Data Mining and Predictive Analysis (eBook)

Intelligence Gathering and Crime Analysis

Colleen McCue (Autor)

eBook Download: PDF | EPUB

2014 | 2. Auflage
422 Seiten
Elsevier Science (Verlag)
978-0-12-800408-1 (ISBN)

Infos

Data Mining and Predictive Analysis: Intelligence Gathering and Crime Analysis, 2nd Edition, describes clearly and simply how crime clusters and other intelligence can be used to deploy security resources most effectively. Rather than being reactive, security agencies can anticipate and prevent crime through the appropriate application of data mining and the use of standard computer programs. Data Mining and Predictive Analysis offers a clear, practical starting point for professionals who need to use data mining in homeland security, security analysis, and operational law enforcement settings. This revised text highlights new and emerging technology, discusses the importance of analytic context for ensuring successful implementation of advanced analytics in the operational setting, and covers new analytic service delivery models that increase ease of use and access to high-end technology and analytic capabilities. The use of predictive analytics in intelligence and security analysis enables the development of meaningful, information based tactics, strategy, and policy decisions in the operational public safety and security environment. - Discusses new and emerging technologies and techniques, including up-to-date information on predictive policing, a key capability in law enforcement and security - Demonstrates the importance of analytic context beyond software - Covers new models for effective delivery of advanced analytics to the operational environment, which have increased access to even the most powerful capabilities - Includes terminology, concepts, practical application of these concepts, and examples to highlight specific techniques and approaches in crime and intelligence analysis

Dr. Colleen McCue is the Senior Director of Social Science and Quantitative Methods at DigitalGlobe. Her areas of expertise within , in the applied public safety and national security environment include the application of data mining and predictive analytics to the analysis of crime and intelligence data, with particular emphasis on deployment strategies; surveillance detection; threat and vulnerability assessment; geospatial predictive analytics; computational modeling and visualization of human behavior; Human, Social, Culture and Behavior (HSCB) modeling and analysis; crisis and conflict mapping; and the behavioral analysis of violent crime in support of anticipation and influence.

Textauszug

Preface

So many things have changed since the first edition of this textbook, particularly as relates to data, technology, and tradecraft. Some things have not changed, however, including my love of science and desire to develop innovative solutions to some of the really challenging public safety and security challenges. Operational security analytics, at its core, is designed to effectively characterize bad behavior in support of information-based approaches to anticipation and influence. Whether “influence” entails prevention, thwarting, mitigation, response, or consequence management, we are trying to change outcomes for the better.

In the beginning of my operational security analytics journey, I became profoundly intrigued by how many of the seasoned detectives I worked with were often able to generate quick yet accurate hypotheses about their cases, sometimes only moments after they had arrived at the scene. Like the “profilers” on television and in the movies, many of them seemed to have an uncanny ability to accurately describe a likely motive and related suspect based merely on a review of the crime scene and some preliminary knowledge regarding the victim’s lifestyle and related risk factors. Over time, I started to acquire this ability as well, although to a lesser degree. It became much easier to read a report and link a specific incident to others, predict future related crimes, or even calculate the likelihood that a particular case would be solved based on the nature of the incident. Drawing on my training as a scientist, I frequently found myself looking for some order in the chaos of crime, trying to generate testable hypotheses regarding emerging trends and patterns, as well as investigative outcomes. Sometimes I was correct. However, even when I was not, I was able to include the information in my ever-expanding internal rule sets regarding crime and criminal behavior.

Prior to working for the Richmond Police Department, I spent several years working with that organization. Perhaps one of the most interesting aspects of this early relationship with the Department was my weekly meeting with the officer in charge of violent crimes. Each week we would discuss the homicides from the previous week, particularly any unique or unusual behavioral characteristics. Over time, we began to generate casual predictions of violent crime trends and patterns that proved to be surprisingly accurate. During the same time period, I also began to examine intentional injuries among incarcerated offenders. As I probed the data and drilled down in an effort to identify potentially actionable patterns of risk, it became apparent that many of the individuals I looked at were not just in the wrong place at the wrong time, as they frequently indicated. Rather, they were in the wrong place at the wrong time doing the wrong things with the wrong people and were assaulted as a result of their involvement in these high-risk activities. As I explored the data further, I found that different patterns of offending were associated with different patterns of risk. This work had immediate implications for violence reduction efforts. It also had implications for the analysis of crime and intelligence data. Fortunately, the field of data mining and predictive analytics had evolved to the point that many of the most sophisticated algorithms were available in a PC environment, so that everyone from a software-challenged psychologist like myself to a beat cop could begin to not only understand but also use these incredibly powerful tools.

Although I did not realize it at the time, a relatively new approach to marketing and business intelligence was emerging at the same time we were engaging in this lively speculation about crime and criminals at the police department. Professionals in the business community were exploiting artificial intelligence and machine learning to characterize and retain customers, increase sales, focus marketing campaigns, and perform a variety of other business-related tasks. For example, each time I went through the checkout counter at my local supermarket, my purchasing habits were coded, collected, and analyzed. This information was aggregated with data from other shoppers and employed in the creation of models about purchasing behavior and how to turn a shopper into a buyer. These models were then used to gently mold my future behavior through everything from direct marketing based on my existing preferences to the strategic stocking of shelves in an effort to encourage me to make additional purchases during my next trip down the aisle. Similarly, data and information were collected and analyzed each time I perused the Internet. As I skipped through web pages, I left cookies, letting the analysts behind the scenes know where I went and when and in what sequence I moved through their sites. All of this information was analyzed and used to make their sites more friendly and easier to navigate or to subtly guide my behavior in a manner that would benefit the online businesses that I visited. The examples of data mining and predictive analytics in our lives are almost endless, but the contrast between my professional and personal lives was profound. Directly comparing the state of public safety analytical capacity to that of the business community only served to underscore this shortcoming. Throughout almost every aspect of my life, data and information were being collected on me and analyzed using sophisticated data mining algorithms; however, the use of these very powerful tools was severely limited or nonexistent in the public safety arena in which I worked. With very few exceptions, data mining and predictive analytics were not readily available for the analysis of crime or intelligence data, particularly at the state and local levels.

Like most Americans, I was profoundly affected by the events of September 11th. In the week of September 10th, 2001, I was attending a specialized course in intelligence analysis in northern Virginia. Like many, I can remember exactly what I was doing that Tuesday morning when I saw the first plane hit the World Trade Center and how I felt as the horror continued to unfold throughout the day. As I drove back to Richmond, Virginia that afternoon (the training had been postponed indefinitely), I saw the smoke rise up over the Beltway from the fire at the Pentagon, which was still burning. Those of us working in the public safety community were inundated with information over the next several days, some of it reliable, much of it not. Like many agencies, we were swamped with the intelligence reports and BOLOs (be on the lookout reports) that came in over the teletype, many of which were duplicative or contradictory. Added to that were the numerous suspicious situation reports from concerned citizens and requests for assistance from the other agencies pursuing the most promising leads. Described as the “volume challenge” by former CIA director George Tenent, the amount of information threatened to overwhelm us. Because of this, it lost its value. There was no way to effectively manage the information, let alone analyze it. In many cases, the only viable option was to catalog the reports in three-ring binders, with the hope that it could be reviewed thoroughly at some later date. Like others in law enforcement, our lives as analysts changed dramatically that day. Our professional work would never again be the same. In addition to violent crimes and vice, we now have the added responsibility of analyzing data related to the war on terrorism and the protection of homeland security, regardless of whether we work at the state, local, or federal level. Moreover, if there was one take-home message from that day as an analyst, particularly in Virginia, it was that the terrorists had been hiding in plain sight among us, sometimes for years, and they had been engaging in a variety of other crimes in an effort to further their terrorist agenda, including identity theft, forgery, and smuggling; not to mention the various immigration laws they violated. Many of these crimes fall within the purview of local law enforcement.

As we moved through the days and weeks following the attacks, I realized that we could do much better as analysts. The subsequent discussions regarding “connecting the dots” highlighted the sad fact that quite a bit of information had been available before the attacks; however, flaws in the sharing and analysis of information resulted in tragic consequences. Although meaningful information sharing remains an important goal, advanced analytical techniques are available now. The same tools that were being used to prevent people from switching their mobile telephone service provider and to stock shelves at our local supermarkets on September 10th can be used to create safer, healthier communities and enhance homeland security. The good news is that these techniques and tools are being used widely in the business community. The key is to apply them to questions or challenges in public safety, law enforcement, and intelligence analysis.

Again, I thoroughly enjoy science and particularly like the new concept of “data science,” which really captures the creative aspects of analysis and associated promise of transdisciplinary approaches. As someone who likes to color outside the lines and explore novel approaches to analysis, I am intrigued by the use of advanced analytics to improve other aspects of my life and see data science as a means to an end; as a means by which to better understand behavior—good, bad and otherwise—so that we can use it to anticipate and influence outcomes, particularly in support of enhanced public safety and security....

Produkt-Details

• Erscheint lt. Verlag: 30.12.2014
• Sprache: englisch
• Themenwelt: Informatik ► Datenbanken ► Data Warehouse / Data Mining
• Themenwelt: Mathematik / Informatik ► Mathematik
• Themenwelt: Recht / Steuern ► EU / Internationales Recht
• Themenwelt: Recht / Steuern ► Strafrecht ► Kriminologie
• Themenwelt: Recht / Steuern ► Strafrecht ► Strafverfahrensrecht
• Themenwelt: Sozialwissenschaften ► Kommunikation / Medien ► Buchhandel / Bibliothekswesen
• Themenwelt: Sozialwissenschaften ► Politik / Verwaltung
• ISBN-10: 0-12-800408-8 / 0128004088
• ISBN-13: 978-0-12-800408-1 / 9780128004081