Search

Abstract Textual content of citations has been widely applied to access the impact of research outputs, observe knowledge transfer across domains and generate summaries of research articles. These applications, however, experience the problem of correct identification of citational context windows, and alternately use text in a fixed-size window around the citation instance. Citation contexts might contain terms not referenced by the target citation. When application requirement is only the reference terms, this observation ought to be taken into account. This paper tries to find out whether the reference terms of a target citation can be represented as topics derived from its citation contexts or not. We, specifically, consider cases where complete statements are referred to by a citation and cases where there are no explicit reference terms. Evaluation results show that in a majority of cases, reference terms of a citation can be replaced with topics generated from its contexts. [ABSTRACT FROM AUTHOR]

The existing research is about location tracking either completely indoor or altogether on open air by utilizing various sensors and procedures based on inter-networking or internet of things. A limited arrangement has been proposed to embrace a single location area detecting innovation that based on both environments using physical devices. This paper intends to track a client position in both indoor and open-air situations by utilizing a single remote sensor with negligible subsequent bumble. The huge majority of the existing localization arrangements use additional equipment supplies to accomplish higher limitation precision. Received signal strength indicator (RSSI) was the standard element of remote sensors; the localization methods won't have any extra equipment. To overcome the problems, we need to study how to grow more precision distance for the strength of the signal model and 3dimensional measurement techniques. The Path Loss Exponent (PLE) determines the separation of RSS. In this event that RSS diminish at the separation, PLE will measure the rate; its esteem depends on the extraordinary proliferation environment. For positioning, the system requires location data. Most part use two-dimensional positioning as a part of the system with moving toward localization with three-dimensional positioning with weighted centroid restriction calculation as a proposed work. Here considers the triangular pyramid to perform three-dimensional positioning instead of three circle intersection method, in which the pyramid is divided into four two-dimensional surfaces. It consists of different parts of the environment where transmitter and receiver’s parts have the different PLE when compared to others existing research. Many researchers used two-dimensional positioning system, however, some circumstances we have to go through in three-dimensional position system. So, the proposed idea is basically based on the three-dimensional centroid limitation of systems. [ABSTRACT FROM AUTHOR]

This paper presents a model for improving query performance in Virtual Data Warehouse (VDW) by simulating VDW environment on a cellular phone billing and customer care system which involves processing millions of Call Detail Records (CDRs) generated by thousands of counters across the country. Processing aggregations on millions of CDRs requires expensive systems, especially when analysing customers’ traffic trends and encompasses several performance optimization techniques used for improvement of query performance in VDW. In this regard, VDW offers several advantages such as real-time analytic reports, reduced maintenance, low cost solution and flexible data integration, but performance is still one of its critical shortcomings. This paper enhances performance of VDW by using techniques like partitioned materialized views, index performance optimization, query rewrite in materialized views, analytic functions, sub-queries and enabling parallel execution etc. The study uses Oracle 10 g as a backend database; Oracle management console and SQL query analyser are used for monitoring performance concerns during validation of VDW model; standard PL/SQL developer is used for extracting and loading test data; and finally, Hyperion Development suite is used for testing time comparisons of datasets both in normal OLTP and simulated VDW environments. [ABSTRACT FROM AUTHOR]