Wen-Tin Lee

Discovering Early Aspects through Goals Interactions

Aspect-oriented approaches have resulted in a great impact on the processing of system properties during the development of software systems. However, a systematic way for eliciting early aspects from requirements is still underdeveloped to better integrate early aspects with the analysis results. As an attempt towards the investigation of the interactions among goals, scenarios, and early aspects, we proposed, in this work, a goal-driven approach to the discovery of early aspects by means of a bidding process that organizes goals into goal clusters for discovering aspectual candidates by exploring the interactions among goals and use cases. By introducing early aspects, the goal-driven approach can be further enhanced to deal with the crosscutting properties in the analysis stage of software development. The proposed approach is illustrated using the problem domains of meeting scheduler system.

Framework for Enhancing Mobile Availability of RESTful Services

Providing users of mobile devices uninterrupted access to web services in unstable network conditions continues to be a problem. Numerous methods for service caching have been proposed; however, most studies fail to consider two crucial factors: (1) Context of network connectivity: Smartphones are used in a variety of wireless network conditions, such as high-speed networks, unstable networks, and areas without an available network connection; and (2) Service failure handling: Current service caching mechanisms are able to deal with temporary unavailability, but they cannot handle long-time service failures or malfunctions. This paper proposes a connectivity-aware, risk-driven (CARD) approach to the delivery of RESTful services. The CARD approach is encapsulated in the form of a client-side library for use by mobile applications (apps) to invoke backend RESTful services in a highly-available manner. The CARD approach has two main features: 1) the ability to perform actions specific to the network conditions, such as the application of prefetch services and caching services when connected to high speed wireless networks to ensure that cached services are used for unstable wireless networks, and allowing users to request cached service responses from other users when no wireless network can be accessed. 2) The proposed risk-driven analysis method enables the provision of a reasonable service recovery plan when the original service malfunctions. Experiments demonstrate that the proposed CARD approach expands the availability of service and shortens service response times under a variety of network conditions.

Fuzzy logic as a basic for use case point estimation

Project estimation based on function point or use case point (UCP) methods provide only fixed complexity grades which can not deal with the uncertain and imprecise conditions. This study, therefore, provides a fuzzy size estimation procedure for goal-driven use case model based on UCP using fuzzy theory. We propose a metric to calculate the unadjusted use case points of goal-driven use cases based on the relations between each use cases and goals with the fuzzy membership functions and the fuzzy rules. Furthermore, the technical and environmental factors are considered to calculate the use case points which can be used to estimate the implementation time and effort of the system under development. The proposed approach is illustrated by a benchmark problem domain of a meeting scheduler systems.

Service Brick Composition Framework for Smartphones

Nowadays, smart phones have become the most important mobile devices and can access thousands of mobile applications. However, retrieving required information from various resources by the smart phone is still a time-consuming, costly and inconvenient task for users. To address this issue, we introduce the concept of service brick, and propose a service brick composition framework to integrate required services and information. Notably, the web-based framework is devised according to the characteristics of smart phones, including relatively narrow network bandwidth and relatively small size screen. Besides, a visual design tool is provided to let the user build brick-based applications easily. To demonstrate the feasibility of the proposed approach, two illustrative examples, music composite app and travel composite app, are offered.

State-Driven and Brick-Based Mobile Mashup

Mobile applications (i.e., Mobile apps or apps) are becoming an important software delivery model. Users can employ a wide range of services associated with mobile apps, such as entertainment, news, travel, and social networking. Unfortunately, the retrieval of information from multiple apps, services, or local resources can be time-consuming, costly, and inconvenient. This paper proposes a novel mobile mash up approach, referred to as brick-based, state-driven mobile service composition (BSMSC) to overcome these difficulties. BSMSC comprises two primary mechanisms: (1) Android-fragment-based service bricks, and (2) a state-driven linkage for composite Restful services, which supports one-shot service flow execution as well as stateful service flow execution. The proposed BSMSC approach makes it possible to assemble fully-fledged, reconfigurable mobile mash up applications.

Efficient Service Discovery Using Similarity-Distance-Based Relaxation

Web service discovery is the process of locating web services to satisfy the requirements of service requesters. It plays an important role in building loosely coupled service-oriented applications. Nowadays, although many service discovery mechanisms are available, a lot of efforts do not focus on linking actual business services in real world and most methods do not provide an appropriate relaxation mechanism. Therefore, this study proposes a systematic service discovery approach to address the above issues. The main features of the proposed approach include (1) retrieving services through multiple filters, including kernel property filter, provider constraint filter, and user constraint filter, (2) calculating service ranking scores based on QoS (Quality of Service) constraint values and setting of QoS importance rank, and (3) assisting users in locating appropriate services by using the proposed similarity-distance-based relaxation mechanism.

Applying Software Effort Estimation Model Based on Work Breakdown Structure

Software effort estimation is an important activity for effective project management. To support the prediction of efforts in the early stage of a software project development, we propose an effort estimation model along with a software effort estimation process to estimate project efforts from high level WBS. The proposed estimation model is composed of four original task assignment factors - team size, concurrency, intensity, and fragmentation and one newly proposed factor, named WBS effort. To demonstrate the feasibility of proposed model, a case study is presented. The result of the case study shows that comparing the proposed model with those that using original task assignment effort estimation model, the accuracy of the prediction is improved by applying the five factors. For the data examined, the proposed model also shows to have better results in goodness of fit and quality of estimation than original one.

Improving Software Maintenance for Pattern-Based Software Development: A Comment Refactoring Approach

The effectiveness of the design patterns has been an important research topic in the software engineering area. Since uncontrolled use of design patterns results in serious problem of maintenance. Good code comments can help maintainers understand the design of the software, thus avoid the wrong utilization of design patterns. In this paper we will in our refactoring approach that can add pattern comment line to improve the software maintains for a pattern-based software.

Provision and consumption of linked open data: a define–produce–invoke approach

Open data is becoming increasingly popular in a wide range of service domains; however, most open datasets in Taiwan remain separate. The lack of linked open data (LOD) makes it difficult to locate and combine open datasets for the creation of innovative applications. In this study, we sought to facilitate the spread of open data in Taiwan using a novel approach referred to as define–produce–invoke (DPI). The proposed scheme employs a newly defined data query language, called LODQL (LOD query language), to allow the definition of rules for the generation of LOD by data experts. We also developed an LOD engine, which is able to produce linked open datasets and allow application developers to access LOD by invoking RESTful services. This scheme also allows data visualizations indicating the relevance of open datasets and the associations among open data items. Experiments demonstrate the feasibility and effectiveness of the proposed DPI approach.

LODQL: A Language for Creation, Query, and Service Generation of Linked Open Data

Nowadays, open data has become more and more popular and is widely used in a variety of service domains. However, in Taiwan, most open datasets are still separate while linked open data (LOD) is still lacking. Without LOD, application developers are difficult to locate and combine appropriate open datasets to create innovative applications. Therefore, to facilitate the consumption and the spreading of open data in Taiwan, we propose a new data query language, referred to as LODQL (Linked Open Data Query Language), to allow the definition of rules for building LOD by data experts and the consumption of LOD via RESTful services by application developers. We designed the LODE (LODQL Engine) to parse the LODQL, build LOD in RDF (Resource Description Framework) according to LODQL, and generate the corresponding RESTful services that extract required LOD based on the user query. LODE is also able to perform data visualization to show relevance of open datasets and relations between open data items. The developed system implementation and the evaluation experiments fully demonstrate that the feasibility and effectiveness of LODQL and LODE.