Text

theses

Over the past few years, we have seen a significant grown in interest for open data, specifically open government data (OGD). This led to the availability of a large number of public sectors’ datasets made available to the citizens or any other interested stakeholder. Thanks to the pressure being placed on all types of government organizations in order to release their raw data. The main motivations for publicizing access to raw materials and make it more transparent are that it can help provide higher returns from the public utilization of such data, can provide policymakers with supportive data that can assess the process of making better decisions, can generate wealth through the development and creation of new and innovative products and services or enhance the current ones, and can involve the citizens to monitor and analyze publicly available datasets to help them evaluate and assess the performance of their governments. This dissertation consists of three essays on open government data and data analytics. It explores and contributes to the literature in introducing a new model for effective and efficient open governments, introduces application of visualizations and data analytics of governmental data and a text mining analysis for government-related financial data.
The first essay considers the process and the use of open government data (OGD) initiatives by focusing on financial reporting with the use of procurement contracts. It pursues two main arguments regarding (i) the possibility of disseminating more financial data, especially procurement tenders, and how this could transform relationships between different levels of government and citizens; (ii) and discusses the level of data transparency based on the definition of the open data model; and more importantly, the study introduces a new model for effective and efficient open government data by adding new dimensions to the open data concept when utilized by governments.
The second essay investigates the use of visualization and cluster analysis techniques in governmental, publicly available datasets. It examines the utilization of advanced data mining techniques such as hierarchical, k-means clustering and visualization in two case studies. In the first case study, we explore the literature for the use of emerging data mining techniques in auditing. Then we apply k-means and hierarchical clustering on U.S. states financial statements data. In particular, we demonstrate how cluster analysis could be applied as supportive tools for auditing governmental bodies. The second case study utilizes the Volcker Alliance’s Survey data results. The survey produces extensive information about how the different U.S. states score on an annual basis on budgeting using five measures. We apply cluster analysis and visualization on the budget data. On both case studies, we demonstrate how visualization and data analytics especially cluster analysis could be used on governmental data and to help gain more insights about financial statements and budgeting.
The third essay focuses on text mining analytics for government-related financial data. Specially, utilizing a text mining implementation of the Financial Industry Business Ontology (FIBO) to extract financial information from the social media platform Twitter regarding financial and budget information in the public sector, namely the two public-private agencies of the Port Authority of New York and New Jersey (PANYNJ), and the New York Metropolitan Transportation Agency (MTA). This research initiative develops a methodology to classify tweets that are related to financial bonds. We apply a frame and slot approach from the artificial intelligence literature to operationalize the FIBO ontology in a public sector/municipalities business context. FIBO is part of the Enterprise Data Management Council (EDMC) and Object Management Group (OMG) family of specifications. FIBO provides standards for defining the facts, terms, and relationships associated with financial concepts. One contribution of this paper is that it is the first to recognize that the FIBO structure provides a grammar of financial concepts which can be used to classify social media. We show that this grammar can be used to mine semantic meaning from unstructured textual data. The Twitter stream is monitored and analyzed with frames derived from FIBO and using keywords. The ability of the FIBO frames to detect semantic meaning in tweets is compared with naïve keyword analysis and by determining the number of false positive classified from the Twitter stream. Using FIBO frames, constituent semantic structures can be uncovered to predict reactions to policies and programs and perform other environmental scanning more quickly than by following the feeds manually.

ETD_9892

doi:10.7282/t3-8vf8-6926

Ph.D.

Includes bibliographical references

ETD doctoral

The author owns the copyright to this work.