SBIR Phase I: Wavelet-based Data Exploration for Auditing, Decision Support, and Corporate Performance Monitoring

这个小型企业创新研究(SBIR)第一阶段项目将为审计软件提供可视化和分析功能，使非科学用户能够更快、更准确地识别和监控模式、趋势和异常。现有的审计工具要求审计人员随机抽样数据或使用不熟悉的复杂统计方法来检查数据。这个问题因企业系统收集的大量数据以及FAR和萨班斯-奥克斯利法案等监管复杂性而变得复杂。小波提供了一种方法，用于在多个分辨率级别上探索更大的数据量，同时突出显示异常、趋势和差异。将小波方法应用于预算和财务控制环境带来两个重大挑战：创建一种内在可伸缩和及时的方法；第二，使审计员、决策者和支助系统能够直观地了解原始数据并对其采取行动。这项第一阶段研究的目标是论证利用小波作为加强审计和财务控制方法的基础的可行性。SBIR倡议的更广泛影响将不仅是提高财务审计和管理决策者的决策质量和及时性，而且还将提高执行和监测业务流程的系统的决策质量和及时性。这项研究工作的结果将提高几个部门的经济生产率，并为学术界和工业界带来决策支持和数据工程方面的新研究机会。

This Small Business Innovation Research (SBIR) Phase I project will provide a visualization and analysis capability to auditing software that will enable non-scientifically oriented users to identify and monitor patterns, trends, and exceptions faster and with greater accuracy. Existing auditing tools require auditors to randomly sample data or use unfamiliar complex statistical methods to examine data. This problem is complicated by volume of data collected by enterprise systems and regulatory complexity such as the FAR and the Sarbanes-Oxley Act. Wavelets provide a method for exploring greater volumes of data at multiple levels of resolution while highlighting exceptions, trends, and variances. Applying wavelet methods to budgetary and financial control environments, presents two significant challenges: creating a methodology that is inherently scalable and timely; and secondly, enabling auditors, decision-makers, and support systems to visualize and act on the knowledge derived from the raw data. The objective of this Phase 1 research is to demonstrate the feasibility of utilizing wavelets as the basis for an enhanced auditing and financial control methodology. The broader impact of this SBIR initiative will be to improve the decision quality and timeliness of not only financial auditing and management decision-makers, but also of the systems that implement and monitor business processes. The result of this research effort will improve the economic productivity of several sectors and introduce new research opportunities in decision support and data engineering to both academia and industry.