An offshoot of data mining is the concept of opinion mining which aims to extract opinions from text documents. This concept is also referred to as sentiment analysis, sentiment classification, opinion extraction, etc. While text mining revolves around extracting factual information, opinion mining is not only concerned with the topics or subjects a document handles but also with the opinions expressed on such topics. This deals with aspects such as sentiment analysis or classification and attempts to extract opinions from the statements and expressions that appear in documents. Identifying the sources for opinion mining is also a concern as it has to handle large volumes of documents before it narrows down the relevant documents. Researchers have identified in the past several linguistic rules to identify words that express opinions and their semantic orientations which reveal whether the opinions extracted are positive, negative or neutral. The authors, Shishir Shandilya and Suresh Jain, in their paper, “Automatic Extraction and Classification of Opinions of Product Reviews from Web Documents”, have proposed to develop a system for some of the tasks involved in opinion mining, such as word segmentation, morphological analysis and automatic knowledge acquisition. The designed system has components such as crawlers and classifiers which are used to extract domain-specific data and uses linguistic rules to perform sentiment analysis.

The developments in wireless mobile communication technologies and protocols have enabled the implementation of various mobile vehicular ad hoc networks. For example, we have seen in the past wireless on-board units on the vehicles that allow each other to communicate with or without road-side units using short-range communication protocols using transceivers and transponders. The vehicular ad hoc networks help to implement applications such as safety and traffic monitoring applications, intelligent transport systems and electronic fee collections. While implementing such networks, the issues related to connectivity, bandwidth, link stability, route failure tolerance, etc. need to be addressed. Ensuring Quality of Services (QoS) has drawn the attention of the researchers while providing support for mobile users with real-time and multimedia applications in mobile ad hoc networks. The authors, B R Arun Kumar, Lokanatha C Reddy and Prakash S Hiremath, in their research paper, “Cross-Layer Design for Quality of Service Multicasting in Mobile Ad Hoc Networks”, have investigated the issues which significantly affect the multicast routing, using graph theory and MATLAB software. The investigation has resulted in a cross-layer design for such protocol that may exploit the dependence across the traditional network layers to obtain performance gains in QoS-enabled networks.

Vehicular ad hoc networks allow vehicles to disseminate messages such as road conditions to other vehicles, and the trustworthiness of these messages will increase the safety of the transportation system. While care must be exerted to ensure that vehicle-generated messages do not convey inaccurate or false content, it is also essential that the communication protocols are used to deliver the messages in time and with high reliability. Vehicular communication demands some of the requirements that are quite different from the conventional mobile networks, and hence requires the need to develop new communication methodologies and protocols. A number of MAC protocols were proposed in the literature for wireless networks and the performance of these protocols was compared using various parameters, such as delay, which defines the average time spent by the message in the queue; throughput, which defines the channel capacity used for message transmission; stability, which ensures the working of the communication system when high load is faced by MAC; and robustness against channel fading. Ranjeet Singh Tomar and Shekhar Verma, in their paper, “Vehicular Communication for Enhancing Safety in Transportation System”, propose a MAC protocol to minimize channel allocation time and overhead management in vehicular networks.

Component-based software development is aimed at developing reusable components that interact with each other while implementing the required functionalities in a software system. Components were created and maintained in binary format, and they are conceived as independent and reusable building blocks to provide a unique service which may be used with similar services provided by other components. Though a component-based approach differs from that of object-oriented framework, the current component technologies such as COM, CORBA and Enterprise JavaBeans are all built upon object-oriented programming. This emphasizes the need to investigate the techniques of modeling, design and verification of object systems and the construction of component systems on underlying object systems. It is well known that the Unified Modeling Language (UML) is used as a graphical language for depicting, specifying, constructing, and documenting the object-oriented models. UML has been used for expressing models of Object-Oriented Analysis and Design (OOAD) under the auspices of the Object Management Group (OMG), and currently it is the de facto standard for object-oriented modeling. Gufran Ahmad Ansari, Samir A Elsagheer Mohamed and Master Prince Syed, in their paper, “Modeling of a Component-Based Software Through UML Using Object-Oriented Approach”, have proposed a UML class model for developing a component-based software.

-- M Raja
Consulting Editor