Kochi University of Technology, JapanSoo-Yong Shin Sungkyunkwan University, South KoreaYanfeng Shu CSIRO, AustraliaTony Smith University of Waikato, New ZealandChattrakul Sombattheera Mahasarakham University, ThailandInsu Song James Cook University, AustraliaSafeeullah Soomro Virginia State University, USATasanawan Soonklang Silpakorn University, ThailandMarkus Stumptner University of South Australia, AustraliaMerlin Teodosia Suarez De La Salle University, PhilippinesXin Sun Catholic University of Lublin, PolandBoontawee Suntisrivaraporn DTAC, ThailandThepchai Supnithi National Electronics and Computer Technology Center,

ThailandDavid Taniar Monash University, Australia

Organization xi

Thanaruk Theeramunkong Thammasat University, ThailandMichael Thielscher University of New South Wales, AustraliaSatoshi Tojo Japan Advanced Institute of Science and Technology,

JapanShikui Tu Shanghai Jiao Tong University, ChinaMiroslav Velev Aries Design Automation, USAMuriel Visani Hanoi University of Science and Technology, Vietnam

and La Rochelle University, FranceToby Walsh University of New South Wales, AustraliaXiao Wang Beijing University of Posts and Telecommunications,

ChinaPaul Weng Shanghai Jiao Tong University, ChinaPeter Whigham University of Otago, New ZealandWayne Wobcke University of New South Wales, AustraliaSartra Wongthanavasu Khon Kaen University, ThailandBrendon J. Woodford University of Otago, New ZealandKaibo Xie University of Amsterdam, The NetherlandsMing Xu Xi’an Jiaotong-Liverpool University, ChinaShuxiang Xu University of Tasmania, AustraliaHui Xue Southeast University, ChinaMing Yang Nanjing Normal University, ChinaRoland Yap National University of Singapore, SingaporeKenichi Yoshida University of Tsukuba, JapanTakaya Yuizono Japan Advanced Institute of Science and Technology,

JapanChengqi Zhang University of Technology Sydney, AustraliaDu Zhang California State University, USAMin-Ling Zhang Southeast University, ChinaShichao Zhang Central South University, ChinaWen Zhang Beijing University of Technology, ChinaYu Zhang Southern University of Science and Technology, ChinaZhao Zhang Hefei University of Technology, ChinaZili Zhang Deakin University, AustraliaYanchang Zhao Data61, CSIRO, AustraliaShuigeng Zhou Fudan University, ChinaXingquan Zhu Florida Atlantic University, USA

Additional Reviewers

Aitchison, MatthewAkhtar, NaveedAlgar, ShannonAlmeida, YuriBoudou, JosephBurie, Jean-ChristopheChandra, Abel

Cheng, CharibethDamigos, MatthewDong, HuanfangDu Preez-Wilkinson, NathanielEffendy, SuhendryEng, Bah TeeFeng, Xuening

xii Organization

Fu, KerenGao, YiGeng, ChuanxingHabault, GuillaumeHang, Jun-YiHe, ZhengqiHoang, AnhHuynh, DuInventado, Paul SalvadorJan, ZohaibJannai, TokotokoJia, BinbinJiang, ZhaohuiKalogeros, EleftheriosKarim, AbdulKumar, ShiuLai, YongLaosen, KanjanaLee, Nung KionLee, ZhiyiLi, WeikaiLiang, YanyanLiu, JiexiLiu, XiaxueLiu, YanliLuke, Jing YuanMahdi, GhulamMayer, WolfgangMendonça, FábioMing, ZuhengMittelmann, MunyqueNguyen, Duy HungNguyen, Hong-HuyNguyen, Mau ToanNguyen, Minh HieuNguyen, Minh LeNguyen, Trung ThanhNikafshan Rad, HimaOkubo, YoshiakiOng, EthelOstertag, Cécilia

Phiboonbanakit, ThananutPhua, Yin JunPongpinigpinyo, SuneePreto, SandroQian, JunqiQiao, YukaiRiahi, VahidRodrigues, PedroRosenberg, ManouSa-Ngamuang, ChaitawatScherrer, RomaneSelway, MattSharma, RoneshSong, GeSu Yin, MyatSubash, AdityaTan, HongweiTang, JiahuaTeh, Chee SiongTettamanzi, AndreaTian, QingTran, VuVo, Duc VinhWang, Deng-BaoWang, KaixiangWang, ShuwenWang, YuchenWang, YunyunWilhelm, MarcoWu, LinzeXiangru, YuXing, GuanyuXue, HaoYan, WenzhuYang, WanqiYang, YikunYi, HuangYin, ZeYu, GuanbaoZhang, JianyiZhang, Jiaqiang

Organization xiii

Contents – Part I

AI Foundations/Decision Theory

Designing Bounded Min-Knapsack Bandits Algorithm for SustainableDemand Response. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3

Akansha Singh, P. Meghana Reddy, Shweta Jain, and Sujit Gujar

Designing Refund Bonus Schemes for Provision Point Mechanismin Civic Crowdfunding . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18

Sankarshan Damle, Moin Hussain Moti, Praphul Chandra,and Sujit Gujar

Federated Learning for Non-IID Data: From Theory to Algorithm. . . . . . . . . 33Bojian Wei, Jian Li, Yong Liu, and Weiping Wang

Fixed-Price Diffusion Mechanism Design. . . . . . . . . . . . . . . . . . . . . . . . . . 49Tianyi Zhang, Dengji Zhao, Wen Zhang, and Xuming He

Multiclass Classification Using Dilute Bandit Feedback . . . . . . . . . . . . . . . . 63Gaurav Batra and Naresh Manwani

A Study of Misinformation Games . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 76Constantinos Varsos, Giorgos Flouris, Marina Bitsaki,and Michail Fasoulakis

Influence-Driven Explanations for Bayesian Network Classifiers . . . . . . . . . . 88Emanuele Albini, Antonio Rago, Pietro Baroni, and Francesca Toni

Public Project with Minimum Expected Release Delay . . . . . . . . . . . . . . . . 101Guanhua Wang and Mingyu Guo

Strategy Proof Mechanisms for Facility Location at Limited Locations . . . . . 113Toby Walsh

Applications of AI

A Consistency Enhanced Deep Lmser Network for Face Sketch Synthesis . . . 127Qingjie Sheng, Shikui Tu, and Lei Xu

A Cost-Efficient Framework for Scene Text Detection in the Wild . . . . . . . . 139Gangyan Zeng, Yuan Zhang, Yu Zhou, and Xiaomeng Yang

A Dueling-DDPG Architecture for Mobile Robots Path Planning Basedon Laser Range Findings . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 154

Panpan Zhao, Jinfang Zheng, Qinglin Zhou, Chen Lyu, and Lei Lyu

A Fully Dynamic Context Guided Reasoning and Reconsidering Networkfor Video Captioning. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 169

Xia Feng, Xinyu He, Rui Huang, and Caihua Liu

Adaptive Prediction of Hip Joint Center from X-ray Images UsingGeneralized Regularized Extreme Learning Machine and GlobalizedBounded Nelder-Mead Strategy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 183

Fuchang Han, Shenghui Liao, Yiyong Jiang, Shu Liu, Yuqian Zhao,and Xiantao Shen

Adversarial Training for Image Captioning Incorporating RelationAttention . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 197

Tianyu Chen, Zhixin Li, Canlong Zhang, and Huifang Ma

Element Re-identification in Crowdtesting . . . . . . . . . . . . . . . . . . . . . . . . . 212Li Zhang and Wei-Tek Tsai

Flame and Smoke Detection Algorithm for UAV Basedon Improved YOLOv4-Tiny . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 226

Ruinan Wu, Changchun Hua, Weili Ding, Yifan Wang, and Yubao Wang

Improving Protein Backbone Angle Prediction Using Hidden MarkovModels in Deep Learning. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 239

Fereshteh Mataeimoghadam, M. A. Hakim Newton, Rianon Zaman,and Abdul Sattar

Magic Mirror Stealth: Interactive Automatic Picture Editing System . . . . . . . 252MengDi Zhou, BoHeng Hu, and Si Liu

Off-TANet: A Lightweight Neural Micro-expression Recognizer withOptical Flow Features and Integrated Attention Mechanism . . . . . . . . . . . . . 266

Jiahao Zhang, Feng Liu, and Aimin Zhou

Pulmonary Nodule Classification of CT Images with Attribute Self-guidedGraph Convolutional V-Shape Networks . . . . . . . . . . . . . . . . . . . . . . . . . . 280

Xiangbo Zhang, Kun Wang, Xiaohong Zhang, and Sheng Huang

Semantic Structural and Occlusive Feature Fusionfor Pedestrian Detection . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 293

Hui Wang, Yu Zhang, Hongchang Ke, Ning Wei, and Zhongyu Xu

VTLayout: Fusion of Visual and Text Features for DocumentLayout Analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 308

Shoubin Li, Xuyan Ma, Shuaiqun Pan, Jun Hu, Lin Shi, and Qing Wang

xvi Contents – Part I

An Initial Study of Machine Learning Underspecification Using FeatureAttribution Explainable AI Algorithms: A COVID-19 Virus TransmissionCase Study. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 323

James Hinns, Xiuyi Fan, Siyuan Liu, Veera Raghava Reddy Kovvuri,Mehmet Orcun Yalcin, and Markus Roggenbach

Generation of Environment-Irrelevant Adversarial DigitalCamouflage Patterns . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 336

Xu Teng, Hui Zhang, Bo Li, Chunming Yang, and Xujian Zhao

Magnitude-Weighted Mean-Shift Clustering with Leave-One-OutBandwidth Estimation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 347

Yuki Yamagishi, Kazumi Saito, Kazuro Hirahara, and Naonori Ueda

Pervasive Monitoring of Gastrointestinal Health of Newborn Babies . . . . . . . 359Insu Song, Yi Huang, Tieh Hee Hai Guan Koh,and Victor Samuel Rajadurai

Price and Time Optimization for Utility-Aware Taxi Dispatching . . . . . . . . . 370Yuya Hikima, Masahiro Kohjima, Yasunori Akagi, Takeshi Kurashima,and Hiroyuki Toda

VAN: Voting and Attention Based Network for Unsupervised MedicalImage Registration . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 382

Zhiang Zu, Guixu Zhang, Yaxin Peng, Zhen Ye, and Chaomin Shen

Data Mining and Knowledge Discovery

MGEoT: A Multi-grained Ensemble Method for TimeSeries Classification . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 397

Ziyi Wang, Yujie Zhou, Chun Li, Lin Shang, and Bing Xue

Mining Skyline Frequent-Utility Itemsets with Utility Filtering . . . . . . . . . . . 411Wei Song, Chuanlong Zheng, and Philippe Fournier-Viger

Network Embedding with Topology-Aware Textual Representations . . . . . . . 425Jiaxing Chen, Zenan Xu, and Qinliang Su

Online Discriminative Semantic-Preserving Hashing for Large-ScaleCross-Modal Retrieval . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 440

Jinhan Yi, Yi He, and Xin Liu

Empirical Study on the Impact of Different Sets of Parameters of GradientBoosting Algorithms for Time-Series Forecasting with LightGBM . . . . . . . . 454

Filipa S. Barros, Vitor Cerqueira, and Carlos Soares

Contents – Part I xvii

Evolutionary Computation/Optimisation

A Two-Stage Efficient Evolutionary Neural Architecture Search Methodfor Image Classification . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 469

Gonglin Yuan, Bing Xue, and Mengjie Zhang

Adaptive Relaxations for Multistage Robust Optimization . . . . . . . . . . . . . . 485Michael Hartisch

ALGNN: Auto-Designed Lightweight Graph Neural Network. . . . . . . . . . . . 500Rongshen Cai, Qian Tao, Yufei Tang, and Min Shi

Automatic Graph Learning with Evolutionary Algorithms:An Experimental Study . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 513

Chenyang Bu, Yi Lu, and Fei Liu

Dendritic Cell Algorithm with Group Particle Swarm Optimizationfor Input Signal Generation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 527

Dan Zhang and Yiwen Liang

Knowledge Representation and Reasoning

Building Trust for Belief Revision. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 543Aaron Hunter

Correcting Large Knowledge Bases Using Guided Inductive LogicLearning Rules . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 556

Yan Wu, Zili Zhang, and Guodong Wang

High-Quality Noise Detection for Knowledge Graph Embeddingwith Rule-Based Triple Confidence . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 572

Yan Hong, Chenyang Bu, and Xindong Wu

Multi-agent Epistemic Planning with Inconsistent Beliefs, Trust and Lies . . . . 586Francesco Fabiano, Alessandro Burigana, Agostino Dovier,Enrico Pontelli, and Tran Cao Son

Correction to: Federated Learning for Non-IID Data: From Theoryto Algorithm. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . C1

Bojian Wei, Jian Li, Yong Liu, and Weiping Wang

Author Index . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 599

xviii Contents – Part I

Contents – Part II

Natural Language Processing

A Calibration Method for Sentiment Time Series by Deep Clustering . . . . . . 3Jingyi Wu, Baopu Qiu, and Lin Shang

A Weak Supervision Approach with Adversarial Training for NamedEntity Recognition . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17

Jianxuan Shao, Chenyang Bu, Shengwei Ji, and Xindong Wu

An Attention-Based Approach to Accelerating Sequence GenerativeAdversarial Nets . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 31

Minglei Gao, Sai Zhang, Xiaowang Zhang, and Zhiyong Feng

Autoregressive Pre-training Model-Assisted Low-Resource NeuralMachine Translation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 46

Nier Wu, Hongxu Hou, Yatu Ji, and Wei Zheng

Combining Improvements for Exploiting Dependency Trees in NeuralSemantic Parsing. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 58

Defeng Xie, Jianmin Ji, Jiafei Xu, and Ran Ji

Deep Semantic Fusion Representation Based on Special Mechanismof Information Transmission for Joint Entity-Relation Extraction. . . . . . . . . . 73

Wenqiang Xu, Shiqun Yin, Junfeng Zhao, and Ting Pu

Exploiting News Article Structure for Automatic Corpus Generationof Entailment Datasets . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 86

Jan Christian Blaise Cruz, Jose Kristian Resabal, James Lin,Dan John Velasco, and Charibeth Cheng

Fake News Detection Using Multiple-View Text Representation . . . . . . . . . . 100Tuan Ha and Xiaoying Gao

Generating Pseudo Connectives with MLMs for Implicit DiscourseRelation Recognition . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 113

Congcong Jiang, Tieyun Qian, Zhuang Chen, Kejian Tang,Shaohui Zhan, and Tao Zhan

Graph Convolutional Network Exploring Label Relations for Multi-labelText Classification. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 127

Ting Pu, Shiqun Yin, Wenwen Li, and Wenqiang Xu

Improving Long Content Question Generation with Multi-levelPassage Encoding . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 140

Peide Zhu

Learning Vietnamese-English Code-Switching Speech Synthesis ModelUnder Limited Code-Switched Data Scenario . . . . . . . . . . . . . . . . . . . . . . . 153

Cuong Manh Nguyen, Lam Viet Phung, Cuc Thi Bui, Trang Van Truong,and Huy Tien Nguyen

Multi-task Text Normalization Approach for Speech Synthesis . . . . . . . . . . . 164Cuc Thi Bui, Trang Van Truong, Cuong Manh Nguyen, Lam Viet Phung,Manh Tien Nguyen, and Huy Tien Nguyen

Performance-Driven Reinforcement Learning Approach for AbstractiveText Summarization . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 177

Trang-Phuong N. Nguyen, Nam-Chi Van, and Nhi-Thao Tran

Punctuation Prediction in Vietnamese ASRs Using Transformer-BasedModels . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 191

Viet The Bui and Oanh Thi Tran

Rumor Detection on Microblogs Using Dual-Grained Feature via GraphNeural Networks . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 205

Shouzhi Xu, Xiaodi Liu, Kai Ma, Fangmin Dong, Shunzhi Xiang,and Changsong Bing

Short Text Clustering Using Joint Optimization of Feature Representationsand Cluster Assignments . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 217

Liping Sun, Tingli Du, Xiaoyu Duan, and Yonglong Luo

Soft-BAC: Soft Bidirectional Alignment Cost for End-to-End AutomaticSpeech Recognition . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 232

Yonghe Wang, Hui Zhang, Feilong Bao, and Guanglai Gao

Span Labeling Approach for Vietnamese and ChineseWord Segmentation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 244

Duc-Vu Nguyen, Linh-Bao Vo, Dang Van Thin,and Ngan Luu-Thuy Nguyen

VSEC: Transformer-Based Model for Vietnamese Spelling Correction . . . . . . 259Dinh-Truong Do, Ha Thanh Nguyen, Thang Ngoc Bui,and Hieu Dinh Vo

What Emotion Is Hate? Incorporating Emotion Information into the HateSpeech Detection Task. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 273

Kosisochukwu Judith Madukwe, Xiaoying Gao, and Bing Xue

xx Contents – Part II

Enhanced Named Entity Recognition with Semantic Dependency . . . . . . . . . 287Peng Wang, Zhe Wang, Xiaowang Zhang, Kewen Wang,and Zhiyong Feng

Improving Sentence-Level Relation Classification via Machine ReadingComprehension and Reinforcement Learning . . . . . . . . . . . . . . . . . . . . . . . 299

Bo Xu, Zhengqi Zhang, Xiangsan Zhao, Hui Song, and Ming Du

Multi-modal and Multi-perspective Machine Translation by CollectingDiverse Alignments . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 311

Lin Li, Turghun Tayir, Kaixi Hu, and Dong Zhou

Simplifying Paragraph-Level Question Generation via TransformerLanguage Models . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 323

Luis Enrico Lopez, Diane Kathryn Cruz, Jan Christian Blaise Cruz,and Charibeth Cheng

Neural Networks and Deep Learning

ABAE: Utilize Attention to Boost Graph Auto-Encoder . . . . . . . . . . . . . . . . 337Tianyu Liu, Yifan Li, Yujie Sun, Lixin Cui, and Lu Bai

Adversarial Examples Defense via Combining Data Transformationsand RBF Layers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 349

Jingjie Li, Jiaquan Gao, and Xiao-Xin Li

An Improved Deep Model for Knowledge Tracingand Question-Difficulty Discovery. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 362

Huan Dai, Yupei Zhang, Yue Yun, and Xuequn Shang

ARNet: Accurate and Real-Time Network for Crowd Counting . . . . . . . . . . 376Yinfeng Xia, Qing He, Wenyue Wei, and Baoqun Yin

Deep Recommendation Model Based on BiLSTM and BERT. . . . . . . . . . . . 390Changwei Liu and Xiaowen Deng

GCMNet: Gated Cascade Multi-scale Network for Crowd Counting . . . . . . . 403Jinfang Zheng, Panpan Zhao, Jinyang Xie, Chen Lyu, and Lei Lyu

GIAD: Generative Inpainting-Based Anomaly Detection viaSelf-Supervised Learning for Human Monitoring. . . . . . . . . . . . . . . . . . . . . 418

Ning Dong and Einoshin Suzuki

Heterogeneous Graph Attention Network for User Geolocation . . . . . . . . . . . 433Xuan Zhang, FuQiang Lin, DiWen Dong, WangQun Chen, and Bo Liu

Contents – Part II xxi

Hyperbolic Tangent Polynomial Parity Cyclic Learning Rate for DeepNeural Network . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 448

Hong Lin, Xiaodong Yang, Binyan Wu, and Ruyan Xiong

Infrared Image Super-Resolution via HeterogeneousConvolutional WGAN . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 461

Yongsong Huang, Zetao Jiang, Qingzhong Wang, Qi Jiang,and Guoming Pang

Knowledge Compensation Network with Divisible Feature Learningfor Unsupervised Domain Adaptive Person Re-identification . . . . . . . . . . . . 473

Jiajing Hong, Yang Zhang, and Yuesheng Zhu

LoCo-VAE: Modeling Short-Term Preference as Joint Effect of Long-TermPreference and Context-Aware Impact in Recommendation . . . . . . . . . . . . . 487

Jianping Liu, Bo Wang, Ruifang He, Bin Wu, Shuo Zhang, Yuexian Hou,and Qinxue Jiang

Multi-scale Edge-Based U-Shape Network for Salient Object Detection . . . . . 501Han Sun, Yetong Bian, Ningzhong Liu, and Huiyu Zhou

Reconstruct Anomaly to Normal: Adversarially Learned and LatentVector-Constrained Autoencoder for Time-Series Anomaly Detection . . . . . . 515

Chunkai Zhang, Wei Zuo, Shaocong Li, Xuan Wang, Peiyi Han,and Chuanyi Liu

Robust Ensembling Network for Unsupervised Domain Adaptation . . . . . . . . 530Han Sun, Lei Lin, Ningzhong Liu, and Huiyu Zhou

SPAN: Subgraph Prediction Attention Network for Dynamic Graphs. . . . . . . 544Yuan Li, Chuanchang Chen, Yubo Tao, and Hai Lin

WINVC: One-Shot Voice Conversion with Weight AdaptiveInstance Normalization . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 559

Shengjie Huang, Mingjie Chen, Yanyan Xu, Dengfeng Ke,and Thomas Hain

Fusion Graph Convolutional Collaborative Filtering . . . . . . . . . . . . . . . . . . 574Zeqi Zhang, Ying Liu, and Fengli Sun

Multi-label Learning by Exploiting Imbalanced Label Correlations . . . . . . . . 585Shiqiao Gu, Liu Yang, Yaning Li, and Hui Li

Random Sparsity Defense Against Adversarial Attack . . . . . . . . . . . . . . . . . 597Nianyan Hu, Ting Lu, Wenjing Guo, Qiubo Huang, Guohua Liu,Shan Chang, Jiafei Song, and Yiyang Luo

Author Index . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 609

xxii Contents – Part II

Contents – Part III

Reinforcement Learning

Consistency Regularization for Ensemble Model BasedReinforcement Learning . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3

Ruonan Jia, Qingming Li, Wenzhen Huang, Junge Zhang, and Xiu Li

Detecting and Learning Against Unknown Opponentsfor Automated Negotiations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17

Leling Wu, Siqi Chen, Xiaoyang Gao, Yan Zheng, and Jianye Hao

Diversity-Based Trajectory and Goal Selection with HindsightExperience Replay . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 32

Tianhong Dai, Hengyan Liu, Kai Arulkumaran, Guangyu Ren,and Anil Anthony Bharath

Off-Policy Training for Truncated TD(k) Boosted Soft Actor-Critic . . . . . . . . 46Shiyu Huang, Bin Wang, Hang Su, Dong Li, Jianye Hao, Jun Zhu,and Ting Chen

Adaptive Warm-Start MCTS in AlphaZero-Like DeepReinforcement Learning . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 60

Hui Wang, Mike Preuss, and Aske Plaat

Batch-Constraint Inverse Reinforcement Learning . . . . . . . . . . . . . . . . . . . . 72Mao Chen, Li Wan, Chunyan Gou, Jiaolu Liao, and Shengjiang Wu

KG-RL: A Knowledge-Guided Reinforcement Learning for MassiveBattle Games . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 83

Shiyang Zhou, Weiya Ren, Xiaoguang Ren, Xianya Mi, and Xiaodong Yi

Vision and Perception

A Semi-supervised Defect Detection Method Based on Image Inpainting . . . . 97Huibin Cao, Yongxuan Lai, Quan Chen, and Fan Yang

ANF: Attention-Based Noise Filtering Strategy for UnsupervisedFew-Shot Classification . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 109

Guangsen Ni, Hongguang Zhang, Jing Zhao, Liyang Xu, Wenjing Yang,and Long Lan

Asymmetric Mutual Learning for Unsupervised Cross-DomainPerson Re-identification . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 124

Danyang Huang, Lei Zhang, Qishuai Diao, Wei Wu, and Zhong Zhou

Collaborative Positional-Motion Excitation Module for EfficientAction Recognition . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 138

Tamam Alsarhan and Hongtao Lu

Graph Attention Convolutional Network with Motion Tempo Enhancementfor Skeleton-Based Action Recognition . . . . . . . . . . . . . . . . . . . . . . . . . . . 152

Ruwen Bai, Xiang Meng, Bo Meng, Miao Jiang, Junxing Ren,Yang Yang, Min Li, and Degang Sun

Learning to Synthesize and Remove Rain Unsupervisedly . . . . . . . . . . . . . . 166Yinhe Qi, Meng Pan, and Zhi Jin

Object Bounding Box-Aware Embedding for Point CloudInstance Segmentation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 182

Lixue Cheng, Taihai Yang, and Lizhuang Ma

Objects as Extreme Points . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 195Yang Yang, Min Li, Bo Meng, Zihao Huang, Junxing Ren,and Degang Sun

Occlusion-Aware Facial Expression Recognition Based RegionRe-weight Network . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 209

Xinghai Zhang, Xingming Zhang, Jinzhao Zhou, and Yubei Lin

Online Multi-Object Tracking with Pose-Guided Object Location and DualSelf-Attention Network . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 223

Xin Zhang, Shihao Wang, Yuanzhe Yang, Chengxiang Chu,and Zhong Zhou

Random Walk Erasing with Attention Calibration for Action Recognition . . . 236Yuze Tian, Xian Zhong, Wenxuan Liu, Xuemei Jia, Shilei Zhao,and Mang Ye

RGB-D Based Visual Navigation Using Direction Estimation Module . . . . . . 252Chao Luo, Sheng Bi, Min Dong, and Hongxu Nie

Semi-supervised Single Image Deraining with DiscreteWavelet Transform . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 265

Xin Cui, Wei Shang, Dongwei Ren, Pengfei Zhu, and Yankun Gao

Simple Light-Weight Network for Human Pose Estimation. . . . . . . . . . . . . . 279Bin Sun and Mingguo Zhao

xxiv Contents – Part III

SIN: Superpixel Interpolation Network. . . . . . . . . . . . . . . . . . . . . . . . . . . . 293Qing Yuan, Songfeng Lu, Yan Huang, and Wuxin Sha

SPANet: Spatial and Part-Aware Aggregation Network for 3DObject Detection . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 308

Yangyang Ye

Subspace Enhancement and Colorization Network for Infrared VideoAction Recognition . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 321

Lu Xu, Xian Zhong, Wenxuan Liu, Shilei Zhao, Zhengwei Yang,and Luo Zhong

Thinking in Patch: Towards Generalizable Forgery Detectionwith Patch Transformation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 337

Xueqi Zhang, Shuo Wang, Chenyu Liu, Min Zhang, Xiaohan Liu,and Haiyong Xie

When Distortion Meets Perceptual Quality: A Multi-taskLearning Pipeline . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 353

Jing Wen and Qianyu Guo

Feature Adaption with Predicted Boxes for Oriented Object Detectionin Aerial Images . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 366

Minhao Zou, Ziye Hu, Yuxiang Guan, Zhongxue Gan, Chun Guan,and Siyang Leng

Few-Shot Crowd Counting via Self-supervised Learning . . . . . . . . . . . . . . . 379Jiefeng Long, Chun Li, and Lin Shang

Low-Rank Orthonormal Analysis Dictionary Learning for ImageClassification . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 391

Kun Jiang, Zhaoli Liu, and Qindong Sun

MRAC-Net: Multi-resolution Anisotropic Convolutional Network for 3DPoint Cloud Completion. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 403

Sheng Liu, Dingda Li, Wenhao Huang, Yifeng Cao,and Shengyong Chen

Nonlinear Parametric Transformation and Generation of Images Basedon a Network with the CWNL Layer. . . . . . . . . . . . . . . . . . . . . . . . . . . . . 415

Slawomir Golak

PupilFace: A Cascaded Face Detection and Location NetworkFusing Attention . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 426

Xiang Li and Jiancheng Zou

Author Index . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 439

Contents – Part III xxv