Search

We investigate modeling strategies for English code-switched words as found in a Swahili spoken term detection system. Code switching, where speakers switch language in a conversation, occurs frequently in multilingual environments, and typically de- teriorates STD performance. Analysis is performed in the context of the IARPA Babel program which focuses on rapid STD system development for under-resourced languages. Our results show that approaches that specifically target the modeling of code-switched words, significantly improve the detection performance of these words. [ABSTRACT FROM AUTHOR]

The elaborate analysis of a business process (BP) typically informs its potential for business process redesign (BPR), but the latter is usually conducted in a non-systematic way. The purpose of this paper is the introduction of the Business Process Redesign Capacity Assessment (BP-RCA) framework that assesses the redesign capability of BP models, prior to their implementation. This study combines key redesign features introduced by domain experts, to a conceptual framework that takes into consideration an inclusive set of BPR components in three consecutive phases, towards facilitating organizations in the practice of redesign decision making. In this paper, an illustrative case study is used to present the initial phase (selection) of the framework. To assess the usability of the BP-RCA, the authors reviewed twelve established redesign initiatives from literature which proved to implicitly follow similar steps to the proposed framework. The findings indicate that the BP-RCA framework provides a systematic exploration of fundamental redesign aspects and can be used as a reliable measurement of the redesign capacity of candidate BP models. The framework also provides practitioners with the necessary methodology for increasing the BPR effectiveness, the robustness of the varying initiatives and the overall innovativeness of businesses. [ABSTRACT FROM AUTHOR]

This paper addresses the inspection and evaluation of design reuse in the conceptual schema of CMS-based Web applications in order to improve their design quality. We have utilized WebML as the design platform of our approach and we capture design reuse by detecting all the recurrent design structures within the hypertext schema of an application. We propose a methodology for automatically extracting the hypertext schema of an application which is then submitted to a pattern-based analysis in order to detect all the incorporated recurrent patterns implying design reuse. A pattern is a design structure consisting of a core specification, i.e., an invariant composition of WebML elements that characterizes the pattern and by a number of pattern variants which extend the core specification with all the valid modalities in which the pattern composition can start (starting variants) or terminate (termination variants). Finally, we apply a number of metrics on the identified patterns to evaluate the appropriateness of reuse which results in categorizing them as effective or poor design solutions. By applying the methodology on a web application, developers can inspect all the design structures that occur in its design probably due to design reuse and identify: (i) either effective reusable design solutions for implementing a certain behavior that can also be used as building blocks in future designs and can even form the basis for the discovery of design patterns in CMS domain, or (ii) ineffective recurrent design fragments causing design inconsistencies which imply the need for refactoring in order to improve the application׳s quality. [ABSTRACT FROM AUTHOR]

It is vastly acknowledged that analyzing social networks is a very challenging research area. Take as a striking example the organization of vertices in clusters, with many edges joining vertices of the same cluster and comparatively few edges joining vertices of different clusters. This comprises a fundamental aspect, which concerns the detection of user communities. In certain fields such as sociology and computer science where interactions and associations are often represented in the form of graphs, detecting communities is of vital importance. This paper addresses the need for an efficient and innovative methodology for community detection that will also leverage users’ behavior on emotional level. Ekman emotional scale is the key point with which the methodology analyzes user’s tweets in order to determine their emotional behavior. Consequently, the derived communities are estimated with the use of three different metrics, while the weighted version of a modularity community detection algorithm is utilized. There is substantial evidence indicating that our proposed methodology creates influential enough communities. [ABSTRACT FROM AUTHOR]