Balaji Rajendran

Socio-contextual Filters for Discovering Similar Knowledge-Gathering Tasks in Generic Information Systems

The task of knowledge-gathering through an Information System has become increasingly challenging, due to the multitude of activities that are being facilitated through the system. A user-centric approach is presented in this paper where various activities executed by a user, for a knowledge-gathering task, are mapped to certain cognitive states in our model and the transitions between those states are used to indicate the progress made by the user. We propose a socio-contextual filtering algorithm for discovering similar tasks that were executed by other users and claim that such a socio-contextually related task would help in reducing the cognitive load, efforts and the time required for a user, naive to a given knowledge gathering task. We demonstrate this through the fewer number of state transitions that occur in our model for a guided user.

Contextual Strategies for Detecting Spam in Academic Portals

The emergence of social networking platforms in online space and its ever increasing user base has opened up a new arena for the spammers to exploit. Spam, in these kinds of platforms and such other interactive tools like forums, instant messaging, could be created easily and difficult to stop it from spreading, which necessitates the development of better detection strategies. In this paper, we present a contextual strategy for detecting spam in a restricted domain such as an academic portal. The proposed method uses the relationship between the concepts of the domain and the concepts of the individual message fragments to determine the relevancy of the message to the given context and marks the outliers. The strategy has been tested using a prototype system which had networking and interactive features for the participants to share information, and the results indicated that the contextual strategy was fairly successful in detecting spam.

Socio-contextual model of knowledge sharing among user assisting agents

A goal of user assisting agents is to provide effective assistance to their users in their tasks. The problem becomes challenging when the users are involved in a cognitive activity such as a knowledge gathering task through the web, and when the agents cannot clearly know the task of their users in advance. The challenge grows when the agent themselves do not possess the knowledge required for such tasks. It is in this scenario that we propose a socio-contextual model of knowledge sharing within a community of user assisting agents, provided through the environment of an open domain knowledge portal, which are used by the users for their web based knowledge gathering tasks. The agents involve themselves in a knowledge sharing exercise by implementing the socio-contextual model that may allow each agent to gain knowledge about a task of their interest from other fellow agents through which they can assist their respective users. We evaluate our model through an experiment involving many knowledge gathering tasks from diverse domains and the results indicate interesting implications of the model with respect to the agents, their tasks and their community.

Socio-contextual Network Mining for User Assistance in Web-based Knowledge Gathering Tasks

Web-based Knowledge Gathering (WKG) is a specialized and complex information seeking task carried out by many users on the web, for their various learning, and decision-making requirements. We construct a contextual semantic structure by observing the actions of the users involved in WKG task, in order to gain an understanding of their task and requirement. We also build a knowledge warehouse in the form of a master Semantic Link Network (SLX) that accommodates and assimilates all the contextual semantic structures. This master SLX, which is a socio-contextual network, is then mined to provide contextual inputs to the current users through their agents. We validated our approach through experiments and analyzed the benefits to the users in terms of resource explorations and the time saved. The results are positive enough to motivate us to implement in a larger scale.

FOSS Solutions for Community Development

The use of ICT solutions in developmental activities and the deployment of them in modern telecentres have been widely accepted phenomena. In this article, we examine the use of FOSS – Free/Open source software in ICT solutions and categorize them into two types: ‘FOSS in’ and ‘FOSS for’, in order to emphasize the difference between those solutions, where FOSS play a meager or symbolic role and those solutions that are specifically developed to address a community need. We discuss about our experience in developing a FOSS-based ICT solution that is specifically built to address the needs of a community information system. We illustrate our three-tiered multi-stakeholder model of development, deployment and usage. We also describe our experience in the process of development and pilot deployment of our solution in various locations of India, and emphasize on the synergies between our model and the FOSS mode of ICT solution development.

Digital Tokens: A Scheme for Enabling Trust Between Customers and Electronic Marketplaces

In electronic marketplaces, when the supply of a particular product is limited, and when there is a huge demand for the same, the questions of transparency and integrity prop up. We propose Digital Tokens—defined using proven cryptographic techniques—as a mechanism to assure trust for customers, and issued by a reliable, transparent and third-party intermediary, called digital token service provider (DTSP). The digital tokens are issued to a customer on behalf of a vendor and could be authenticated by both Vendor and the DTSP. This paper details the architecture involving the DTSP, protocols for communication, implementation details, the potential uses and benefits of the system and performance evaluation of such a system.

A Survey of Storage Remote Replication Software

Data availability at all times is the key requirement for major business operations and the financial impact of business outage could be disastrous (see table 1). There could be many majors threats to information availability, be it natural disasters, unplanned or planned occurrences. Remote replication is the process of creating copies of data known as replica at geographically distance location referred to as remote sites. With the advent of intelligent storage systems, there had been quite a pool of replication technologies emerging to avert business outage. In this paper we would describe important concepts of business continuity and remote replication mechanisms along with a study of major remote replication software available from major players in the storage technology.

An intelligent contextual support system for intrusion detection tasks

The cognitively challenging task of Intrusion Detection, undertaken by Network Security Engineers (NSE) even through a sophisticated Intrusion Detection System (IDS), demands considerable effort and time in figuring out the real intrusion attempts. We believe that a support system providing contextual assistance in the task of Intrusion detection would be of immense help to the Network Security Engineers. In this paper, we discuss about the features and mechanisms of such an intelligent contextual support system, along with the implementation and quantitative evaluation of it by using certain observable parameters of performance. The results indicated improvement in the productivity of the NSE and also the usability of the IDS.