Atish P. Sinha

Cognitive fit in requirements modeling: a study of object and process methodologies

Requirements modeling constitutes one of the most important phases of the systems development life cycle. Despite the proliferation of methodologies and models for requirements analysis, empirical work examining their relative efficacy is limited. This paper presents an empirical examination of object-oriented and process-oriented methodologies as applied to object-oriented and process-oriented tasks. The conceptual basis of the research model is derived from the theory of cognitive fit, which posits that superior problem-solving performance will result when the problem-solving task and the problem-solving tool emphasize the same type of information. Two groups of subjects participated in an experiment that required them to construct solutions to two requirements-modeling tasks, one process-oriented and the other object-oriented. One group employed the object-oriented tool while the other used the process-oriented tool. As predicted by the theory of cognitive fit, superior performance was observed when the process-oriented tool was applied to the process-oriented task. For the object-oriented task, however, the performance effects of cognitive fit require further investigation since there was no difference in subject performance across the two tools.

An empirical investigation of the key determinants of data warehouse adoption

Data warehousing (DW) has emerged as one of the most powerful decision support technologies during the last decade. However, despite the fact that it has been around for some time, DW has experienced limited spread/use and relatively high failure rates. Treating DW as a major IT infrastructural innovation, we propose a comprehensive research model - grounded in IT adoption and organizational theories - that examines the impact of various organizational and technological (innovation) factors on DW adoption. Seven factors - five organizational and two technological - are tested in the model. The study employed rigorous measurement scales of the research variables to develop a survey instrument and targeted 2500 organizations in both manufacturing and services segments within two major states in the United States. A total of 196 firms (276 executives), of which nearly 55% were adopters, responded to the survey. The results from a logistic regression model, initially conceptualizing a direct effect of each of the seven variables on adoption, indicate that five of the seven variables (three organizational factors - commitment, size, and absorptive capacity - and two innovation characteristics - relative advantage and low complexity) are key determinants of DW adoption. Although scope for DW and preexisting data environment within the organization were favorable for adopter firms, they did not emerge as key determinants. However, the study provided an opportunity to explore a more complex set of relationships. This alternative structural model (using LISREL) provides a much richer explanation of the relationships among the antecedent variables and with adoption, the dependent variable. The study, especially the revised conceptualization, contributes to existing research by proposing and empirically testing a fairly comprehensive model of organizational adoption of an information technology (IT) innovation, more specifically a DSS technology. The findings of the study have interesting implications with respect to IT/DW adoption, both for researchers and practitioners.

A comparison of data warehousing methodologies

Using a common set of attributes to determine which methodology to use in a particular data warehousing project.

Incorporating domain knowledge into data mining classifiers: An application in indirect lending

Data mining techniques have been applied to solve classification problems for a variety of applications such as credit scoring, bankruptcy prediction, insurance underwriting, and management fraud detection. In many of those application domains, there exist human experts whose knowledge could have a bearing on the effectiveness of the classification decision. The lack of research in combining data mining techniques with domain knowledge has prompted researchers to identify the fusion of data mining and knowledge-based expert systems as an important future direction. In this paper, we compare the performance of seven data mining classification methods-naive Bayes, logistic regression, decision tree, decision table, neural network, k-nearest neighbor, and support vector machine-with and without incorporating domain knowledge. The application we focus on is in the domain of indirect bank lending. An expert system capturing a lending experts knowledge of rating a borrowers credit is used in combination with data mining to study if the incorporation of domain knowledge improves classification performance. We use two performance measures: misclassification cost and AUC (area under the curve). A 2x7 factorial, repeated-measures ANOVA, with the two factors being domain knowledge (present or absent) and data mining method (seven methods), as well as a special statistical test for comparing AUCs, is used for analyzing the results. Analysis of the results reveals that incorporation of domain knowledge significantly improves classification performance with respect to both misclassification cost and AUC. There is interaction between classification method and domain knowledge. Incorporation of domain knowledge has a higher influence on performance for some methods than for others. Both measures-misclassification cost and AUC-yield similar results, indicating that the findings of the study are robust.

Effects of feature construction on classification performance: An empirical study in bank failure prediction

While extensive research in data mining has been devoted to developing better classification algorithms, relatively little research has been conducted to examine the effects of feature construction, guided by domain knowledge, on classification performance. However, in many application domains, domain knowledge can be used to construct higher-level features to potentially improve performance. For example, past research and regulatory practice in early warning of bank failures has resulted in various explanatory variables, in the form of financial ratios, that are constructed based on bank accounting variables and are believed to be more effective than the original variables in identifying potential problem banks. In this study, we empirically compare the performance of two sets of classifiers for bank failure prediction, one built using raw accounting variables and the other built using constructed financial ratios. Four popular data mining methods are used to learn the classifiers: logistic regression, decision tree, neural network, and k-nearest neighbor. We evaluate the classifiers on the basis of expected misclassification cost under a wide range of possible settings. The results of the study strongly indicate that feature construction, guided by domain knowledge, significantly improves classifier performance and that the degree of improvement varies significantly across the methods.

Object-oriented modeling with UML: a study of developers' perceptions

The object-oriented (OO) approach provides a powerful and effective environment for modeling and building complex systems. It supports a variety of techniques for analyzing, designing, and implementing flexible and robust real-world systems, providing benefits such as encapsulation, polymorphism, inheritance, and reusability [5, 9]. During the analysis phase, systems analysts abstract concepts from the application domain and describe what the intended system must do, rather than how it will be done. They specify the functional behavior of the system, independently of concerns relating to the environment in which it is ultimately implemented. During the design phase, systems designers define how the application-oriented analysis model will be realized in the implementation environment. Finally, during the implementation phase, the design models are translated into program code using a chosen programming language and, often, into a database implementation using a specific DBMS. The Unified Modeling Language (UML) was adopted as a standard for OO modeling by the Object Management Group (OMG) in 1997. The UML has already found widespread use in diverse domains such as e-commerce, command and control, computer games, medical electronics, banking, insurance, telephony, robotics,

On the usability of OO representations

OO OO OO OO OO OO OO OO OO OO OO OO OO OO OO OO OO OO OO OO OO OO OO OO OO OO OO OO OO OO OO OOO OO O OO OO OO OO OO OO OO OO OO OO O OO OO OO OO OO OO OO OO OO OO OO OO OO OO OO OO OO OO OO OO OO OO OO OO OO OO OO OO OO OO OO OO O OO OO OO OO OO OO OO OO OO OO OO OO OO O OO OO OO OO OO OO OO OO OO OO OO OO OOOO OO OO OO OO OO OO OO OO OO OO OO OO OO OO OO OO OOOO OO OO OO OO OO OO OO OO OO OO OO OOOO OO OO OO COMMUNICATIONS OF THE ACM October 2000/Vol. 43, No. 10 83 On the Usability of oo Representations

The influence of reviewer engagement characteristics on online review helpfulness: A text regression model

Abstract The era of Web 2.0 is witnessing the proliferation of online social media platforms, which develop new business models by leveraging user-generated content. One rapidly growing source of user-generated data is online reviews, which play a very important role in disseminating information, facilitating trust, and promoting commerce in the e-marketplace. In this paper, we develop and compare several text regression models for predicting the helpfulness of online reviews. In addition to using review words as predictors, we examine the influence of reviewer engagement characteristics such as reputation, commitment, and current activity. We employ a reviewers RFM (Recency, Frequency, Monetary Value) dimensions to characterize his/her overall engagement and investigate if the inclusion of those dimensions helps improve the prediction of online review helpfulness. Empirical findings from text mining experiments conducted using reviews from Yelp and Amazon offer strong support to our thesis. We find that both review text and reviewer engagement characteristics help predict review helpfulness. The hybrid approach of combining the textual features of bag-of-words model and RFM dimensions produces the best prediction results. Furthermore, our approach facilitates the estimation of the helpfulness of new reviews instantly, making it possible for social media platforms to dynamically adjust the presentation of those reviews on their websites.

Cognitive fit: an empirical study of recursion and iteration

A laboratory experiment was conducted to assess the basic theory and extensions to the theory for recursive tasks across programming languages. The experiment used 34 LISP and 48 PASCAL computer science students in two repeated measures designs. Findings of the study are reported and analyzed. The results strongly suggest that investigation of programming constructs should take place in the context of specific programming languages. Since a number of languages provide similar kinds of programming constructs, it is difficult for programmers to choose those implementations that best suit their needs. One way of encouraging the use of desirable constructs would be to develop languages adapted to certain types of tasks. Such an approach would inherently lead to cognitive fit and the attendant performance benefits would be realized. >

Data Warehousing Process Maturity: An Exploratory Study of Factors Influencing User Perceptions

This paper explores the factors influencing perceptions of data warehousing process maturity. Data warehousing, like software development, is a process, which can be expressed in terms of components such as artifacts and workflows. In software engineering, the Capability Maturity Model (CMM) was developed to define different levels of software process maturity. We draw upon the concepts underlying CMM to define different maturity levels for a data warehousing process (DWP). Based on the literature in software development and maturity, we identify a set of features for characterizing the levels of data warehousing process maturity and conduct an exploratory field study to empirically examine if those indeed are factors influencing perceptions of maturity. Our focus in this paper is on managerial perceptions of DWP. The results of this exploratory study indicate that several factors-data quality, alignment of architecture, change management, organizational readiness, and data warehouse size-have an impact on DWP maturity, as perceived by IT professionals. From a practical standpoint, the results provide useful pointers, both managerial and technological, to organizations aspiring to elevate their data warehousing processes to more mature levels. This paper also opens up several areas for future research, including instrument development for assessing DWP maturity