Data mart

About: Data mart is a research topic. Over the lifetime, 559 publications have been published within this topic receiving 8550 citations.

Incident Database-based Framework for Establishing Industrial Safety Performance Assessments

Abstract: Many organizations collect data on industrial incidents. These organizations differ from each other in their interests, data collection procedures, definitions, and scope, and each of them is analyzing its data to achieve its goal and to accomplish its mission. However, there were no attempts to explore the potential hidden in integrating data sources. Extensive efforts are required in order to integrate information from different data sources as well as to identify the effects of the individual aspects of data collection procedures on the quality and completeness of the data. This work presents a methodology for incident data collection from various sources, and the opportunity that exists in a combined data mart for industrial safety performance assessment and identification of trends. It also presents the major challenges associated with utilizing databases, furthermore integrating these data sources for conducting industry-wide performance assessments. The study also discusses how the proposed analysis can be used to determine the areas for major reduction of losses and number of incidents. Introduction The fallout of dioxin caused by a runaway reaction at Seveso, Italy, in 1976, and the 1984 disaster of Bhopal, India, led to major changes in laws over the world. Federal and industrial entities devoted major efforts toward risk reduction and hazard control. Most of the organizations in the chemical industry integrated their systems for safety. Beyond measuring performance within facilities (Keren, 2003, 2009), and among facilities (Keren, 2002, 2005), most of the efforts in the development of safety performance measurements are invested toward measuring the industry as a whole and with some efforts directed toward performance measurements of federal agencies. The Occupational Safety and Health Administration (OSHA) is a federal agency under the authority of the Department of Labor (DOL) and is responsible for the safety and health of employees in the work place. OSHA’s incident rate is a statistical index that measures illnesses and injuries per 100 worker years (DoT, 2009). The Fatality Accident Rate (FAR) is a European index mostly used by the British and is a statistical index that measures the number of fatalities per 1000 employees working their entire lifetime (50 working years per employee). Indices such as FAR and Incident Rate, which represent failure to effectively control risks, are called Trailing Indicators. These indices are important, and can be used to measure performance in industries. While OSHA’s Bureau of Labor Statistics produces several such indices, these indices are normalized, and are per industrial sectors. There is still a need for a framework that will allow to conduct estimation of variety of performance indices for more specific, yet inclusive, industrial arenas. A most desirable type of arenas are performance indices for product-based (e.g., Chlorine, Ammonia, etc.). Newell (2001) presents a very well developed concept of process safety performance measurements. In his work, he analyzes in detail OSHA’s database as a sole source of data for performance measurements. Newell recommends Dr. Nir Keren is an Assistant Professor of Occupational Safety at the Department of Agricultural and Biosystems Engineering and a Graduate Faculty in the Human Computer Interaction program at Iowa State University. His research interest fall into two broad categories: safety decision making and harnessing incident databases to enhance loss prevention. In safety decision making, his interest is in developing naturalistic decision making models; more specifically, in developing predictive models for emergency responders’ decision making under a variety of domains such as environmental constraints, organizational climates and cultures, and personal propensities. In the incident databases area, his interest is in quantitative risk assessment, risk analysis of transportation of hazardous material, and assessment of industrial safety performance. Dr. Keren received his Ph.D. at Texas A&M University, College Station, Texas. He earned his B.S. in Mechanical Engineering and M.S. in Management and Safety Engineering, both from the Ben Gurion University, Beer-Sheva, Israel.

Crime data warehousing and crime pattern discovery

Abstract: The structure of crime activity logs stored in police databases is not designed for decision support systems and hence not for complex crime analysis. This paper shows how crime log data could be converted to significantly useful information using data warehousing and data mining techniques. Crime incident data and other relevant data are organized with applying data warehousing concepts. Spatial association rule mining is used for finding interesting local relationship patterns of crime incidents with other spatial features. This paper shows how to materialize spatial relationship information between crime activities and task-relevant other features into data marts, and to discover interesting crime patterns from the data mart using a spatial association rule mining technique. A proof of concept is carried out with real crime data and points of interest in a study area to illustrate and evaluate the proposed approach. The case study results show the usefulness of such data warehousing and spatial association rule mining for crime data analysis.

An architecture and services for constructing data marts from online data sources

Abstract: The Agri sector has shown an exponential growth in both the requirement for and the production and availability of data. In parallel with this growth, Agri organisations often have a need to integrate their in-house data with international, web-based datasets. Generally, data is freely available from official government sources but there is very little unity between sources, often leading to significant manual overhead in the development of data integration systems and the preparation of reports. While this has led to an increased use of data warehousing technology in the Agri sector, the issues of cost in terms of both time to access data and the financial costs of generating the Extract-Transform-Load layers remain high. In this work, we examine more lightweight data marts in an infrastructure which can support on-demand queries. We focus on the construction of data marts which combine both enterprise and web data, and present an evaluation which verifies the transformation process from source to data mart.