Joined: 19 Sep 2002
|Posted: Wed Nov 19, 2003 2:40 pm Post subject: Book Review - Investigative Data Mining for Security
Investigative Data Mining for Security and Criminal Detection
Author: Jesus Mena
Publisher: Butterworth Heinemann (Direct to book)
Book Specifications: Softcover, 452 Pages
Line Illustrations: 20
Measurements: 178 X 235 mm
Publication Date: 17 February 2003
Suggested Publisher Price: £34.99 / $49.99
User Level: Non-Technical
Amazon.co.uk: Investigative Data Mining for Security and Criminal Detection
Amazon.com: Investigative Data Mining for Security and Criminal Detection
From the back cover: "Investigative Data Mining for Security and Criminal Detection is the first book to outline how data mining technologies can be used to combat crime in the 21st century. It introduces security managers, law enforcement investigators, counter intelligence agents, fraud specialists and information security analysts to data mining techniques and shows how they can be used as investigative tools. Readers will learn how to search public and private databases and networks to flag potential security threads and root out criminal activities even before they occur".
This book shows the reader how to analyze what sometimes appear to be small unconnected pieces of data in order to build up a detailed view of a problem or situation, and looking at what this data may be used for in a wide range of applications. It also discusses the collection and verification of data from multiple sources.
As a book written in the "Post 9/11" era that deals with issues such as crime and terrorism, the terrible events which occured on 9/11 are referenced frequently through out the book. Many ideas are raised looking at how information warfare can be used to predict and prevent acts of terrorism and other crimes in the 21st century.
This book is primarly aimed at anyone involved in crime prevention and detection worldwide, from decision makers to people who actually implement data analysis of any sort, this could even be extended to include those responsible for Internal Investigations within a corperate environment. This book is not really aimed at the IT implementer at a technical level.
Content & Overview
Think for a moment about how marketing companies often use data about us when selling items? They will use seemingly innocent data, such as a persons age to target sales of a product to all the people of a particular age group, eg, they wouldnt try to sell pensions to toddlers (well, they might ), but they would say, "hey, we have a toddler here, a suitable product for him/her would be xyz latest toy". They know that there is a good chance a toddler will be interested in the xyz latest toy based on sales information about xzy latest toy to others in his/her age group. It is this sort of thinking that this book encourages and details, but applied to a different subject, namely crime fighting.
The book covers indepth the inital collection, format, and quality of data, which as the book explains, is a vital part of any data analysis system. Keeping data in a rigid format makes processing of the data at a later stage much easier. As an example, what is otherwise known as "free form text" is discouraged, and multiple choice is suggested as an alternative due to much stricter controls on the data input. Free form text is also a bad idea as many people from different parts of the country will have different ways of writing, slang terms will be used in one place, but not another, something that is either "yes" or "no", or 0 or 1 cannot be misread to mean the other.
Out of all the chapters in the book, I must say I did particularly like Criminal Patterns, which looks at the modus operandi of several different types of crimes including Financial crimes, medical scams and Insurance Fraud. Real life examples are used in this chapter and throughout the book which is a definate plus point.
- Precrime Data Mining
- Investigative Data Warehousing
- Link Analysis: Visualizing Associations
- Intelligent Agents: Software Detectives
- Text Mining: Clustering Concepts
- Neural Networks: Classifying Patterns
- Machine Learning: Developing Profiles
- NetFraud: A Case Study
- Criminal Patterns: Detection Techniques
- Intrusion Detection: Techniques and Systems
- The Entity Validation System(EVS): A conceptual Architecture.
- Mapping Crime: Clustering Case Work
Layout & Style
The book is cleanly presented and includes screenshots of software used for data mining and analysis. Charts are used to explain how pieces of information link together in a descriptive manner, and are also used as examples of what some data analysis software can produce when used correctly.
Each chapter has an introduction approximatly 1-2 page long which briefly outlines the ideas & contents of the chapter, then the issues outlined are explored in greater depth. Sometimes the book is a bit repitive, especally in the early chapters, but i suppose that is part of making you remember the most important base information the rest of the book is built on. With that as the exception, the writing style is fairly constant throughout.
The book is laid out in a logical manner, starting with a gentle introduction on data usage, the collection of data("mining"), moving on to the analysis, pattern finding and predictions later on. Each chapter has an "internet resources" section at the end, giving details of places to continue reading and researching the topic of discussion for that section.
Personally I found the book good in that it put forward a broad range of ideas in a way that was easy to understand, but found it a little overwhelming in places due to the volume of information contained within. I think the book is a good purchase for anyone involved in data analysis or data collection in a general sense, not just in a crime prevention sense, as the book explains new ideas that can be applied to many types of data systems.
This book earns extra points for its useful Appendix section (which in most books i dont normally give a 2nd look), but this one actually has some quite useful information and resources within. "1,000 Online Sources for the investigative data miner" is litterally just that. A link to a page, and a brief description of exactly what is there. The links in this section were relevant and useful. "Intrusion Detection Systems (IDS) Products, Services Freeware, and Projects" also had links to some interesting web projects, some of which had been mentioned within the book already.
Overall Rating: A positive 8 out of 10
This review is copyright 2003 by the author and Security-Forums Dot Com, and may not be reproduced in any form in any media without the express permission of the author, or Security-Forums Dot Com.