Hiroshi Igaki

Detecting Service Chains and Feature Interactions in Sensor-Driven Home Network Services

Sensor-driven services often cause chain reactions, since one service may generate an environmental impact that automatically triggers another service. We first propose a framework that can formalize and detect such service chains based on ECA (event, condition, action) rules. Although the service chain can be a major source of feature interactions, not all service chains lead to harmful interactions. Therefore, we then propose a method that identifies feature interactions within the service chains. Specifically, we characterize the degree of deviation of every service chain by evaluating the gap between expected and actual service states. An experimental evaluation demonstrates that the proposed method successfully detects 11 service chains and 6 feature interactions within 7 practical sensor-driven services.

Extracting service candidates from procedural programs based on process dependency analysis

To support legacy migration to service-oriented architecture (SOA), this paper presents a method that derives candidates of SOA services from procedural programs. In SOA, every service is supposed to be a process (procedure) with (1) open interface, (2) self-containedness, (3) coarse granularity for business. We identify such services from the source code and its data flow diagram (DFD), by analyzing data and control dependency among processes. Specifically, we first obtain the DFD with reverse-engineering techniques. For each layer of the DFD we classify every data flow into three categories. Using the data category and control among procedures, we characterize four types of dependency. We finally apply six rules that aggregate mutually dependent procedures and extract them as a service. A case study with a liquor shop inventory control system extracts service candidates with various granularity.

Using formal methods to increase confidence in a home network system implementation: a case study

A home network system consists of multiple networked appliances, intended to provide more convenient and comfortable living for home users. Before being deployed, one has to guarantee the correctness, the safety, and the security of the system. Here, we present the approach chosen to validate the Java implementation of a home network system. We rely on the Java Modelling Language to formally specify and validate an abstraction of the system.

Implementing Multi-Vendor Home Network System with Vendor-Neutral Services and Dynamic Service Binding

The home network system (HNS) consists of networked household appliances, intended to provide value-added services. The conventional HNS has been built on the single- vendor system, which severely limits potential of the HNS. To overcome the problem, this paper presents a method that constructs the HNS with multi-vendor appliances. The proposed method first defines vendor-neutral standard services, with which various HNS applications and services are developed. Then, we exploit a dynamic service binding mechanism, which binds each standard service on a vendor- specific API of an appliance during run-time. With this mechanism, common HNS applications and services can be achieved by various combinations of multi-vendor appliances. Moreover, replacing any appliance with another never affects the execution of the applications. We have implemented the proposed method using Apache Axis Web services and Rhino JavaScript engine. The experimental evaluation showed that our implementation works well for a practical HNS with sufficiently small overhead.

Fu-Rin-Ka-Zan: Quantitative Analysis of Developers' Characteristics Based on Project Historical Data

In software development, multiple developers compose a team, and perform a development process. To succeed the development, understanding developers characteristics are needed. However, there is no way to measure the characteristics of the developer by quantitative analysis. Therefore, a project manager often assigns team members based on his/her own sense, considering experience year of them. This paper proposes a measurement framework for developers characteristics using quantitative historical data. The proposed framework is useful for measuring developers characteristics and building a team based on quantitative analysis. We collected historical data from a PBL project for graduate students, and applied our framework to the collected data. As a result, our framework measured developers characteristics suitability in our subjective view.

On detecting service chains in sensor-driven home network services

When multiple sensor-driven services are deployed in the same environment, execution of a service may trigger other services, successively. Such chain reactions of services often cause undesirable feature interactions. This paper presents a framework that can characterize and detect the service chains within the home network system (HNS). We first introduce the ECA rules to describe the services, and then propose an environment effect model to capture how each device in the HNS affects the environment. Finally, we develop an algorithm that detects the service chains with concrete enabling conditions. A case study with 7 practical services shows that the proposed method successfully detects 11 service chains, in which 6 harmful feature interactions are identified.