The Design and Implementation of a Corporate Householding Knowledge Processor to Improve Data Quality

Abstract

Advances in Corporate Householding are needed to address certain categories of data quality problems caused by data misinterpretation. In this paper, we first summarize some of these data quality problems and our more recent results from studying corporate householding applications and knowledge exploration. Then we outline a technical approach to a Corporate Householding Knowledge Processor (CHKP) to solve a particularly important type of corporate householding problem -entity aggregation. We illustrate the operation of the CHKP by using a motivational example in account consolidation. Our CHKP design and implementation uses and expands on the COntext INterchange (COIN) technology to manage and process corporate householding knowledge.

Reasoning About Temporal Context Using Ontology and Abductive Constraint Logic Programming

“…For example, instead of depending on time periods, it depends on purposes (SEC filing, risk assessment, taxation, etc) that the rules differ for whether the total revenue of a corporation should include those of foreign branches, subsidiaries, subsidiaries of branches and subsidiaries, and other companies majority-owned by the corporation. As is demonstrated in the Corporate Householding research [16], COIN framework can be applied to this scenario as well to represent and reasoning about those complex rules.…”

A Context-Based Approach to Reconciling Data Interpretation Conflicts in Web Services Composition

“…In Example 2 we use the modifier of value-added taxes and the conversion cvtVAT to deal with the difference of value-added taxes, a kind of conceptual-level data misinterpretation problems. Other conceptual-level problems like those of "Corporate Householding" (Madnick et al 2003) and temporal-level problems ) can also be modeled using appropriate modifiers and addressed in a similar way. With the ontology/context modeling and semantic annotation in place, all the possible data misinterpretation problems in Table 2 that may occur in Web services composition can be addressed by the approach.…”

“…For example, a retail price reported by European services usually includes the valueadded taxes, while retail prices reported by US services, especially for purchases to be done in a store, usually do not include the value-added taxes. 6 An even more challenging problem in this category is referred to as "Corporate Householding" (Madnick et al 2003) which refers to misinterpretation of corporate household data. For example, the answer to "What were the total sales of IBM" varies depending on whether the sales of majority owned subsidiaries of IBM should be included or not.…”

A Simplified Approach for Quality Management in Data Warehouse

“…For this, organizations may consider correcting defects manually or hiring agencies that specialize in data enhancement and cleansing. Error detection and correction can also be automated by the adoption of methods that optimize inspection in retrieval of data from data warehouse while generating new information [10], integrity rule-based systems [11], and software agents that detect quality violations [12]. Some ETL (Extraction, Transformation, and Loading) tools also support the automation of error detection and correction.…”