Vivek Kumar Singh

Videopedia: Lecture Video Recommendation for Educational Blogs Using Topic Modeling

Two main sources of educational material for online learning are e-learning blogs like Wikipedia, Edublogs, etc., and online videos hosted on various sites like YouTube, Videolectures.net, etc. Students would benefit if both the text and videos are presented to them in an integrated platform. As the two types of systems are separately designed, the major challenge in leveraging both sources is how to obtain video materials, which are relevant to an e-learning blog. We aim to build a system that seamlessly integrates both the text-based blogs and online videos and recommends relevant videos for explaining the concepts given in a blog. Our algorithm uses content extracted from video transcripts generated by closed captions. We use topic modeling to map videos and blogs in the common semantic space of topics. After matching videos and blogs in the space of topics, videos with high similarity values are recommended for the blogs. The initial results are plausible and confirm the effectiveness of the proposed scheme.

Study of combined effects of zero copy and pre-processing on Input/Output performance of multimedia on demand streaming server

The growing broadband internet connectivity has made services like Media on Demand (MoD) and News on Demand (NoD) a huge business. Media servers play an important role and critical part of providing real time MoD services. The major challenges are the handling of multimedia files and their real time streaming, which require fast processing as well as transmission of multimedia by the server. One of the important features of media server design is Input/Output (I/O) performance of multimedia files to network. This performance also affects the overall delay in case of end-to-end QoS. In this paper, we have proposed an idea to improve the I/O performance by enhancing zero copy methodology with pre-processed multimedia. Our proposed ideas implementation results present five times reduction of total transmission and processing time delay in comparison to without zero copying and pre-processing in media server. Moreover, there is a 15% reduction in CPU usage of the multimedia server.

A content engagement score for online learning platforms

Engagement on online learning platforms is essential for user retention, learning, and performance. However, there is a paucity of research addressing latent engagement measurement using user activities. In this work in progress paper, we present a novel engagement score consisting of three sub-dimensions - cognitive engagement, emotional engagement, and behavioral engagement using a comprehensive set of user activities. We plan to evaluate our score on a large scale online learning platform and compare our score with measurements from a user survey-based engagement scale from the literature.

Estimating operating system process energy consumption in real time

The power consumption in data centers due to Information and Communication Technologies (ICT) is significant across the globe. With recent developments in Service Oriented Architecture (SOA), we notice a paradigm shift in computing. Desktops (PCs) and laptops are being replaced by smart phones and tablets. A major impact of this architecture is a shift of computing resources from personal desktops and laptops to centralized server farms. This implies increases in power consumption in the large-scale servers used in these infrastructures. In such a scenario, optimizing the IT resources for power consumption is a necessity. The first step of such an optimization at the application level is the knowledge of how much energy the application is consuming. A major challenge in this domain is to develop a software-based energy metering tool that can measure the energy consumptions at the OS process level. We have developed an OS process-level power metering tool that can accurately estimate the energy usage based on system resource usages, and demonstrated that our tool provides energy measurement for complex e-business applications with greater than 95% accuracy.

EPGLU: Enhanced PGLU timer for paging and location update optimization in IEEE 802.16m based WiMAX networks

Power saving in Mobile Wireless Broadband networks is one of the vibrant, important and current areas of research. Mobile WiMAX based on IEEE 802.16e standard provides different Power Saving Modes (PSM) like sleep mode and idle mode to save power in Subscriber Station/Mobile Station(s). The forthcoming IEEE 802.16m standard has extended idle mode with introduction of concept of primary and secondary paging groups with PGLU (Paging Group Location Update) timer for optimization of power saving during paging and location update. In this paper, authors have provided comparative study of power saving due to enhancements in IEEE 802.16m standard with respect to IEEE 802.16e especially in idle mode. In addition to this, authors have proposed a novel idea of Enhanced PGLU (EPGLU) timer to further optimize the paging and location update in 802.16m standard. The simulation results show up to 60 percent decrease in paging signalling overhead and a marginal increase in location update in Mobile Station (MS).

Study of latency in Enhanced-PGLU (EPGLU) timer based paging in IEEE 802.16m based WiMAX networks

Paging latency Is one of the Important parameters for determining the Quality of Service (QoS) required by the telecom and broadband applications. Most of the 4G applications (especially real time applications) require upper bound on paging latency. In this paper, authors have provided a novel algorithm for Enhanced-PGLU (EPGLU) to reduce the paging delay and have provided the comparative study of paging latency in EPGLU timer based paging with respect to the PGLU (Paging Group Location Update) timer based paging available in forth coming IEEE 802.16m standards. The results show 27% decrease in paging latency with EPGLU based proposed algorithm as compared to PGLU based paging provided in IEEE 802.16m standards.