Abstract

Ranking the search results is an important research problem in WWW. HITS, PageRank and variations of these algorithms are widely used approaches for ranking. In this paper we proposed a new ranking algorithm to rank the search results by introducing the concept of “leader-page”. The notion of “leader-page” is defined by extending the concept of “leader” from leadership theory. In leadership theory, a leader is defined as a person, who has more contacts with the other members of the group, both initiates and receives communication, and whose characteristics are the most similar to the group’s own characteristics. Similarly, in WWW, the proposed approach identifies leader-pages and assigns “leadership score” to them based on several kinds of cyclic and similarity relationships it establishes with other web pages. Given a set of key words (search query), the proposed approach ranks the related pages based on the corresponding “leadership score”. The experiment results show that the proposed approach gives high leadership score to resourceful pages as compared to the results of the HITS algorithm and the Google search engine

Abstract

With the rapid growth of the Internet, most of the textualdata in the form of newspapers, magazines and journals tend to be avail-able on-line. Summarizing these texts can aid the users access the in-formation content at a faster pace. However, doing this task manuallyis expensive and time-consuming. Automatic text summarization is asolution for dealing with this problem. For a given text, a text summa-rization algorithm selects a few salient sentences based on certain fea-tures. In the literature, weight-based, foci-based, and machine learningapproaches have been proposed. In this paper, we propose a popularity-based approach for text summarization. A popularity of the sentence isdetermined based on the number of other sentences similar to it. Usingthe notion of popularity, it is possible to extract potential sentences forsummarization that could not be extracted by the existing approaches.The experimental results show that by applying both popularity andweight-based criteria it is possible to extract effective summaries.

Abstract

No Results Found

Abstract

The growing trend towards commercial crops with ever changing technology packages necessitate a transformation in the existing agricultural extension system with emerging technologies like IT.In this context IIIT , Hyderabad has developed and implemented eSagu, a Web Based Agricultural Expert Advice Dissemination System in Atmakur mandal, Warangal district during 2004-05 covering 1,051 cotton farms. eSagu system of extension has succeeded in acquisition of infrastructure and human resources as envisaged and proved its organizational ability in coordinating different branches for a smooth functioning on proposed lines. Wholistic information in the form of clear images of crop (through digital photographs and zooming) and other information i.e. soil, weather, crop history has helped scientists to provide effective advice. The system was able to offer collective expert advice from one place with in 24 hours of response time to the farmers at the other end. Further, it is only the information that moved, while the farmers and scientists remained at their respective working places. The system has proved its technical efficiency in terms of pest identification and prediction with appropriate advices based on IPM practices to all sections of farmers. In the process of information dissemination, the system was able to provide 20,000 advices and accumulated 1,11,000 crop photographs in a period of one year. In the field of research it was able to identify new pests like stem borer and early detection of Gray Mildew disease and share the information with other research agencies. As an experiment, the project has tested feasibility and acceptability of IT for tapping its potential as an alternative to the existing extension system. It also offers scope for further reduction in the cost of delivery of advice provided a cluster based approach is adopted in the future

Abstract

In view of technology/extension gaps in Indian agriculture and to exploit ICT revolution, International Institute of Information Technology, Hyderabad, A.P., India had developed eSagu model of extension system and implemented it for the cotton crop in three villages of Oorugonda, Gudeppad and Oglapur covering 749 farmers and 1041 farms during 2004-05 crop season. The main objective is to build a cost effective and scalable agricultural expert advice dissemination system to all the farmers. The three-tier system consists of farmers as end users, coordinators as intermediaries to obtain crop status through digital photographs and text and communicate the advice to the farmers. The scientists with knowledge system prepare farm advices. The evaluation study clearly brings out the feasibility and acceptability of ICT based model of eSagu. The crop status, the zooming facility of digital photographs, weather data, crop history, soil data, all have enabled the scientists in advice making and prediction. The advices were provided to the farmers in 24-36 hours. eSagu operation has improved the access, knowledge and technology adoption rate as compared to their counter parts in non-project area. It further established that the acquired technology has been quite useful and reflected in increased production by 1.5 quintals per acre, saved fertilizers by 0.76 bags, pesticides by 2.3 sprays per acre. This has resulted in a net gain of Rs. 3820.0 per acre. Farmers’ response to the utility of the project is quite positive and they want the project to continue in the future. In conclusion, the evaluation study reveals that accomplishments of the project are impressive, the system is generally accepted and throws a promise for wider application in the future.

Abstract

Joins on set-valued attributes (set joins) have numerous database applications. In this paper we propose PRETTI (PREfix Tree based seT joIn) – a suite of set join algorithms for containment, overlap and equality join predicates.Our algorithms use prefix trees and inverted indices. These structures are constructed on-the-fly if they are not already precomputed. This feature makes ouralgorithms usable for relations without indices and when joining intermediate results during join queries with more than two relations. Another feature of ouralgorithms is that results are output continuously during their execution and notjust at the end. Experiments on real life datasets show that the total executiontime of our algorithms is significantly less than that of previous approaches, even when the indices required by our algorithms are not precomputed.

Abstract

No Results Found

Abstract

In this paper, we present studies on combining evidence from multiple classifiers to recognize a large number of consonant-vowel (CV) unit.s of speech. Multiple classi­fier systems may lead to a better solution to the complex speech recognition tasks, when the evidence obtained from individual systems is complementary in nature. Hidden Markov models (HIvlMs) are based on the max­imum likelihood (ML) approach for training CV pat­terns of variable lengt.h. Support vector machine (SVM) models are based on discriminative learning approach for training fixed length CV patterns. Because of the differences in the training methods and in the pattern representation used, they may provide complementary evidence for CV classes. Complement.ary evidence avail­able from these classifiers is combined using the sum rule. Effectiveness of the multiple classifier system is demonstrated for recognition of CV units of speech in Indian languages .

Abstract

Multilingual speech recognition system is required for tasksthat use several languages in one speech recognition application. In thispaper, we propose an approach for multilingual speech recognition byspotting consonant-vowel (CV) units. The important features of spottingapproach are that there is no need for automatic segmentation of speechand it is not necessary to use models for higher level units to recognisethe CV units. The main issues in spotting multilingual CV units are thelocation of anchor points and labeling the regions around these anchorpoints using suitable classifiers. The vowel onset points (VOPs) have beenused as anchor points. The distribution capturing ability of autoassocia-tive neural network (AANN) models is explored for detection of VOPsin continuous speech. We explore classification models such as supportvector machines (SVMs) which are capable of discriminating confusableclasses of CV units and generalisation from limited amount of trainingdata. The data for similar CV units across languages are shared to trainthe classifiers for recognition of CV units of speech in multiple languages.We study the spotting approach for recognition of a large number of CVunits in the broadcast news corpus of three Indian languages

Abstract

India has witnessed 6 earthquakes during past 15 years causing around 40,000 casualties and immense property loss. More than 90% of these casualties are due to collapse of buildings and they occurred within very short period of time after the earthquake. Hence, for reducing casualty losses and subsequent consequences due to earthquake, it is very important to study the response of structures due to seismic excitations. For studying the response of the structures and consequent issues after the earthquake disaster, information technology can play a major role. This paper contributes to understand the importance of information technology in earthquake disaster mitigation