Databases Theory and Applications Junhu Wang

download

https://textbookfull.com/product/databases-theory-and-
applications-junhu-wang/

Download more ebook from https://textbookfull.com

We believe these products will be a great fit for you. Click
the link to download now, or visit textbookfull.com
to discover even more!

MongoDB Performance Tuning: Optimizing MongoDB

Databases and their Applications 1st Edition Guy
Harrison

https://textbookfull.com/product/mongodb-performance-tuning-
optimizing-mongodb-databases-and-their-applications-1st-edition-
guy-harrison/

A deep dive into NoSQL databases the use cases and

applications First Edition Raj

https://textbookfull.com/product/a-deep-dive-into-nosql-
databases-the-use-cases-and-applications-first-edition-raj/

Theory of Translation Closedness for Time Scales With

Applications in Translation Functions and Dynamic
Equations Developments in Mathematics 62 Chao Wang

https://textbookfull.com/product/theory-of-translation-
closedness-for-time-scales-with-applications-in-translation-
functions-and-dynamic-equations-developments-in-
mathematics-62-chao-wang/

Seven Databases in Seven Weeks A Guide to Modern

Databases and the NoSQL Movement Perkins

https://textbookfull.com/product/seven-databases-in-seven-weeks-
a-guide-to-modern-databases-and-the-nosql-movement-perkins/
Microcontroller Theory and Applications Rafiquzzaman

https://textbookfull.com/product/microcontroller-theory-and-
applications-rafiquzzaman/

Supercritical Water Processing Technologies for

Environment Energy and Nanomaterial Applications
Shuzhong Wang

https://textbookfull.com/product/supercritical-water-processing-
technologies-for-environment-energy-and-nanomaterial-
applications-shuzhong-wang/

Defoaming : Theory and Industrial Applications Garrett

https://textbookfull.com/product/defoaming-theory-and-industrial-
applications-garrett/

Relative fidelity processing of seismic data : methods

and applications 1 edition Edition Wang

https://textbookfull.com/product/relative-fidelity-processing-of-
seismic-data-methods-and-applications-1-edition-edition-wang/

Eukaryotic Genomic Databases Martin Kollmar

https://textbookfull.com/product/eukaryotic-genomic-databases-
martin-kollmar/
 Junhu Wang · Gao Cong
Jinjun Chen · Jianzhong Qi (Eds.)
LNCS 10837

Databases Theory
and Applications
29th Australasian Database Conference, ADC 2018
Gold Coast, QLD, Australia, May 24–27, 2018
Proceedings

123
Lecture Notes in Computer Science 10837
Commenced Publication in 1973
Founding and Former Series Editors:
Gerhard Goos, Juris Hartmanis, and Jan van Leeuwen

Editorial Board
David Hutchison
Lancaster University, Lancaster, UK
Takeo Kanade
Carnegie Mellon University, Pittsburgh, PA, USA
Josef Kittler
University of Surrey, Guildford, UK
Jon M. Kleinberg
Cornell University, Ithaca, NY, USA
Friedemann Mattern
ETH Zurich, Zurich, Switzerland
John C. Mitchell
Stanford University, Stanford, CA, USA
Moni Naor
Weizmann Institute of Science, Rehovot, Israel
C. Pandu Rangan
Indian Institute of Technology Madras, Chennai, India
Bernhard Steffen
TU Dortmund University, Dortmund, Germany
Demetri Terzopoulos
University of California, Los Angeles, CA, USA
Doug Tygar
University of California, Berkeley, CA, USA
Gerhard Weikum
Max Planck Institute for Informatics, Saarbrücken, Germany
More information about this series at http://www.springer.com/series/7409
Junhu Wang Gao Cong
•

Jinjun Chen Jianzhong Qi (Eds.)

•

Databases Theory
and Applications
29th Australasian Database Conference, ADC 2018
Gold Coast, QLD, Australia, May 24–27, 2018
Proceedings

123
Editors
Junhu Wang Jinjun Chen
ICT Faculty of Information and Communication
Griffith University Technologies
Southport, QLD Swinburne University of Technology
Australia Hawthorn, VIC
Australia
Gao Cong
Nanyang Technological University Jianzhong Qi
Singapore The University of Melbourne
Singapore Melbourne, VIC
Australia

ISSN 0302-9743 ISSN 1611-3349 (electronic)

Lecture Notes in Computer Science
ISBN 978-3-319-92012-2 ISBN 978-3-319-92013-9 (eBook)
https://doi.org/10.1007/978-3-319-92013-9

Library of Congress Control Number: 2018944378

LNCS Sublibrary: SL3 – Information Systems and Applications, incl. Internet/Web, and HCI

© Springer International Publishing AG, part of Springer Nature 2018

This work is subject to copyright. All rights are reserved by the Publisher, whether the whole or part of the
material 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 information
storage and retrieval, electronic adaptation, computer software, or by similar or dissimilar methodology now
known or hereafter developed.
The use of general descriptive names, registered names, trademarks, service marks, etc. in this publication
does not imply, even in the absence of a specific statement, that such names are exempt from the relevant
protective 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 are
believed to be true and accurate at the date of publication. Neither the publisher nor the authors or the editors
give a warranty, express or implied, with respect to the material contained herein or for any errors or
omissions that may have been made. The publisher remains neutral with regard to jurisdictional claims in
published maps and institutional affiliations.

Printed on acid-free paper

This Springer imprint is published by the registered company Springer International Publishing AG
part of Springer Nature
The registered company address is: Gewerbestrasse 11, 6330 Cham, Switzerland
 Preface

It is our great pleasure to present the proceedings of the 29th Australasian Database
Conference (ADC 2018). The Australasian Database Conference is an annual inter-
national forum for sharing the latest research advancements and novel applications of
database systems, data-driven applications, and data analytics between researchers and
practitioners from around the globe, particularly Australia and New Zealand. The
mission of ADC is to share novel research solutions to problems of today’s information
society that fulfil the needs of heterogeneous applications and environments and to
identify new issues and directions for future research. ADC seeks papers from aca-
demia and industry presenting research on all practical and theoretical aspects of
advanced database theory and applications, as well as case studies and implementation
experiences.
ADC 2018 was held during May 23–25, 2018, on the Gold Coast, Australia. As in
previous years, ADC 2018 accepted all the papers that the Program Committee con-
sidered as being of ADC quality without setting any predefined quota. The conference
received 53 submissions, each of which was carefully peer reviewed by at least three
independent reviewers, and in some cases four or five reviewers. Based on the reviewer
comments, we accepted 23 full research papers, six short papers, and three demo
papers. The Program Committee that selected the papers comprised 52 members from
around the world including Australia, China, USA, Finland, Denmark, Switzerland,
Japan, New Zealand, and Singapore. The conference programme also includes keynote
talks and invited tutorials for ADC’s PhD school.
We are grateful to Professor Xiaofang Zhou (University of Queensland, ADC
Steering Committee member) for his helpful advice, Professor Rui Zhang (University
of Melbourne, ADC 2018 General Chair), and Dr. Sen Wang (Griffith University, ADC
2018 Local Organization Chair) for their tireless work in coordinating the conference
activities. We would like to thank all members of the Organizing Committee, and the
many volunteers, for their support in the conference organization. Special thanks go to
the Program Committee members and the external reviewers who contributed their time
and expertise in the paper review process. We would also like to thank the invited
speakers, all authors who submitted their papers, and all conference attendees.

May 2018 Junhu Wang

Gao Cong
Jinjun Chen
Jianzhong Qi
 General Chair’s Welcome Message

Welcome to the proceedings of the 29th Australasian Database Conference (ADC

2018)! ADC is a leading Australia- and New Zealand-based international conference
on research and applications of database systems, data-driven applications, and data
analytics. In the past 10 years, ADC has been held in Brisbane (2017), Sydney (2016),
Melbourne (2015), Brisbane (2014), Adelaide (2013), Melbourne (2012), Perth (2011),
Brisbane (2010), Wellington (2009), and Wollongong (2008). This year, the ADC
conference came to the Gold Coast.
In the past, the ADC conference series was held as part of the Australasian Com-
puter Science Week (ACSW). Starting from 2014, the ADC conferences departed from
ACSW as the database research community in Australasia has grown significantly
larger. Now the new ADC conference has an expanded research program and focuses
on community-building through a PhD School. ADC 2018 was the fifth of this new
ADC conference series.
In addition to 23 full research papers, six short research papers, and three demo
papers carefully selected by the Program Committee, we were also very fortunate to
have five invited talks presented by world-leading researchers: Kyuseok Shim from
Seoul National University, South Korea, Reynold Cheng from The University of Hong
Kong, Hong Kong SAR, Shuai Ma from Beihang University, China, Lina Yao from
The University of New South Wales, Australia, and Hongzhi Yin and Weiqing Wang
from The University of Queensland, Australia. We had a two-day PhD School program
as part of this year’s ADC.
We wish to take this opportunity to thank all speakers, authors, and organizers.
I would also especially like to thank our Organizing Committee members: Program
Committee Chairs Junhu Wang, Gao Cong, and Jinjun Chen, for their dedication in
ensuring a high-quality program, Proceedings Chair Jianzhong Qi, for his effort in
delivering the conference proceedings timely, Local Organization Chairs Sen Wang
and Sibo Wang, for their consideration in covering every detail of the conference
logistics, and Publicity Chair Lijun Chang, for his efforts in disseminating our call for
papers and attracting submissions. Without them, this year’s ADC would not have been
a success.
The Gold Coast is a coastal city and ADC 2018 was held at the Mantra On View
Hotel in the heart of Surfers Paradise. We hope all ADC 2018 participants had a
wonderful experience with the conference and the city.

Rui Zhang
 Organization

General Chair
Rui Zhang University of Melbourne, Australia

Program Chairs
Junhu Wang Griffith University, Australia
Gao Cong Nanyang Technical University, Singapore
Jinjun Chen Swinburne University of Technology, Australia

Proceedings Chair
Jianzhong Qi University of Melbourne, Australia

Publicity Chair
Lijun Chang University of Sydney, Australia

Local Organization Chairs

Sen Wang Griffith University, Australia
Sibo Wang University of Queensland, Australia

Steering Committee
Rao Kotagiri University of Melbourne, Australia
Timos Sellis RMIT University, Australia
Gill Dobbie University of Auckland, New Zealand
Alan Fekete University of Sydney, Australia
Xuemin Lin University of New South Wales, Australia
Yanchun Zhang Victoria University, Australia
Xiaofang Zhou University of Queensland, Australia

Program Committee
Tarique Anwar Swinburne University of Technology, Australia
Zhifeng Bao RMIT University, Australia
Huiping Cao New Mexico State University, USA
Xin Cao University of New South Wales, Australia
Lijun Chang University of Sydney, Australia
Muhammad Aamir Cheema Monash University, Australia
X Organization

Lisi Chen Hong Kong Baptist University, SAR China

Farhana Murtaza RMIT University, Australia
Choudhury
Shumo Chu University of Washington, USA
Kaiyu Feng Nanyang Technological University, Singapore
Janusz Getta University of Wollongong, Australia
Yusuke Gotoh Okayama University, Japan
Tao Guo Nanyang Technological University, Singapore
Michael E. Houle National Institute of Informatics, Japan
Wen Hua University of Queensland, Australia
Guangyan Huang Deakin University, Australia
Zi Huang University of Queensland, Australia
Md. Saiful Islam Griffith University, Australia
Guoliang Li Tsinghua University, China
Jianxin Li University of Western Australia
Xiang Lian Kent State University, USA
Jixue Liu University of South Australia
Jiaheng Lu University of Helsinki, Finland
Parth Nagarkar Arizona State University, USA
Quoc Viet Hung Nguyen Griffith University, Australia
Makoto Onizuka Osaka University, Japan
Miao Qiao Massey University, New Zealand
Lu Qin University of Technology, Sydney, Australia
Goce Ristanoski Data61, Australia
Shazia Sadiq University of Queensland, Australia
Timos Sellis Swinburne University of Technology, Australia
Michael Sheng Macquarie University, Australia
Jingkuan Song University of Queensland, Australia
Bela Stantic Griffith University, Australia
Farhan Tauheed Oracle Labs Zurich, Switzerland
Anwaar Ulhaq Victoria University, Australia
Hua Wang Victoria University, Australia
Hongzhi Wang Harbin Institute of Technology, China
Chuan Xiao Nagoya University, Japan
Yajun Yang Tianjin University, China
Weiren Yu University of New South Wales, Australia
Wenjie Zhang University of New South Wales, Australia
Ying Zhang University of Technology, Sydney, Australia
Kai Zheng University of Electronic Science and Technology
of China, China
Yongluan Zhou University of Copenhagen, Denmark
Rui Zhou Swinburne University of Technology, Australia
Yi Zhou University of Technology, Sydney, Australia
Yuanyuan Zhu Wuhan University, China
Can Wang Griffith University, Australia
 Organization XI

External Reviewers
Taotao Cai University of Western Australia
Xuefeng Chen University of New South Wales, Australia
Qixu Gong New Mexico State University, USA
Yifan Hao New Mexico State University, USA
Nguyen Quoc Viet Hung Griffith University, Australia
Md Zahidul Islam University of South Australia
Saiful Islam Griffith University, Australia
Selasi Kwasie University of South Australia
Yadan Luo University of Queensland, Australia
Yue Qian Dalian University of Technology, China
Nguyen Khoi Tran University of Adelaide, Australia
Edgar Ceh Varela New Mexico State University, USA
Can Wang Griffith University, Australia
Fan Wang Aston University, UK
Lujing Yang University of South Australia
Invited Talks
 MapReduce Algorithms for Big Data Analysis

Kyuseok Shim

Seoul National University

Abstract. There is a growing trend of applications that should handle big data.
However, analyzing big data is very challenging today. For such applications,
the MapReduce framework has recently attracted a lot of attention. MapReduce
is a programming model that allows easy development of scalable parallel
applications to process big data on large clusters of commodity machines.
Google’s MapReduce or its open-source equivalent Hadoop is a powerful tool
for building such applications. In this tutorial, I will first introduce the
MapReduce framework based on Hadoop system available to everyone to run
distributed computing algorithms using MapReduce. I will next discuss how to
design efficient MapReduce algorithms and present the state-of-the-art in
MapReduce algorithms for big data analysis. Since Spark is recently developed
to overcome the shortcomings of MapReduce which is not optimized for of
iterative algorithms and interactive data analysis, I will also present an outline of
Spark as well as the differences between MapReduce and Spark. The intended
audience of this tutorial is professionals who plan to develop efficient
MapReduce algorithms and researchers who should be aware of the
state-of-the-art in MapReduce algorithms available today for big data analysis.

Short Biography. Kyuseok Shim is currently a professor at electrical and computer

engineering department in Seoul National University, Korea. Before that, he was an
assistant professor at computer science department in KAIST and a member of tech-
nical staff for the Serendip Data Mining Project at Bell Laboratories. He was also a
member of the Quest Data Mining Project at the IBM Almaden Research Center and
visited Microsoft Research at Redmond several times as a visiting scientist. Kyuseok
was named an ACM Fellow for his contributions to scalable data mining and query
processing research in 2013. Kyuseok has been working in the area of databases
focusing on data mining, search engines, recommendation systems, MapReduce
algorithms, privacy preservation, query processing and query optimization. His writ-
ings have appeared in a number of professional conferences and journals including
ACM, VLDB and IEEE publications. He served as a Program Committee member for
SIGKDD, SIGMOD, ICDE, ICDM, ICDT, EDBT, PAKDD, VLDB and WWW
conferences. He also served as a Program Committee Co-Chair for PAKDD 2003,
WWW 2014, ICDE 2015 and APWeb 2016. Kyuseok was previously on the editorial
board of VLDB as well as IEEE TKDE Journals and is currently a member of the
VLDB Endowment Board of Trustees. He received the BS degree in electrical engi-
neering from Seoul National University in 1986, and the MS and PhD degrees in
computer science from the University of Maryland, College Park, in 1988 and 1993,
respectively.
 Meta Paths and Meta Structures: Analysing
Large Heterogeneous Information Networks

Reynold Cheng

University of Hong Kong

Abstract. A heterogeneous information network (HIN) is a graph model in

which objects and edges are annotated with types. Large and complex databases,
such as YAGO and DBLP, can be modeled as HINs. A fundamental problem in
HINs is the computation of closeness, or relevance, between two HIN objects.
Relevance measures, such as PCRW, PathSim, and HeteSim, can be used in
various applications, including information retrieval, entity resolution, and
product recommendation. These metrics are based on the use of meta paths,
essentially a sequence of node classes and edge types between two nodes in a
HIN. In this tutorial, we will give a detailed review of meta paths, as well as how
they are used to define relevance. In a large and complex HIN, retrieving meta
paths manually can be complex, expensive, and error-prone. Hence, we will
explore systematic methods for finding meta paths. In particular, we will study a
solution based on the Query-by-Example (QBE) paradigm, which allows us to
discovery meta paths in an effective and efficient manner.
We further generalise the notion of a meta path to “meta structure”, which is
a directed acyclic graph of object types with edge types connecting them. Meta
structure, which is more expressive than the meta path, can describe complex
relationship between two HIN objects (e.g., two papers in DBLP share the same
authors and topics). We develop three relevance measures based on meta
structure. Due to the computational complexity of these measures, we also study
an algorithm with data structures proposed to support their evaluation. Finally,
we will examine solutions for performing query recommendation based on meta
paths. We will also discuss future research directions in HINs.

Short Biography. Dr. Reynold Cheng is an Associate Professor of the Department of

Computer Science in the University of Hong Kong. He was an Assistant Professor in
HKU in 2008–2011. He received his BEng (Computer Engineering) in 1998, and
MPhil (Computer Science and Information Systems) in 2000, from the Department of
Computer Science in the University of Hong Kong. He then obtained his MSc and PhD
from Department of Computer Science of Purdue University in 2003 and 2005
respectively. Dr. Cheng was an Assistant Professor in the Department of Computing
of the Hong Kong Polytechnic University during 2005-08. He was a visiting scientist in
the Institute of Parallel and Distributed Systems in the University of Stuttgart during
the summer of 2006.
Dr. Cheng was granted an Outstanding Young Researcher Award 2011–2012 by
HKU. He was the recipient of the 2010 Research Output Prize in the Department of
Computer Science of HKU. He also received the U21 Fellowship in 2011. He received
 Meta Paths and Meta Structures XVII

the Performance Reward in years 2006 and 2007 awarded by the Hong Kong
Polytechnic University. He is the Chair of the Department Research Postgraduate
Committee, and was the Vice Chairperson of the ACM (Hong Kong Chapter) in 2013.
He is a member of the IEEE, the ACM, the Special Interest Group on Management of
Data (ACM SIGMOD), and the UPE (Upsilon Pi Epsilon Honor Society). He is an
editorial board member of TKDE, DAPD and IS, and was a guest editor for TKDE,
DAPD, and Geoinformatica. He is an area chair of ICDE 2017, a senior PC member for
DASFAA 2015, PC co-chair of APWeb 2015, area chair for CIKM 2014, area chair for
Encyclopedia of Database Systems, program co-chair of SSTD 2013, and a workshop
co-chair of ICDE 2014. He received an Outstanding Service Award in the CIKM 2009
conference. He has served as PC members and reviewer for top conferences (e.g.,
SIGMOD, VLDB, ICDE, EDBT, KDD, ICDM, and CIKM) and journals (e.g., TODS,
TKDE, VLDBJ, IS, and TMC).
 Approximate Computation for Big Data
Analytics

Shuai Ma

Beihang University

Abstract. Over the past a few years, research and development has made sig-
nificant progresses on big data analytics with the supports from both govern-
ments and industries all over the world, such as Spark, IBM Watson and Google
AlphaGo. A fundamental issue for big data analytics is the efficiency, and
various advances towards attacking these issues have been achieved recently,
from theory to algorithms to systems. In this talk, we shall present the idea of
approximate computation for efficient and effective big data analytics: query
approximation and data approximation, based on our recent research experi-
ences. Different from existing approximation techniques, the approximation
computation that we are going to introduce does not necessarily ask for theo-
retically guaranteed approximation solutions, but asks for sufficiently efficient
and effective solutions in practice.

Short Biography. Shuai Ma is a full professor in the School of Computer Science and
Engineering, Beihang University, China. He obtained two PhD degrees: University of
Edinburgh in 2010 and Peking University in 2004, respectively. His research interests
include database theory and systems, and big data. He is a recipient of the best paper
award of VLDB 2010, the best challenge paper award of WISE 2013, the National
Science Fund of China for Excellent Young Scholars in 2013, and the special award of
Chinese Institute of Electronics for progress in science and technology in 2017 (8/15).
He is an Associate Editor of VLDB Journal since 2017.
 Understanding Human Behaviors via Learning
Internet of Things Interactions

Lina Yao

The University of New South Wales

Abstract. Internet of Things (IoT) enables the connection and integration of

physical world and virtual world. A vast amount of interactive data between
human and the real world being created by diverse sensing sources can be readily
collected. Such growing interconnections powered with intelligent approaches
open up a new world of broader possibilities and innovations with a deeper
understanding of human behaviors. In this tutorial, I will introduce the method-
ologies to learn actionable knowledge from the monitored environment, in order
to take actions on the situations and improve decision-making process, present
real-world application examples and discuss the future research directions.

Short Biography. Lina Yao is currently a lecturer in the School of Computer Science
and Engineering, University of New South Wales. Her research interests lie in data
mining and machine learning applications with the focuses on Internet of Things,
recommender systems, human activity recognition and Brain-Computer Interface.
 Mining Geo-social Networks – Spatial Item
Recommendation

Hongzhi Yin and Weiqing Wang

The University of Queensland

Abstract. The rapid development of Web 2.0, location acquisition and wireless
communication technologies has fostered a profusion of geo-social networks
(e.g., Foursquare, Yelp and Google Place). They provide users an online plat-
form to check-in at points of interests (e.g., cinemas, galleries and hotels) and
share their life experiences in the physical world via mobile devices. The new
dimension of location implies extensive knowledge about an individual’s
behaviors and interests by bridging the gap between online social networks and
the physical world. It is crucial to develop spatio-temporal recommendation
services for mobile users to explore the new places, attend new events and find
their potentially preferred spatial items from billions of candidate ones. Com-
pared with traditional recommendation tasks, the spatio-temporal recommen-
dation faces the following new challenges: Travel Locality, Spatial Dynamics of
User Interests, Temporal Dynamics of User Interests, Sequential Influence of
user mobility behaviors and Real-time Requirement. In this talk, I will present
our recent advancement of spatio-temporal recommendation techniques and how
to address these unique challenges.

Short Biography. Dr. Hongzhi Yin is now working as a lecturer in data science and an
ARC DECRA Fellow (Australia Discovery Early Career Researcher Award) with The
University of Queensland, Australia. He received his doctoral degree from Peking
University in July 2014. After graduation, he joined the school of ITEE, the University
of Queensland. He successfully won the ARC DECRA award in 2015 and obtained an
ARC Discovery Project grant as a chief investigator in 2016. His current main research
interests include social media analytic, user profiling, recommender system, especially
spatial-temporal recommendation, topic discovery and event detection, deep learning,
user linkage across social networks, knowledge graph mining and construction. He has
published over 70 peer-reviewed papers in prestigious journals and top international
conferences including ACM TOIS, VLDBJ, IEEE TKDE, ACM TKDD, ACM TIST,
ACM SIGMOD, ACM SIGKDD, VLDB, IEEE ICDE, AAAI, SIGIR, WWW, ACM
Multimedia, ICDM, WSDM and CIKM. He has been actively engaged in professional
services by serving as conference organizers, conference PC members for PVLDB,
SIGIR, ICDE, IJCAI, ICDM, CIKM, DASFAA, ASONAM, MDM, WISE, PAKDD
and reviewer of more than 10 reputed journals such as VLDB Journal, TKDE, TOIS,
TKDD, TWeb, IEEE Transactions on Cybernetics, WWW Journal, Knowledge-based
system and etc.
 Mining Geo-social Networks – Spatial Item Recommendation XXI

Dr. Weiqing Wang is now working as a Research Fellow in the school of ITEE, the
University of Queensland, where she also obtained her PhD in July on 2017. She will
join Monash University as a lecturer in data science in this July. Her major research
interests include user modelling and recommender systems, especially spatial-temporal
recommender systems. She has published over ten peer-reviewed papers in prestigious
journals and top conferences including IEEE TKDE, ACM TOIS, ACM TIST,
ACM SIGKDD, ACM SIGIR, IEEE ICDE, ACM Multimedia, and CIKM.
 Contents

Full Research Papers: Database and Applications

Adaptive Access Path Selection for Hardware-Accelerated DRAM Loads. . . . 3

Markus Dreseler, Timo Gasda, Jan Kossmann, Matthias Uflacker,
and Hasso Plattner

Privacy Preservation for Trajectory Data Publishing by Look-Up

Table Generalization . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15
Nattapon Harnsamut, Juggapong Natwichai, and Surapon Riyana

Trajectory Set Similarity Measure: An EMD-Based Approach . . . . . . . . . . . 28

Dan He, Boyu Ruan, Bolong Zheng, and Xiaofang Zhou

Histogram Construction for Difference Analysis of Spatio-Temporal

Data on Array DBMS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 41
Jing Zhao, Yoshiharu Ishikawa, Chuan Xiao, and Kento Sugiura

Location-Aware Group Preference Queries in Social-Networks . . . . . . . . . . . 53

Ammar Sohail, Arif Hidayat, Muhammad Aamir Cheema,
and David Taniar

Social-Textual Query Processing on Graph Database Systems. . . . . . . . . . . . 68

Oshini Goonetilleke, Timos Sellis, and Xiuzhen Zhang

Using SIMD Instructions to Accelerate Sequence Similarity Searches Inside

a Database System . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 81
Sidath Randeni Kadupitige and Uwe Röhm

Renovating Database Applications with DBAutoAwesome . . . . . . . . . . . . . . 94

Jonathan Adams and Curtis E. Dyreson

Full Research Papers: Data Mining and Applications

Uncovering Attribute-Driven Active Intimate Communities . . . . . . . . . . . . . 109

Md Musfique Anwar, Chengfei Liu, and Jianxin Li

Customer Churn Prediction in Superannuation: A Sequential Pattern

Mining Approach . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 123
Ben Culbert, Bin Fu, James Brownlow, Charles Chu, Qinxue Meng,
and Guandong Xu

Automated Underwriting in Life Insurance: Predictions and Optimisation. . . . 135

Rhys Biddle, Shaowu Liu, Peter Tilocca, and Guandong Xu
XXIV Contents

Maintaining Boolean Top-K Spatial Temporal Results

in Publish-Subscribe Systems . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 147
Maryam Ghafouri, Xiang Wang, Long Yuan, Ying Zhang,
and Xuemin Lin

Interdependent Model for Point-of-Interest Recommendation

via Social Networks . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 161
Jake Hashim-Jones, Can Wang, Md. Saiful Islam, and Bela Stantic

Child Abuse and Domestic Abuse: Content and Feature Analysis

from Social Media Disclosures . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 174
Sudha Subramani, Hua Wang, Md Rafiqul Islam, Anwaar Ulhaq,
and Manjula O’Connor

30 min-Ahead Gridded Solar Irradiance Forecasting Using Satellite Data . . . . 186

Todd Taomae, Lipyeow Lim, Duane Stevens, and Dora Nakafuji

An Efficient Framework for the Analysis of Big Brain Signals Data . . . . . . . 199
Supriya, Siuly, Hua Wang, and Yanchun Zhang

Full Research Papers: Theories and Methodologies

TSAUB: A Temporal-Sentiment-Aware User Behavior Model

for Personalized Recommendation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 211
Qinyong Wang, Hongzhi Yin, Hao Wang, and Zi Huang

Finding Maximal Stable Cores in Social Networks . . . . . . . . . . . . . . . . . . . 224

Alexander Zhou, Fan Zhang, Long Yuan, Ying Zhang,
and Xuemin Lin

Feature Extraction for Smart Sensing Using Multi-perspectives

Transformation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 236
Sanad Al-Maskari, Ibrahim A. Ibrahim, Xue Li, Eimad Abusham,
and Abdulqader Almars

Finding Influential Nodes by a Fast Marginal Ranking Method. . . . . . . . . . . 249

Yipeng Zhang, Ping Zhang, Zhifeng Bao, Zizhe Xie, Qizhi Liu,
and Bang Zhang

Maximizing Reverse k-Nearest Neighbors for Trajectories . . . . . . . . . . . . . . 262

Tamjid Al Rahat, Arif Arman, and Mohammed Eunus Ali

Auto-CES: An Automatic Pruning Method Through Clustering

Ensemble Selection . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 275
Mojtaba Amiri Maskouni, Saeid Hosseini,
Hadi Mohammadzadeh Abachi, Mohammadreza Kangavari,
and Xiaofang Zhou
 Contents XXV

DistClusTree: A Framework for Distributed Stream Clustering . . . . . . . . . . . 288

Zhinoos Razavi Hesabi, Timos Sellis, and Kewen Liao

Short Research Papers

Mobile Application Based Heavy Vehicle Fatigue Compliance

in Australian Operations. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 303
Luke Mirowski and Joel Scanlan

Statistical Discretization of Continuous Attributes

Using Kolmogorov-Smirnov Test . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 309
Hadi Mohammadzadeh Abachi, Saeid Hosseini,
Mojtaba Amiri Maskouni, Mohammadreza Kangavari,
and Ngai-Man Cheung

Econometric Analysis of the Industrial Growth Determinants

in Colombia . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 316
Carolina Henao-Rodríguez, Jenny-Paola Lis-Gutiérrez,
Mercedes Gaitán-Angulo, Luz Elena Malagón,
and Amelec Viloria

Parallelizing String Similarity Join Algorithms . . . . . . . . . . . . . . . . . . . . . . 322

Ling-Chih Yao and Lipyeow Lim

Exploring Human Mobility Patterns in Melbourne Using Social

Media Data . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 328
Ravinder Singh, Yanchun Zhang, and Hua Wang

Bootstrapping Uncertainty in Schema Covering. . . . . . . . . . . . . . . . . . . . . . 336

Nguyen Thanh Toan, Phan Thanh Cong, Duong Chi Thang,
Nguyen Quoc Viet Hung, and Bela Stantic

Demo Papers

TEXUS: Table Extraction System for PDF Documents . . . . . . . . . . . . . . . . 345

Roya Rastan, Hye-Young Paik, John Shepherd, Seung Hwan Ryu,
and Amin Beheshti

Visual Evaluation of SQL Plan Cache Algorithms. . . . . . . . . . . . . . . . . . . . 350

Jan Kossmann, Markus Dreseler, Timo Gasda, Matthias Uflacker,
and Hasso Plattner

Visualising Top-k Alternative Routes. . . . . . . . . . . . . . . . . . . . . . . . . . . . . 354

Lingxiao Li, Muhammad Aamir Cheema, David Taniar,
and Maria Indrawan-Santiago

Author Index . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 359

Full Research Papers: Database
and Applications
 Adaptive Access Path Selection
for Hardware-Accelerated DRAM Loads

Markus Dreseler(B) , Timo Gasda, Jan Kossmann, Matthias Uflacker,

and Hasso Plattner

Hasso Plattner Institute, Potsdam, Germany

[email protected]

Abstract. For modern main memory database systems, the memory

bus is the main bottleneck. Specialized hardware components of large
NUMA systems, such as HPE’s GRU, make it possible to offload mem-
ory transfers. In some cases, this improves the throughput by 30%, but
other scenarios suffer from reduced performance. We show which factors
influence this tradeoff. Based on our experiments, we present an adaptive
prediction model that supports the DBMS in deciding whether to utilize
these components. In addition, we evaluate non-coherent memory access
as an additional access method and discuss its benefits and shortcomings.

1 Introduction
Current in-memory databases are significantly limited by the main memory’s
latency and bandwidth [2]. In the time spent for transferring a cache line from
DRAM to the CPU (roughly 100 ns), a modern CPU can execute 300 instructions
or more. When the compute part of database operators executes in fewer cycles,
the CPU stalls and waits for more data to arrive. This gets exacerbated in NUMA
setups where remote DRAM accesses take roughly 200 ns with a single NUMA
hop. Scale-up systems, as used for big SAP HANA or Oracle databases, can
include multiple NUMA hops and up to 48 TB of memory. These connect up to
eight blades with four processors each to a single, cache-coherent network using
a proprietary interconnect. In such setups, memory latency from one end to the
other can reach hundreds of nanoseconds, making the influence even bigger.
Closely related to memory latency is memory bandwidth. On our test sys-
tem (cf. Sect. 3), we measured a NUMA node-local bandwidth of slightly over
50 GB/s, while remote accesses on the same blade had a reduced bandwidth of
12.5 GB/s and remote blades of 11.5 GB/s. As such, making good use of the
available physical bandwidth is vital. Doing so includes reducing the amount of
data transferred by using compression for a higher logical bandwidth (i.e., more
information transferred per byte) or organizing the data in a cache line-friendly
way. This could be a columnar table layout where each cache line only holds
values from the column that is accessed and cache line bycatch, i.e., data that is
loaded into the CPU but never used, is avoided for column store-friendly queries.
Making the DBMS more aware of NUMA can significantly improve the per-
formance [7]. By ensuring that data is moved across the NUMA network only
c Springer International Publishing AG, part of Springer Nature 2018
J. Wang et al. (Eds.): ADC 2018, LNCS 10837, pp. 3–14, 2018.
https://doi.org/10.1007/978-3-319-92013-9_1
4 M. Dreseler et al.

when it is unavoidable, the memory access costs can be reduced. Still, there
remain cases in which a load from a distant node cannot be avoided. This hap-
pens when joins access data from a remote table or when the data (and thus the
load) is imbalanced and an operator cannot be executed on the optimal node.
In addition to NUMA optimization and better data layouts, developers use
dedicated hardware to increase the effective bandwidth [1,6,9]. There are sev-
eral approaches, but no one-size-fits-all technique. In this paper, we look at one
specific method that is used to improve the physical bandwidth available to
database operators, namely the Global Reference Unit (GRU) built into sys-
tems like SGI’s UV series or HPE’s Superdome Flex. The GRU provides an API
that can be used to offload certain memory operations, allowing the CPU to
work on other data in the meantime. Previous work [3] has shown that this can
result in a performance benefit of up to 30% for table scans. We extend on this
by evaluating which factors lead to an advantage of the GRU over the CPU in
some cases and what causes it to be slower in others. This knowledge can be used
by the DBMS to automatically choose between the CPU and GRU access paths.
Furthermore, we present relaxed cache coherence as another access method.
This paper is organized as follows: Sect. 2 gives background information on
the hardware discussed in this paper. To gather data on the memory bus uti-
lization and better profile the physical properties of database operations, we use
performance counters as described in Sect. 3. These are then used in Sect. 4 to
discuss the factors that influence if one method or another gives the higher effec-
tive bandwidth. Section 5 explains how a DBMS can use these results in order
to choose an access method. In Sect. 6, we show how relaxing cache coherency
could further improve the physical bandwidth of a table scan. Related work is
discussed in Sect. 7 and a summary is given in Sect. 8.

2 Hardware Used for Accelerating DRAM Reads

In the previously mentioned scale-up systems, four NUMA nodes (i.e., proces-
sors) are grouped into blades as shown in Fig. 1. Each node is connected via
QPI links to other processors on the same blade, however the two diagonal QPI
connections are omitted. The free QPI port on each processor is then connected
to one of the two so-called HARPs that are part of each blade.
HARPs connect the entire scale-up system, as
Blade 1

H CPU CPU H
each HARP is directly connected with every other A QPI A
R R NUMAlink
using a special interconnect called NUMAlink. This P CPU CPU P
creates an all-to-all topology that allows the addi-
tion of more CPUs and more memory by attach- H CPU CPU H
Blade 2

A A
ing additional blades to the machine. In order to R
P CPU CPU
R
P
make the memory of one blade accessible to another
blade, the HARPs participate in the QPI ring of
their blades and mimic a NUMA node with a large Fig. 1. General architecture
amount of main memory, i.e., the memory of every of the discussed system
other blade [8].
 Adaptive Access Path Selection for Hardware-Accelerated DRAM Loads 5

The component in the HARP that is responsible for transparently extending

the memory space and translating memory addresses is the Global Reference
Unit (GRU). In addition to this transparent access, the GRU also provides a
user-level API that enables developers to instruct the GRU directly.
Previous work has shown that using the GRU API for database operations
improves throughput by up to 30% when scanning remote data [3]. For a more
in-depth discussion of the GRU hardware and the actual implementation of the
GRU scan, we refer to that paper. This performance advantage is achieved by
using the gru_bcopy function, a block copy instruction that is executed by the
GRU, and a double-buffered scan. We divide the data vector into equally sized
chunks, and allocate two temporary buffers of that chunk size. The asynchronous
bcopy operation is used to copy chunks from the data vector into one of the local
buffers, which is processed while bcopy is filling the other buffer. The buffers are
switched and the process is repeated until the entire vector is scanned.
There are three reasons why this can achieve better performance compared
to a regular CPU scan: Firstly, bcopy executes asynchronously and allows the
CPU to run other computations while the GRU handles the memory transfer.
As a result, the CPU can process one local chunk while the GRU is loading
the other chunk in the background. Secondly, the HARP can access remote
memory more efficiently than the CPU. This is because a CPU’s memory access
performance is limited by the number of outstanding read requests that it can
handle before it stalls and waits for loads to complete. Stalls are especially
noticeable in large NUMA systems with high memory access latencies, because it
takes longer for a read request to return and allow the next request to be issued.
This means that for the systems discussed here, an increase in latency results in a
decrease in bandwidth. The HARPs can handle more outstanding read requests
than a CPU and can therefore achieve higher throughput. Thirdly, because the
cache coherency directories (used to maintain a consistent view of data across
processors) are located on the HARP, we expect the HARPs to handle cache
coherency more efficiently. While this does not improve single operations, the
decreased cache coherency overhead can improve memory performance over time.

3 Quantifying the Memory Bandwidth Utilization

To better utilize the available memory bandwidth, it is vital to understand how

much of it is actually used and where the bottleneck is. This information is both
needed by developers of database operators and by the DBMS itself when it
decides which scan will be used. We use the Intel Processor Counter Monitor
(PCM) to monitor the QPI traffic on relevant nodes.
For every QPI link, PCM can measure the amount of both incoming data and
outgoing traffic. Incoming data (dataIn) only includes the transferred payload.
Outgoing traffic (trafficOut) includes both the payload as well as any overhead
such as cache coherency traffic. The incoming traffic or the outgoing data cannot
be retrieved from PCM, but can be computed in most cases. Also, the informa-
tion from where and to where data flows is not available.
6 M. Dreseler et al.

CPU CPU CPU

GRU GRU GRU

Fig. 2. QPI traffic when scanning a 4 GB vector of integers

Figure 2 displays the QPI traffic as measured1 by PCM for a scan on a 4 GB

vector of integers, i.e., approximately one billion integers. In the benchmark, the
same scan is executed three times, each time with a different memory placement
relative to the executing thread. After each run, the CPU caches are flushed
using CLFLUSH. The data is always located on NUMA node 0 (“storage node”),
while the executing thread is pinned to nodes 0, 2, and 7 (“execution node”) for
the node-local, blade-local, and off-blade cases respectively. For the local scan
using the CPU, no QPI traffic is seen because the traffic is node-local. The GRU
implementation, on the other hand, copies data from the storage node to the
execution node even when these are identical. This explains why a GRU scan
is detrimental for local data. Continuing to blade-local CPU scans (top center),
we see two significant lines, one for outgoing traffic of the storage node, and one
for incoming data of the executing node. The difference between the two can be
attributed to the cache coherency overhead. When using the GRU, the overall
bandwidth is higher, resulting in a lower execution time. A similar image can
be seen for off-blade scans. Here, the bandwidth difference between CPU and
GRU is even more pronounced. The fact that Socket 2 appears in the graph is
surprising at first. We attribute it to additional caches that we have no control
over and take this as a reason to look into non-cache coherent loads in Sect. 6.
The same library can be used to get numbers on the current load on the
memory bus. We gather information about the QPI utilization of all relevant
nodes and feed it into the adaptive model described in Sect. 5.

1
All benchmarks were executed on an SGI UV 300H with 6 TB RAM and eight Intel
E7-8890 v2 processors. Our code was compiled with gcc 7.2 at -O3.
 Adaptive Access Path Selection for Hardware-Accelerated DRAM Loads 7

4 Factors that Influence the CPU/GRU Tradeoff

In this section, we describe how the throughput of the GRU scan is affected
by different factors and how deciding between GRU and CPU scans is only
possible by looking at a combination of these factors. We classify these factors
into two groups: Internal and external factors. The former are parameters that
come from the scan itself, such as the size of the table. The latter are unrelated
to the particular scan operator, for example the utilization of the system.
We use the table scan as an example for an operator with sequential access
patterns. Compared to operators such as the join, it is strictly memory-bound,
so the influences of improved memory bandwidth utilization are better to see.
Latency-bound operators cannot be improved by the bcopy approach.

4.1 Internal Influences

Data Size. One of the most important factors is the size of the scanned table.
Figure 3 shows how the throughput of different access methods is influenced by
the size of the table when accessing an off-blade table. Both approaches reach
their maximum throughput only when the table has a certain size. For GRU
scans, this happens when the data size reaches at least 60 MB. CPU scans deliver
the maximum throughput for smaller table sizes, approximately 1 MB.
This can be explained with fixed setup costs for the table scan as well as
having to wait for the first cache lines to arrive from a remote location. For the
GRU scan, the additional system calls required to obtain the execution contexts
and to execute the bcopy method mean that bigger tables are needed to reach
the break-even point.

16 M

Output Rows

Fig. 3. Influence of the input data size for off-blade CPU and GRU scans - each dot is
one measured data point

Data Locality. Depending on where the input data is located relative to the
executing CPU, it needs to be transferred through zero to multiple NUMA hops.
Figure 4 shows that the throughput for the regular CPU scan changes depending
on NUMA distance. The highest throughput of 8 GB/s is achieved on the same
node. With increasing NUMA distance, the throughput rates decrease. For a
blade-local scan, the throughput rates reach up to 5 GB/s, and scanning an off-
blade vector only nets approximately 3 GB/s. The GRU scan performance stays
8 M. Dreseler et al.

Fig. 4. Influence of the NUMA distance for CPU and GRU scans

stable for all NUMA distances at around 6 GB/s. For both CPU and GRU, a
high variance is measured. This is dependent on the other execution parameters
as described in this section. It shows that there are parameters other than the
data locality, especially for small tables, that play an important role in deciding
if the CPU or the GRU is faster.
For the model described in Sect. 5, we take the latency (instead of the number
of hops) between source and destination node as it describes a linear variable,
not a discrete one. This makes it easier to adapt the model to other systems
where the latency between hops is different.

Result Size. When scanning the input data vector, i.e., the column in an
in-memory database, the operation also needs to save the results. In this imple-
mentation, the scan returns a vector of indexes of the input vector where a
certain search value was found. This means that both the value distribution and
the given search value have an impact on how large the result gets. We have
chosen to take both the data size and the result size as parameters instead of
just using the selectivity. This is because the impact of the selectivity on the
scan cost varies for different data sizes.
Figure 5 shows the performance of scans with different result sizes. As the
output size grows, the amount of time spent for writing the output vector slowly
grows as well, and at some point, surpasses the value comparisons in terms of
runtime. Consequently, after that point the benefits gained from our improved
scanning method become insignificant.

Fig. 5. Influence of the output size when scanning an off-blade vector of 512 KiB
 Adaptive Access Path Selection for Hardware-Accelerated DRAM Loads 9

4.2 External Influences

Background QPI Traffic. In a production environment, the executed operator
is not the only memory-intensive operation on the system. Therefore, it makes
sense to also take a look at the influence of memory traffic caused by other
processes. To identify how this affects performance, we use PCM (as described
in Sect. 3) to measure QPI traffic right before executing our scan operations.
For the benchmark, we generate background traffic by using two types of
background workers. One copies data between the worker and the data node
and then scans the remote data with a SIMD scan on 32 bit integers. The
second worker uses bcopy to asynchronously copy data back and forth between
the worker and data node. This generates both QPI and Numalink traffic.
By doing so, we can vary the background traffic and measure the performance
for varying QPI loads. Our measurements show that the impact of high QPI
utilization is higher on CPU scans. If no parallel workers are consuming QPI
bandwidth, the throughput is unaffected. In the worst case, a busy QPI interface
decreases the throughput by 2 GB/s. For the GRU, scans on a high-load system
only have a throughput that is 1 GB/s lower.

Background HARP Traffic. Because the GRU is responsible for referencing

memory of other blades, it is involved in memory accesses even when the GRU
API is not used. Consequently, the load of the GRUs has an effect on other
memory operations in the system as shown in Fig. 6. Different from previous
figures, all combinations of data sizes and data locality are combined in the
graph. For GRU scans, the scan throughput stays in a small window just below
7 GB/s when the GRU is not busy. When other explicit GRU operations are
run simultaneously, the throughput is reduced significantly. For the CPU scans,
background GRU traffic does not affect the maximum throughput as much.

Fig. 6. Influence of a GRU background traffic for vectors bigger than 32 MB

5 Self-tuning Access Path Selection

The measurements presented in Sect. 4 have shown that directly utilizing the
GRU has performance benefits in some situations. In others, such as with local
data, small tables, or a high number of output results, the CPU is preferred. We
Other documents randomly have
different content
 lifting up the lowly, will be spoken of as a memorial of her in all
the churches; and in not a few homes, of the rich as of the
poor, will be felt the sweet constraint of her beautiful, joyous,
consecrated life. She was not alone; there are many more like
her; and, best of all, there are to be vastly more yet, who will
not be deaf to ‘the high calling.’ The Master has need of them.
The way, on the whole, is infinitely attractive. Thanks for the life
of this woman who did so much, from first to last, to make it
appear so!
“And thanks too for such a death, which, coming in the
sweetest and completest blooming of life’s beauty, when not a
fault had stayed to mar it, and no wasting had ever touched it—
an ending which transfigures all that came before it, and which
now, in the mingling of retrospect and prospect, helps those
who knew her to a deeply surprised sense of the fact that,

‘To Death it is given,

To see how this world is embosomed in heaven.’”

To us, who are blind and cannot see afar off, it is impossible to
perceive, and difficult to believe, that the taking away in the vigor of
womanhood of one who was showing such a capacity and
adaptability for the work of elevating the Teetons can be made to
subserve the furtherance of the cause of Christ. But we must believe
that God, who sees the end from the beginning, and who makes no
mistakes, will bring out of this sore bereavement a harvest of joy;
and that that grave under the window of the mission house in Peoria
bottom will be a testimony to the love of Jesus and the power of his
Gospel, that will thrill and uplift many hearts from Bangor to Fort
Sully. It was a beautiful life of faith and service; and it has only gone
to be perfected in the shadow of the Tree of Life.
S. R. R.
 REV. GIDEON H. POND.
A SUCCESSFUL LIFE.

Born and brought up in Litchfield county, in a town adjoining

Washington, Connecticut, Rev. George Bushnell visited that hill
country in his youth, and was deeply impressed with the manifest
and pervading religious element in the community. Taken there by a
special providence, more than a quarter of a century ago, and
enjoying the privilege of a visit in some of the families, it seemed to
me that it had been a good place to raise men. This was on the line
of the impression made upon me years before that. When I first
met, in the land of the Dakotas, the brothers Samuel W. and Gideon
H. Pond, they were both over six feet high, and “seemed the
children of a king.”
In this hill town of Washington, on the 30th of June, 1810,
Gideon Hollister, the younger of the two brothers, was born. His
parents were Elnathan Judson and Sarah Hollister Pond. Gideon was
the fifth child, and so was called by the Dakotas Hakay. Of his
childhood and youth almost nothing is known to the writer. He had
the advantage of a New England common-school education; perhaps
nothing more. As he grew very rapidly and came to the size and
strength of man early, he made a full hand in the harvest field at the
age of sixteen. To this ambition to be counted a man and do a man’s
work when as yet he should have been a boy, he in after life
ascribed some of his infirmities. This ambition continued with him
through life, and occasional over-work at last undermined a
constitution that might, with care and God’s blessing, have continued
to the end of the century.
He came to the land of the Dakotas, now Minnesota, in the
spring of 1834. The older brother, Samuel, had come out as far as
Galena, Ill., in the summer previous. The pioneer minister of that
country of lead was Rev. Aratus Kent, who desired to retain Mr. Pond
as an adjutant in his great and constantly enlarging work; but Mr.
Pond had heard of the Sioux, or Dakotas, for whose souls no one
cared, and, having decided to go to them, he sent for his brother
Gideon to accompany him.
When they reached Fort Snelling, and made known their errand
to the commanding officer of the post, Major Bliss, and to the
resident Indian agent, Major Taliaferro, they received the hearty
approval and co-operation of both, and the agent at once
recommended them to commence work with the Dakotas of the
Lake Calhoun village, where some steps had already been taken in
the line of civilization. There, on the margin of the lake, they built
their log cabin. Last summer Mr. King’s grand Pavilion, so called, was
completed on the same spot, which gave occasion for Mr. Gideon H.
Pond to tell the story of this first effort in that line:

“Just forty-three years previous to the occurrence above

alluded to, on the same beautiful site, was completed an
humble edifice, built by the hands of two inexperienced New
England boys, just setting out in life-work. The foundation-
stones of that hut were removed to make place for the present
Pavilion, perchance compose a part of it. The old structure was
of oak logs, carefully peeled. The peeling was a mistake. Twelve
feet by sixteen, and eight feet high, were the dimensions of the
edifice. Straight poles from the tamarack grove west of the lake
formed the timbers of the roof, and the roof itself was of the
bark of trees which grew on the bank of what is now called
‘Bassett’s Creek,’ fastened with strings of the inner bark of the
bass wood. A partition of small logs divided the house into two
rooms, and split logs furnished material for a floor. The ceiling
was of slabs from the old government saw-mill, through the
kindness of Major Bliss, who was in command of Fort Snelling.
The door was made of boards split from a log with an axe,
having wooden hinges and fastenings, and was locked by
pulling in the latch-string. The single window was the gift of the
kind-hearted Major Lawrence Taliaferro, United States Indian
 agent. The cash cost of the building was one shilling, New York
currency, for nails used in and about the door. ‘The formal
opening’ exercises consisted in reading a section from the old
book by the name of Bible, and prayer to Him who was its
acknowledged author. The ‘banquet’ consisted of mussels from
the lake, flour and water. The ground was selected by the
Indian chief of the Lake Calhoun band of Dakotas, Man-of-the-
sky, by which he showed good taste. The reason he gave for
the selection was that ‘from that point the loons would be
visible on the lake.’
“The old chief and his pagan people had their homes on the
surface of that ground in the bosom of which now sleep the
bodies of deceased Christians from the city of Minneapolis, the
Lake Wood cemetery, over which these old eyes have
witnessed, dangling in the night breeze, many a Chippewa
scalp, in the midst of horrid chants, yells, and wails, widely
contrasting with the present stillness of that quiet home of
those

‘Who sleep the years away.’

That hut was the home of the first citizen settlers of

Hennepin county, perhaps of Minnesota, the first school-room,
the first house for divine worship, and the first mission station
among the Dakota Indians.”

The departure of Mr. Pond called forth from Gen. Henry H.

Sibley so just and beautiful a tribute, that I can not forbear inserting
a portion, from the Pioneer Press of St. Paul:—

“When the writer came to this country, in 1834, he did not

expect to meet a single white man, except those composing the
garrison at Fort Snelling, a few government officials attached to
the department of Indian affairs, and the traders and voyageurs
employed by the great fur company in its business. There was
but one house, or, rather, log cabin, along the entire distance of
 nearly 300 miles between Prairie du Chien and St. Peters, now
Mendota, and that was at a point below Lake Pepin, near the
present town of Wabashaw. What was his surprise then to find
that his advent had been preceded in the spring of the same
year by two young Americans, Samuel W. Pond and Gideon H.
Pond, brothers, scarcely out of their teens, who had built for
themselves a small hut at the Indian village of Lake Calhoun,
and had determined to consecrate their lives to the work of
civilizing and Christianizing the wild Sioux. For many long years
these devoted men labored in the cause, through manifold
difficulties and discouragements, sustained by a faith that the
seed sown would make itself manifest in God’s good time. The
efforts then made to reclaim the savages from their mode of
life, the influence of their blameless and religious walk and
conversation upon those with whom they were brought in daily
contact, and the self-denial and personal sacrifices required at
their hands, are doubtless treasured up in a higher than human
record.”

General Sibley mentions an incident belonging to this period of

their residence at Lake Calhoun, which never before came to my
knowledge:—

“Gifted with an uncommonly fine constitution, the subject

of this sketch met with an accident in his early days, from the
effects of which it is questionable if he ever entirely recovered.
He broke through the ice at Lake Harriet in the early part of the
winter, and as there was no one at hand to afford aid, he only
saved his life after a desperate struggle, by continuing to
fracture the frozen surface until he reached shallow water, when
he succeeded in extricating himself. His long immersion and
exhaustive efforts brought on a severe attack of pneumonia,
which for many days threatened a fatal termination.”

My own personal acquaintance with Mr. Pond commenced in the

summer of 1837. He was then, and had been for a year previous, at
Lac-qui-parle. In September my wife and I joined that station, and
the first event occurring after that, which has impressed itself upon
my memory, was the marriage of Mr. Pond and Miss Sarah Poage,
sister of Mrs. Dr. Williamson. This was the first marriage ceremony I
had been called upon to perform; and Mr. Pond signalized it by
making a feast, and calling, according to the Saviour’s injunction,
“the poor, the maimed, the halt, and the blind.” And there was a
plenty of such to be called in that Dakota village. They could not
recompense him, but “he shall be recompensed at the resurrection
of the just.”
Mr. Pond had long been yearning to see what was inside of an
Indian. He sometimes said he wanted to be an Indian, if only for a
little while, that he might know how an Indian felt, and by what
motives he could be moved. When the early spring of 1838 came,
and the ducks began to come northward, a half-dozen Dakota
families started from Lac-qui-parle to hunt and trap on the upper
part of the Chippewa River, in the neighborhood of where the town
of Benson now is. Mr. Pond went with them and was gone two
weeks. It was in the month of April, and the streams were flooded
and the water was cold. There should have been enough of game
easily obtained to feed the party. But it did not prove so. A cold spell
came on, the ducks disappeared, and Mr. Pond and his Indian
hunters were reduced to scanty fare, and sometimes they had
nothing for a whole day. But Mr. Pond was seeing inside of Indians
and was quite willing to starve a good deal. However, his stay with
them, and their hunt for that time as well, was suddenly terminated,
by the appearance of the Ojibwa chief Hole-in-the-Day and ten men
with him. They came to smoke the peace-pipe, they said. They were
royally feasted by three of the families, who killed their dogs to feed
the strangers, who, in turn, arose in the night and killed the
Dakotas. As God would have it, Mr. Pond was not then with those
three tents, and so he escaped.
No one had started with more of a determination to master the
Dakota language than Gideon H. Pond. And no one of the older
missionaries succeeded so well in learning to talk just like a Dakota.
Indeed, he must have had a peculiar aptitude for acquiring
language; for in these first years of missionary life, he learned to
read French and Latin and Greek, so that the second Mrs. Pond
writes: “When I came, and for a number of years, he read from the
Greek Testament at our family worship in the morning. Afterward he
used his Latin Bible, and still later his French Testament.”
In this line of literary work General Sibley’s testimony is
appreciative. He says:—

“Indeed, to them, and to their veteran co-laborers, Rev. T.

S. Williamson and Rev. S. R. Riggs, the credit is to be ascribed
of having produced this rude and rich Dakota tongue to the
learned world in a written and systematic shape, the lexicon
prepared by their joint labors forming one of the publications of
the Smithsonian Institute at Washington City, which has justly
elicited the commendation of experts in philological lore, as a
most valuable contribution to that branch of literature.”

While Mr. Pond was naturally ambitious, he was also peculiarly

sensitive and retiring. When the writer was left with him at Lac-qui-
parle, Dr. Williamson having gone to Ohio for the winter, although so
much better master of the Dakota than I was at that time, he was
unwilling to take more than a secondary part in the Sabbath
services. “Dr. Williamson and you are ministers,” he would say. And
even years afterward, when he and his family had removed to the
neighborhood of Fort Snelling, and he and his brother had built at
Oak Grove, with the people of their first love, Gideon H. could hardly
be persuaded that it was his duty to become a preacher of the
Gospel. I remember more than one long conversation I had with him
on the subject. He seemed to shrink from it as a little child, although
he was then thirty-seven years old.
In the spring of 1847, he and Mr. Robert Hopkins were licensed
by the Dakota presbytery, and ordained in the autumn of 1848. We
were not disappointed in our men. Mr. Hopkins gave evidence of
large adaptation to the missionary work; but in less than three years
he heard the call of the Master, and went up through a flood of
waters. Mr. Pond, notwithstanding his hesitation in accepting the
office, became a most acceptable and efficient and successful
preacher and pastor.
After the treaties of 1851, these Lower Sioux were removed to
the Upper Minnesota. White people came in immediately and took
possession of their lands. Mr. Pond elected to remain and labor
among the white people. He very soon organized a church, which in
a short time became a working, benevolent church—for some years
the banner Presbyterian church of Minnesota in the way of
benevolence. When, in 1873, Mr. Pond resigned his pastorate, he
wrote in his diary, “I have preached to the people of Bloomington
twenty years.” He received home mission aid only a few years.
We are very glad to have placed at our disposal so much of the
private journal of the late Rev. G. H. Pond as relates to the
wonderful work of God among the Dakotas in prison at Mankato,
Minn., in the winter of 1862-63. The facts, in the main, have been
published before; but the story, as told so simply and graphically by
Mr. Pond, may well bear repeating. Mr. Pond arrived at Mankato
Saturday, January 31, 1863, and remained until the afternoon of
Tuesday, February 3:—

“There are over three hundred Indians in prison, the most

of whom are in chains. There is a degree of religious interest
manifested by them, which is incredible. They huddle
themselves together every morning and evening in the prison,
and read the Scriptures, sing hymns, confess one to another,
exhort one another, and pray together. They say that their
whole lives have been wicked—that they have adhered to the
superstitions of their ancestors until they have reduced
themselves to their present state of wretchedness and ruin.
They declare that they have left it all, and will leave all forever;
that they do and will embrace the religion of Jesus Christ, and
adhere to it as long as they live; and that this is their only hope,
both in this world and in the next. They say that before they
came to this state of mind—this determination—their hearts
failed them with fear, but now they have much mental ease and
comfort.
“About fifty men of the Lake Calhoun band expressed a
wish to be baptized by me, rather than by any one else, on the
ground that my brother and myself had been their first and
chief instructors in religion. After consultation with Rev. Marcus
Hicks of Mankato, Dr. Williamson and I decided to grant their
request, and administer to them the Christian ordinance of
baptism. We made the conditions as plain as we could, and we
proclaimed there in the prison that we would baptize such as
felt ready heartily to comply with the conditions—commanding
that none should come forward to receive the rite who did not
do it heartily to the God of heaven, whose eye penetrated each
of their hearts. All, by a hearty—apparently hearty—response,
signified their desire to receive the rite on the conditions
offered.
“As soon as preparations could be completed, and we had
provided ourselves with a basin of water, they came forward,
one by one, as their names were called, and were baptized into
the name of the Father, the Son, and the Holy Ghost, while each
subject stood with his right hand raised and head bowed, and
many of them with the eyes closed, with an appearance of
profound reverence. As each one passed from the place where
he stood to be baptized, one or the other of us stopped him and
addressed to him, in a low voice, a few words, such as our
knowledge of his previous character and the solemnities of the
occasion suggested. The effect of this, in most cases, seemed to
very much deepen the solemnity of the ceremony. I varied my
words, in this part of the exercises, to suit the case of the
person; and when gray-haired medicine-men stood literally
trembling before me, as I laid one hand on their heads, the
effect on my mind was such that at times my tongue faltered.
The words which I used in this part of the service were the
following, or something nearly like them in substance: ‘My
brother, this is the mark of God which is placed upon you. You
will carry it while you live. It introduces you into the great family
of God, who looked down from heaven, not upon your head, but
into your heart. This ends your superstition, and from this time
you are to call God your Father. Remember to honor him. Be
resolved to do his will.’ It made me glad to hear them respond,
‘Yes, I will.’
“When we were through, and all were again seated, we
sung a hymn appropriate to the occasion, in which many of
them joined, and then prayed. I then said to them, ‘Hitherto I
have addressed you as friends; now I call you brothers. For
years we have contended together on this subject of religion;
now our contentions cease. We have one Father—we are one
family. I must now leave you, and probably shall see you no
more in this world. While you remain in this prison, you have
time to attend to religion. You can do nothing else. Your
adherence to the Medicine Sack and the Wotawe has brought
you to ruin. Our Lord Jesus Christ can save you. Seek him with
all your heart. He looks not on your heads nor on your lips, but
into your bosoms. Brothers, I will make use of a term of
brotherly salutation, to which you have been accustomed in
your medicine dance, and say to you, Brothers, I spread my
hands over you and bless you.’ The hearty answer of three
hundred voices made me feel glad.
“The outbreak and events which followed it have, under
God, broken into shivers the power of the priests of devils,
which has hitherto ruled these wretched tribes. They were
before bound in the chains and confined in the prison of
Paganism, as the prisoners in the prison at Philippi were bound
with chains. The outbreak and its attendant consequences have
been like the earthquake to shake the foundation of their
prison, and every one’s bonds have been loosed. Like the jailer,
in anxious fear they have cried, ‘Sirs, what must we do to be
saved?’ They have been told to believe on the Lord Jesus Christ,
who will still save unto the uttermost all that come unto God by
him. They say they repent and forsake their sins—that they
 believe on him, that they trust in him, and will obey him.
Therefore they have been baptized into the name of the Father
and of the Son and of the Holy Ghost, three hundred in a day.”

In the spring of 1853, Mrs. Sarah Poage Pond departed, after a

lingering illness of eighteen months, and left a “blessed memory.”
There were seven children by this marriage, all of whom are living,
and have families of their own, but George, who died while in the
Lane Theological Seminary. In the summer of 1854, Mr. Pond was
married to his second wife, Mrs. Agnes C. J. Hopkins, widow of Rev.
Robert Hopkins. The second Mrs. Pond brought her three children,
making the united family of children at that time ten. Six have been
added since. And there are twenty-two grandchildren, six of whom
are members of the Church of Christ, together with all the children
and their companions. Is not that a successful life? Counting the
widowed mother and those who have come into the family by
marriage, there are, I understand, just fifty who mourn the
departure of the patriarch father. A little more than two-score years
ago, he was one; and now behold a multitude!
Mary Frances Hopkins, who came into the family when a girl,
and afterward married Edward R. Pond, the son, writes thus: “To me
he was as near an own father as it is possible for one to be who is
so by adoption, and I shall always be glad I was allowed to call him
father.”
The members of the synod of Minnesota will remember with
great pleasure Mr. Pond’s presence with them at their last meeting at
St. Paul, in the middle of October. For some years past, he has
frequently been unable to be present. This time he seemed to be
more vigorous than usual, and greatly entertained the synod and
people of St. Paul with his terse and graphic presentation of some of
the Lord’s workings in behalf of the Dakotas.
During the meeting I was quartered with Mrs. Governor
Ramsay. On Saturday I was charged with a message to Mr. Pond,
inviting him to come and spend the night at the governor’s. We
passed a profitable evening together, and he and I talked long of the
way in which the Lord had led us; of the great prosperity he had
given us in our families and in our work. Neither of us thought,
probably, that that would be our last talk this side the golden city.
The next day, Sabbath, he preached in the morning, for Rev. D. R.
Breed, in the House of Hope, which, probably, was his last sermon.
In the evening he was with us in the Opera House, at a meeting in
the interest of home and foreign missions.
“His health gradually failed,” Mrs. Pond writes, “from the time of
his return from synod, though he did not call himself sick until the
11th of January, 1878, and he died on Sabbath, the 20th, about
noon.” She adds: “His interest in the Indians, for whom he labored
so long, was very deep, and he always spoke of them with loving
tenderness, and often with tears. One of the last things he did was
to look over his old Dakota hymns, revised by J. P. W. and A. L. R.,
and sent to him for his consent to the proposed alterations.”
“His simple faith in the Lord Jesus caused him all the time to live
a life of self-denial, that he might do more to spread the knowledge
of Jesus’ love to those who knew it not.” The love of Christ
constrained him, and was his ruling passion.
Of his last days the daughter says:—
“He really died of consumption. The nine days he was confined
to bed he suffered much; but his mind was mostly clear, and he was
very glad to go. I think the summons was no more sudden to him
than to Elijah. He was to the last loving and trustful, brave and
patient. To his brother Samuel, as he came to his sick bed, he said:
‘So we go to see each other die.’ Some time before he had visited
Samuel when he did not expect to recover. ‘My struggles are over.
The Lord has taken care of me, and he will take care of the rest of
you. My hope is in the Lord,’ he said.
“Toward the last it was hard for him to converse, and he bade
us no formal farewell. But the words, as we noted them down, were
words of cheer and comfort: ‘You have nothing to fear, for the
present or the future.’ And so was given to him the victory over
death, through faith in Jesus.”
 Is that dying? He sleeps with his fathers. He has gone to see
the King in his beauty, in a land not very far off.
As loving hands ministered to him in his sickness, loving hearts
mourned at his death. On the Wednesday following he was buried. A
half a dozen brothers in the ministry were present at his funeral,
and, fittingly, Mr. Breed of the House of Hope preached the sermon.
This is success.
S. R. R.
 SOLOMON.
In the summer of 1874 Rev. John P. Williamson made a tour up
the Missouri River as far as Fort Peck. His judgment was that there
was no opening at that place for the establishment of a new mission,
but that something might possibly be done by native Dakotas. In the
meantime, we had heard from the regions farther north than Fort
Peck, where some of our church-members had gone after the
outbreak of 1862. Somewhere up in Manitoba, near Fort Ellice, was
Henok Appearing Cloud, with his relatives. His mother, Mazaskawin,
—Silver-Woman,—was a member of the Hazelwood church, and his
father, Wamde-okeya,—Eagle Help,—had been my old helper in
Dakota translations. These were all near relatives of Solomon
Toonkanshaecheye, one of our native pastors.
Dr. Williamson, by correspondence with the Presbyterian Board,
obtained an appropriation of several hundred dollars to send a native
missionary to these Dakotas in Canada. Solomon gladly accepted the
undertaking, and in the month of June, 1875, started for Manitoba
with Samuel Hopkins for a companion.
They were received with a great deal of joy by their friends,
who entreated them to stay, or come back again if they left. But
provisions were very scarce, and hard to be obtained; and hence
they determined to return to the Sisseton agency before winter.
While in Manitoba they had taught and preached the Gospel, and
baptized and received several persons to the fellowship of the
church. Solomon wrote, before he returned, “Indeed, there is no
food; they have laid up nothing at all; so that, when winter comes,
where they will obtain food, and how they will live, no one knows.
But I have already found something of what I have been seeking,
and very reluctantly I turn away from the work.”
Solomon and Samuel returned to Sisseton, but their visit had
created a larger desire for education and the privileges of the
Gospel. In the March following, Henok Appearing Cloud wrote that
he had taught school during the winter, and conducted religious
meetings, as he “wanted the Word of God to grow.” In much
simplicity, he adds: “Although I am poor, and often starving, I keep
my heart just as though I were rich. When I read again in the
Sacred Book what Jesus, the Lord, has promised us, my heart is
glad. I am thinking, if a minister will only come this summer and
stay with us a little while, our hearts will rejoice. If he comes to stay
with us a long time, we will rejoice more. But as we are so often in a
starving condition, I know it will be hard for any one to come.”
Rev. John Black of Keldonan Manse, near Winnipeg, heard of
this visit of Solomon to Manitoba, and of the desire of those Dakotas
to have a missionary. He at once became deeply interested in the
movement, and wrote to Dr. Williamson, at St. Peter, proposing that
the Presbyterian Missionary Society of Canada should take upon
themselves the charge of supporting Solomon as a missionary
among the Dakotas of the Dominion. But when the matter was
brought before the missionary committee, they decided that the
condition of their finances would not allow them to add to their
burdens at that time. It was not, however, given up, and a year later
the arrangement was consummated. In the Word Carrier for
December, 1877, appeared this editorial:—

“The most important event occurring in our missionary

work during the month of October is the departure of Rev.
Solomon Toonkanshaecheye, with his family, for Fort Ellice, in
the Dominion of Canada. This has been under advisement by
the Presbyterian Foreign Missionary Society of Canada for two
years past. Rev. John Black of Keldonan Manse, Manitoba, has
been working for it. A year ago the funds of the society would
not admit of enlargement in their operations. This year their
way has been made clear, and the invitation has come to
Solomon to be their missionary among the Dakotas on the
Assinaboine River. They pay his expenses of removal, and
promise him $600 salary.
 “He has gone. Agent Hooper of Sisseton agency furnished
him with the necessary pass, and essentially aided him in his
outfit, and so we sent him off on the tenth day of October,
invoking God’s blessing upon him and his by the way, and
abundant success for him in his prospective work. From the
commencement of negotiations in regard to this matter it has
been of special interest to Dr. T. S. Williamson of St. Peter. He
has conducted the correspondence with Mr. Black. And now,
while the good doctor was lying nigh unto death, as he
supposed, the arrangement has gone into effect. If this prove to
be his last work on earth (may the good Lord cause otherwise),
it will be a matter of joy on his part that thus the Gospel is
carried to regions beyond, by so good and trustworthy a man as
we have found Solomon to be all through these years.”

Thus was the work commenced. Dr. Williamson did not pass
from us then, but lived nearly two years longer, and was cheered by
the news of progress in this far-off land. This being among our first
efforts to do evangelistic work by sending away our native ministers,
our hearts were much bound up in it. The church of Long Hollow
was reluctant to give up their pastor, and to me it was giving up one
whom I had learned to trust, and, in some measure, to depend
upon, among my native pastors. But it was evidently God’s call, and
he has already justified himself, even in our eyes. Solomon found a
people prepared of the Lord, and, in the summer of 1878, he reports
a church organized with thirteen members, which they named Paha-
cho-kam-ya—Middle Hill—of which Henok was elected elder.
In the next winter Solomon and Henok made a missionary tour
of some weeks, of which we have the following report. The letter is
dated “February 22, 1879, at Middle Hill, near Fort Ellice, North-west
Territory”:—

“This winter it seemed proper that I should visit the

Dakotas living in the extreme settlements, to proclaim to them
the Word of God. I first asked counsel of God, and prayed that
he would even now have mercy on the people of these end
villages, and send his Holy Spirit to cause them to listen to his
Word. Then I sent word to the people that I was coming.
“Then I started with Mr. Enoch, my elder. The first night we
came to three teepees of our own people at Large Lake, and
held a meeting with them. The next morning we started, and
slept four nights. On the fifth day we came to a large
encampment on Elm River. There were a great number of tents,
which we visited, and prayed with them, being well received.
But as I came to where there were two men, and prayed with
them, I told them about him whose name was Jesus—that he
was the Helper Man, because he was the Son of God. That he
came to earth, made a sacrifice of himself, and died, that he
might reconcile all men to God; that he made himself alive
again; that, although men have destroyed themselves before
God, whosoever knows the meaning of the name of Jesus, and
fears for his own soul, and prays, he shall find mercy, and be
brought near to God. That is the Name. And he is the Saviour of
men, and so will be your Saviour also, I said.
“Then one of them in a frightened way answered me: ‘I
supposed you were a Dakota, of those who live in cabins. It is
not proper that you should say these things. As for me, I do not
want them. Those who wish may follow in that way; but I will
not. You who hold such things should stay at home. What do
you come here for?’
“Walking-nest then said: ‘You are Cloudman’s son, I
suppose, and so you are my cousin. Cousin, when we first came
to this country there was a white minister who talked to us and
said: “Your hands are full of blood; therefore, when your hands
become white, we will teach you.” So he said, and when you
brought a book from the south, while they were looking at it,
blood dropped from above upon it; and behold, as the white
minister said, I conclude we are not yet good. Therefore, my
cousin, I am not pleased with your coming,’ he said.
“But there were only two men who talked in this way. We
left them and visited every house in the camp. Many may have
felt as those men did, but did not say it openly. The men said
they were glad, and welcomed us into their tents.
“The next day I came into a sick man’s tent whose name
was Hepan, lying near to death. I talked with him, and prayed
to God for him. Then he told me how he longed to hear from his
friends down south, and mentioned over half a dozen names of
his relatives. A woman also, who was present, said: ‘I want to
know if my friends are yet living.’
“Then we continued our visiting from house to house.
Sometimes we found only children in the tent; sometimes there
were men and women, and I prayed with them and told them a
word of Jesus. So we came to the teepees in the valley. Then I
met Iron Buffalo. There we spent the Sabbath, and held
meeting, having twenty-three persons present. A chief man,
whose name is War-club-maker, called them together.
“Our meetings there being finished, we departed and came
to the Wahpaton village. They were making four sacred feasts.
We did not go into them. But, visiting other houses, we passed
on about five miles, when night came upon us. Still we went on
to the end of the settlement, where we held a meeting. The
teepee was small, but there I found a sick man who listened to
the Word. This was Chaskay, the son of Taoyatedoota. He said
he was going to die, and from what source he should hear any
word of prayer, or any comforting word of God, was not
manifest. But now he had heard these things, and was very
glad, he said. This way was the best upon earth, and he
believed in it now. So, while we remained there, he wanted us
to pray with and for him, he said.
“We spent one day there, and the second day we started
home, and came to Hunka’s tent, and so proceeded homeward.
When we had reached the other end of the settlement, we
learned that the white ministers were to hold a meeting of
presbytery. They sent word to us to come, and so in the night,
with my Hoonkayape, Mr. Enoch, I went back. They asked us to
give an account of our missionary journey among the Dakotas.
 And so we told them where we had been and what we had
done. Also, we gave an account of things at Middle Hill, where
we live. When we had finished, they all clapped their hands.
Then they said they wanted to hear us sing a hymn of praise to
God in Dakota. We sang ‘Wakantanka Towaste,’ and at the close
they clapped their hands again.
“Then two men arose, one after the other. The first said: ‘I
have not expected to see such things so soon among the
Dakotas. But now I see great things, which I like very much.’
The other man spoke in the same way.
“Men and women had come together in their prayer-house,
and so there was a large assembly.
“Then the minister of that church arose and said: ‘White
people, who have grown up hearing of this way of salvation, are
expected to believe in it, and I have been accustomed to rejoice
in the multiplication of the Christian church; but I rejoice more
over this work among the Dakotas.’”

Both of these men came home to watch and wait by the sick-
bed of dear children. Nancy Maza-chankoo-win,—Iron Road Woman,
—the daughter of Henok, died April 28, 1879. She was thirteen years
old, read the Dakota Bible well, and was quite a singer in the prayer
assemblies. They say: “We all thought a great deal of her; but now
she too has gone up to sing in the House of Jesus, because she was
called.”
From Middle Hill, near Fort Ellice in Manitoba, comes a letter
written on May 20 by our friend Solomon. He reports seven
members added by profession of faith to his church in April, and ten
children baptized. There, as here, the season has been a sickly one,
and many deaths have occurred. For three months he has had
sickness in his own family. His story is pathetic. “Now,” he says, “my
son Abraham is dead. Seven years ago, at Long Hollow, in the
country of the Coteau des Prairies, he was born on January 12,
1872. And on the 23d of June following, at a communion season at
Good Will Church, he was baptized. When Mr. Riggs poured the
water on him, he was called Abraham. And then in the country of
the north, from Middle Hill, May 9, 1879, on that day, his soul was
carried home to the House of Jesus.
“Five months after he was born, I wanted to have him baptized.
I always remember the thought I had about it. Soon after a child is
born, it is proper to have it baptized. I believed that baptism alone
was not to be trusted in, and when one is baptized now it is finished
is not thinkable. But in Luke 18:16, our Lord Jesus says: ‘Suffer the
little children to come unto me’; and so taking them to Jesus is
good, since his heart is set on permitting them to come. Therefore, I
wanted this my son to go to Jesus.
“And so from the time he could hear me speak, I have
endeavored to train him up in all gentleness and obedience, in truth
and in peace. Now, for two years in this country he has been my
little helper. When some could not say their letters, he taught them.
He also taught them to pray. And when any were told to repeat the
commandments, and were ashamed to do so, he repeated them
first, for he remembered them all. Hence, I was very much attached
to him. But this last winter he was taken sick, and from the first it
seemed that he would not get well. But while he lived it was possible
to help him, and so we did to the extent of our ability. He failed
gradually. He was a long time sick. But he was not afraid to die. He
often prayed. When he was dying, but quite conscious of everything
that took place, then he prayed, and we listened. He repeated the
prayer of the Lord Jesus audibly to the end. That was the last voice
we heard from him. Perhaps when our time comes, and they come
for us to climb up to the hill of the mountain of Jehovah, then we
think we shall hear his new voice. Therefore, although we are sad,
we do not cry immoderately.”
That was a beautiful child-life, and a beautiful child-death. Who
shall say there are not now Dakota children in heaven? To have been
the means, under God, of opening in this desert such a well of faith
and salvation is quite a sufficient reward for a lifetime of work.
S. R. R.
Welcome to our website – the ideal destination for book lovers and
knowledge seekers. With a mission to inspire endlessly, we offer a
vast collection of books, ranging from classic literary works to
specialized publications, self-development books, and children's
literature. Each book is a new journey of discovery, expanding
knowledge and enriching the soul of the reade

Our website is not just a platform for buying books, but a bridge
connecting readers to the timeless values of culture and wisdom. With
an elegant, user-friendly interface and an intelligent search system,
we are committed to providing a quick and convenient shopping
experience. Additionally, our special promotions and home delivery
services ensure that you save time and fully enjoy the joy of reading.

Let us accompany you on the journey of exploring knowledge and

personal growth!

textbookfull.com