Joonho Kwon

Scalable and efficient web services composition based on a relational database

Abstract: Recently, there has been growing interest in web services composition. Web services composition gives us a possibility to fulfil the user request when no single web service can satisfy the functionality required by the user. In this paper, we propose a new system called PSR for the scalable and efficient web services composition search using a relational database. In contrast to previous work, the PSR system pre-computes web services composition using joins and indices and also supports semantic matching of web services composition. We demonstrate that our pre-computing web services composition approach in RDBMS yields lower execution time for processing user queries despite of and shows good scalability when handling a large number of web services and user queries.

Non-redundant web services composition based on a two-phase algorithm

Recently, there has been growing interest in developing web services composition search systems. Current solutions have the drawback of including redundant web services in the results. In this paper, we proposed a non-redundant web services composition search system called NRC, which is based on a two-phase algorithm. In the NRC system, the Link Index is built over web services according to their connectivity. In the forward phase, the candidate compositions are efficiently found by searching the Link Index. In the backward phase, the candidate compositions decomposed into several non-redundant web services compositions by using the concept of tokens. Results of experiments involving data sets with different characteristics show the performance benefits of the NRC techniques in comparison to state-of-the-art composition approaches.

Fast XML document filtering by sequencing twig patterns

XML-enabled publish-subscribe (pub-sub) systems have emerged as an increasingly important tool for e-commerce and Internet applications. In a typical pub-sub system, subscribed users specify their interests in a profile expressed in the XPath language. Each new data content is then matched against the user profiles so that the content is delivered only to the interested subscribers. As the number of subscribed users and their profiles can grow very large, the scalability of the service is critical to the success of pub-sub systems. In this article, we propose a novel scalable filtering system called iFiST that transforms user profiles of a twig pattern expressed in XPath into sequences using the Prüfers method. Consequently, instead of breaking a twig pattern into multiple linear paths and matching them separately, FiST performs holistic matching of twig patterns with each incoming document in a bottom-up fashion. FiST organizes the sequences into a dynamic hash-based index for efficient filtering, and exploits the commonality among user profiles to enable shared processing during the filtering phase. We demonstrate that the holistic matching approach reduces filtering cost and memory consumption, thereby improving the scalability of FiST.

Extracting Trends of Traffic Congestion Using a NoSQL Database

Recently, there has been growing interest in monitoring the road traffic data. Most of the work focused on real-time traffic data. In this paper, we propose an Extrac system which extracts trend of traffic congestions from historical traffic data and answers the queries about the trends. In Extrac system, we first convert the historical traffic data into traffic patterns then summarize it by applying MapReduce style algorithms. These traffic patterns are store into a NoSQL database. Our implementation demonstrates the feasibility of Extrac for querying traffic information and generating the result.

A SIMULATION NETWORK MODEL TO EVALUATE RFID MIDDLEWARES

High-performance radio-frequency identification (RFID) is a challenging issue for large-scale enterprises. As a key component of an RFID system, RFID middleware is an important factor to measure the performance of the system. To evaluate the feasibility of an RFID middleware, the performance of the RFID middleware should be carefully evaluated in various RFID-enabled business environments. However, the construction of an RFID testbed requires a lot of time, money, and human resources because it involves numerous tagged items and a large number of deployed readers. We must provide a meaningful input tag stream representing various business activities, rather than random data. This paper presents a novel simulation model for the virtual construction of RFID testbeds. To ensure the semantic validity of the input tag stream, the proposed RFID simulation network (RSN) extends Petri nets by including sets of functions that represent unique characteristics of RFID environments such as the uncertainty of communications and tag movement patterns. By configuring appropriate functions, the RSN automatically generates an input tag stream that matches the distribution of real data. We demonstrate that the RSN model correctly reflects data from real-world environments by comparing input tag streams from real RFID equipment and from the RSN model.

Road Traffic Analytic Query Processing Based on a Timeline Modeling

Due to the growing number of collected traffic data, big data technology enables us to obtain in-depth analysis of road traffic data. For better understanding for traffic behaviors, the analyzed data can be provided in a timeline view. In this paper, we formally define a timeline model for traffic data and propose an algorithm for constructing a timeline data structure from raw traffic data. We also suggest traffic analytic query processing algorithms which utilize the timeline information. Experimental results demonstrate the efficiency of the proposed algorithm on real traffic datasets from a Busan Intelligent Transport System (ITS) center[1].

Computing traffic congestion degree using SNS-based graph structure

Social networking site (SNS) messages can contain subjective traffic information, including congestion-related expressions such as “bad traffic” or “traffic is crazy”. Moreover, they also contain heterogeneous levels of location information, such as a point (latitude, longitude), a road, or an area name, which complicates the process of collecting related traffic information. This paper aims to use SNS messages for monitoring traffic conditions on a road by computing the traffic congestion degree. The process begins by classifying those SNS messages that are related to a road in terms of location information and constructing an initial graph structure to store each message. Because of the heterogeneous location types, we need to combine the initial graph structures based on their spatial references. We can then measure the subjective congestion by computing an expression score using our rule-based approach.

Generation of RFID test datasets using RSN tool

This paper presents an RSN Tool to generate input datasets for testing RFID middleware. As RFID middleware takes an important role in entire RFID systems, its performance should be carefully evaluated under various business conditions. In general, evaluation of the RFID middleware requires a huge cost because the numerous RFID readers and tags need to be deployed to acquire the tag event stream. To facilitate low-cost testing of the middleware, we propose the RSN Tool which provides means of designing a virtual RFID infrastructure and generates a tag event stream automatically for the virtual infrastructure. Using the RSN Tool, we can easily obtain a semantically valid dataset, which captures both physical characteristics of RF communications and business activities of tags’ movements. This is a major differentiation point of our work compared from previous works, which merely create the randomized dataset based on a set of virtual RFID readers. We also discuss a step-by-step usage of the RSN Tool from the creation of a virtual infrastructure to the generation of tag events. The experimental analysis shows that the RSN Tool can create the near-real dataset, which closely reflects business activities of the real RFID infrastructure.

Querying and extracting timeline information from road traffic sensor data

The escalation of traffic congestion in urban cities has urged many countries to use intelligent transportation system (ITS) centers to collect historical traffic sensor data from multiple heterogeneous sources. By analyzing historical traffic data, we can obtain valuable insights into traffic behavior. Many existing applications have been proposed with limited analysis results because of the inability to cope with several types of analytical queries. In this paper, we propose the QET (querying and extracting timeline information) system—a novel analytical query processing method based on a timeline model for road traffic sensor data. To address query performance, we build a TQ-index (timeline query-index) that exploits spatio-temporal features of timeline modeling. We also propose an intuitive timeline visualization method to display congestion events obtained from specified query parameters. In addition, we demonstrate the benefit of our system through a performance evaluation using a Busan ITS dataset and a Seattle freeway dataset.

Querying Road Traffic Data from a Document Store

We present a novel system called Extrac for querying a large database of road traffic information. Such traffic data are collected from an ITS (Intelligent transportation systems) center of Busan and represents speed values of all road segments of Busan for every 5 minutes. Extrac stores the collected traffic data into a NoSQL document database and supports a traffic congestion queries. It adopts a suite of new approaches for (a) transformation of traffic data into pattern summaries based on a MapReduce framework, and (b) efficient congestion query processing which utilizes single value decomposition (SVD) of transformed matrices. We demonstrate the Extract systems using real traffic data of Busan metropolitan city.