Tech Blog

Information Technology / Machine Learning / Data Analysis / Big Data / System Integration

How to Run Kdd2013AuthorPaperIdentification Benchmark

kddcup PostgreSQL yum pip Linux

What is Kdd2013AuthorPaperIdentification?

KDD Cup is the well-known data mining competition of the annual ACM SIGKDD International Conference on Knowledge Discovery and Data Mining.
KDD Cup 2013 -> https://www.kaggle.com/c/kdd-cup-2013-author-paper-identification-challenge

How to Build Benchmark Environment?

OS:
- CentOS-6.4-x86_64

Add exclude to /etc/yum.repos.d/CentOS-Base.repo file [base] and [updates] sections to exclude PostgreSQL Packages:

[base]
...
exclude=postgresql*
 
[updates]
...
exclude=postgresql*

Install PostgreSQL 9.2 and others:

# Only PostgreSQL 9.2 can store dataRev2.postgres
rpm -Uvh http://yum.postgresql.org/9.2/redhat/rhel-6-x86_64/pgdg-centos92-9.2-6.noarch.rpm
yum install postgresql92 postgresql92-server postgresql92-contrib postgresql92-libs postgresql92-devel

# Python installation
yum install python
yum install python-devel

# Numpy installation
yum install numpy

# Scipy installation
yum install scipy

# g++ installation
yum install gcc-c++

# pip installation
yum install pip

# psycopg2 installation (not available by yum)
pip install psycopg2

# scikit-learn installation (not available by yum)
pip install -U scikit-learn

How to Run Benchmark?

See below:

https://github.com/benhamner/Kdd2013AuthorPaperIdentification

References