Due to freedom to express views, opinions, news, etc and easier method to disseminate the information to large population worldwide, social media platforms are inundated with big streaming data characterized by both short text and long normal text. Getting the glimpse of ongoing events happening over social media is quintessential from the viewpoint of understanding the trends, and for this, topic modeling is the most important step. With reference to increase in proliferation of big data streaming from social media platforms, it is crucial to perform large scale topic modeling to extract the topics dynamically in an online manner. This paper proposes an adaptive framework for dynamic topic modeling from big data using deep learning approach. Approach based on approximation of online latent semantic indexing constrained by regularization has been put forth. The model is designed using deep network of feed forward layers. The framework works in an adaptive manner in the sense that model is extracts incrementally according to streaming data and retrieves dynamic topics. In order to get the trends and evolution of topics, the framework supports temporal topic modeling, and enables to detect implicit and explicit aspects from sentences also.

Big Data which is one of the newest technologies in the present field of science and technology has created an enormous drift of technology to a salient data architecture. The next thing that comes right after big data is Hadoop which has motivated the complete Big Data Environment to its jurisdiction and has reinforced the complete storage and analysis of big data. This paper discusses a hierarchical architecture of Hadoop Nodes namely Namenodes and Datanodes for maintaining a High Availability Hadoop Distributed File System. The High Availability Hadoop Distributed File System architecture establishes itself onto the two fundamental model of Hadoop that is Master-Slave Architecture and elimination of single point node failure. The architecture will be of such utilization that there will be an optimum load on the data nodes and moreover there will be no loss of any data in comparison to the size of data.

With the evolution of higher computing speed, efficient communication technologies, and advanced display techniques the legacy HCI techniques become obsolete and are no more helpful in accurate and fast flow of information in present day computing devices. Hence the need of user friendly human machine interfaces for real time interfaces for human computer interaction have to be designed and developed to make the man machine interaction more intuitive and user friendly. The vision based hand gesture recognition affords users with the ability to interact with computers in more natural and intuitive ways. These gesture recognition systems generally consist of three main modules like hand segmentation, hand tracking and gesture recognition from hand features, designed using different image processing techniques which are further integrated with different applications. An increase use of new interfaces based on hand gesture recognition designed to cope up with the computing devices for interaction. This paper is an effort to provide a comparative analysis between such real time vision based hand gesture recognition systems which are based on interaction using single and multiple hand gestures. Single hand gesture based recognition systems (SHGRS) have fewer complexes to implement, with a constraint to the count of different gestures which is large enough with various permutations and combinations of gesture, which is possible with multiple hands in multiple hand gesture recognition systems (MHGRS). The thorough comparative analysis has been done on various other vital parameters for the recognition systems.