This paper proposes a number of explicit and implicit ratings in product recommendation system fo... more This paper proposes a number of explicit and implicit ratings in product recommendation system for Business-to-customer ecommerce purposes. The system recommends the products to a new user. It depends on the purchase pattern of previous users whose purchase pattern is close to that of a user who asks for a recommendation. The system is based on weighted cosine similarity measure to find out the closest user profile among the profiles of all users in database. It also implements Association rule mining rule in recommending the products. Also, this product recommendation system takes into consideration the time of transaction of purchasing the items, thus eliminating sequence recognition problem. Experimental result shows for implicit rating, the proposed method gives acceptable performance in recommending the products. It also shows introduction of association rule improves the performance measure of recommendation system.
International Journal of Intelligent Information and Database Systems, 2009
ABSTRACT An important component of any web search engine is its crawler, which is also known as r... more ABSTRACT An important component of any web search engine is its crawler, which is also known as robot or spider. An efficient set of crawlers make any search engine more powerful, apart from its other measures of performance, such as its ranking algorithm, storage mechanism, indexing techniques, etc. In this paper, we have proposed an extended technique for crawling over the World Wide Web (WWW) on behalf of a search engine. This is an approach with multiple crawlers working in parallel combined with the mechanism of focused crawling (Chakrabarti et al., 1999a, 2002; Mukhopadhyay et al., 2006). In this approach, the total structure of any website is divided into several number of levels based on the hyperlink-structure for downloading web pages from that website (Chakrabarti et al., 1999b; Mukhopadhyay and Singh, 2004). The number of crawlers of each level is not fixed, rather dynamic in this context. It is determined at execution time on demand basis using threaded program based on the number of hyperlinks of a specific web page. This paper also proposes a focused hierarchical crawling technique, where crawlers are created dynamically at runtime for different domains to crawl the web pages with the essence of resource sharing.
In this report, We investigated the nonlinear phenomena in the study of flow dynamics of velocity... more In this report, We investigated the nonlinear phenomena in the study of flow dynamics of velocity component. In our studies, we observed nonlinear term in the vertical component of velocity by vittori et al [7] The time series simulation of vertical component which increases with Reynold stress value. We developed direct numerical simulation under two dimensional grid system to study the flow dynamics and vorticity parameter. Flow pattern and flow dynamics near wavy boundary wall in the victinity of ripple bottom was readdressed under direct numerical simualtion(DNS) framework. Both horizontal and vertical component of fluid velocity were studied under pulating force of flow. Vorticity is calculated under complex framework by taking into higher order interaction term. we tried to carry out similar simulation with same particle ejection in the viscous bed using DNS simulation for pulsating flow. Our focus was to observe particle motion using DNS simulation and study the particle phase under vortex structures formed here.
Thin Film Technologies and Special Applications, 1982
ABSTRACT Amorphous Ge-SisH alloys have been prepared by r.f. sputtering in an argon-hydrogen atmo... more ABSTRACT Amorphous Ge-SisH alloys have been prepared by r.f. sputtering in an argon-hydrogen atmosphere of varying hydrogen partial pressure and substrate temperature. The correlated data of the vibrational properties as measured by IR absorption and Raman scattering on these samples reveal several hydrogenic and lattice modes. The hydrogenic modes which grow with increasing hydrogenation can be identified with various Ge (Si)-H type molecular modes. The absorption due to Si-H modes appears stronger than that due to Ge-H modes, indicating a pre-ferential H attachment to Si atoms. Dehydrogenation by annealing at successively high temperatures leads to selective reduction/disappearance of certain modes assigned to Ge (Si)-H vibrations. Also, the observed variation (0.1 to 0.7) in the depolarisation ratio for the Raman spectra of asdeposited samples become stabilised at 0.5 with annealing. These data are interpreted in terms of the film growth process and associated structural inhomogeneities and H- bonding configurations.
A new amorphous semiconductor alloy system Si"C"F~has been prepared by rf sputtering of polycryst... more A new amorphous semiconductor alloy system Si"C"F~has been prepared by rf sputtering of polycrystalline SiC in an Ar+SiF4 atmosphere. Dark conductivity and optical absorption of thin films are measured as functions of F concentration. Infrared spectra indicate a preferential attachment of fluorine to carbon over silicon. The bonded fluorine concentration is estimated to be as high as 40 at. %. The principal reststrahlen band shifts to higher frequencies and appears to sharpen with the increase of fluorine concentration. Fluorinated films are observed to be resistant to high-temperature annealing.
Proceedings - 10th International Conference on Information Technology, ICIT 2007, 2007
In this paper, an approach for storing Markov tree, used in various versions of PPM model while p... more In this paper, an approach for storing Markov tree, used in various versions of PPM model while predicting next Web-page is proposed. Markov tree requires huge amount of memory. This problem is solved using Cellular Automata which is considered as a fast and inexpensive mechanism. The proposed technique utilizes non-linear Single Cycle Multiple Attractor Cellular Automata (SMACA) which replaces Markov tree for minimizing the memory requirement.
Proceedings - 2007 International Symposium on Information Technology Convergence, ISITC 2007, 2007
Abstract Search for any topic or a word using Web search engine depends on its ranking mechanism.... more Abstract Search for any topic or a word using Web search engine depends on its ranking mechanism. Generally, search engine sorts through the millions of Web pages and then present the significant Web-pages that match search topic of the user. These matches will be further ranked, so that the most relevant ones come first. This paper proposes an alternate way to rank hyper-linked Web page through analysis of link structure. This link structure is stored in an efficient way to minimize the storage space using Galois Extension Field GF ( ...
Journal of Convergence Information Technology, 2009
Abstract Markov Models have been widely used for predicting next Web-page from the users&... more Abstract Markov Models have been widely used for predicting next Web-page from the users' navigational behavior recorded in the Web-log. This usage-based technique can be combined with the structural properties of the Web-pages to achieve better prediction accuracy. This paper proposes one of the pre-fetching techniques relying both on Markov Model and Ranking which considers the structural properties of the Web. In this paper, prediction accuracy is realized as a linear function of transition probability of first order ...
Journal of Convergence Information Technology, 2009
Clustering of non-homogenous documents has become an increasing challenge and opportunity with th... more Clustering of non-homogenous documents has become an increasing challenge and opportunity with the huge proliferation of World Wide Web. It has become difficult to retrieve the desired information without proper clustering of Web-page with the increase in information on the WWW. Several new ideas have been proposed in recent years. Among them partitioning approach is still widely used clustering approach for its simplicity. This paper proposes a partitioning approach to cluster the Web-page based on information provided by the hyperlink structure of Web-pages and also by the content of the Web-pages. The proposed approach of Web-page clustering exhibits better result than K-medoid partitioning clustering approach as the centroids are chosen by HITS Algorithm. The partitioning approach like Kmediod, K-means require number of clusters apriori. It has been observed that the performance of these approaches depend on the initial selection centroids of the clusters. These two problems have been solved by the approach proposed in this paper. Experimental result supports our approach as better concept.
Web Search Engine uses forward indexing and inverted indexing as a part of its functional design.... more Web Search Engine uses forward indexing and inverted indexing as a part of its functional design. This indexing mechanism helps retrieving data from the database based on user query. In this paper, an efficient solution to handle the indexing problem is proposed with the introduction of Nonlinear Single Cycle Multiple Attractor Cellular Automata (SMACA). This work simultaneously shows generation of
2008 International Conference on Information Technology, 2008
This paper presents an alternative product recommendation system for Business-to-customer ecommer... more This paper presents an alternative product recommendation system for Business-to-customer ecommerce purposes. The system recommends the products to a new user. It depends on the purchase pattern of previous users whose purchase pattern are close to that of new user. The system is based on vector space model to find out the closest user profile among the profiles of all users in database. It also implements Association rule mining based recommendation system, taking into consideration the order of purchase, in recommending more than one product. To make the association rule memoryefficient, cellular automata is used.
9th International Conference on Information Technology (ICIT'06), 2006
Search for a Web-page using Crawler based Search Engine depends on its Ranking Mechanism. Search ... more Search for a Web-page using Crawler based Search Engine depends on its Ranking Mechanism. Search Engine sorts through millions of Web-pages and present the significant Web-pages through Ranking method. A new approach is proposed here adopting Cellular Automata.
Radiation Effects and Defects in Solids - RADIAT EFF DEFECT SOLID, 1983
Thin films of RF sputtered amorphous germanium, silicon and silicon carbide have been irradiated ... more Thin films of RF sputtered amorphous germanium, silicon and silicon carbide have been irradiated in a reactor with fast-neutron fluence of up to 3.6 × 10 n/cm at a temperature not exceeding 50°C. The irradiation induced changes in electrical and structural properties have been studied. The effect of irradiation on dark conductivity is appreciable though the activation energy gap and phonon spectra do not seem to undergo much change. The results could be explained as a combined effect of two processes (i) the rearrangement of the atoms of the amorphous medium in the presence of neutron-fluence to produce a more complete amorphous network, and (ii) the production of defect clusters in certain regions of the amorphous medium, giving rise to isolated zones with high concentration of dangling bonds. Our results, however, indicate that in the amorphous thin films studied, the first process is predominant throughout the volume.
ABSTRACT The d.c. conductivity of RF sputtered hydrogenated amorphous SiC films have been studied... more ABSTRACT The d.c. conductivity of RF sputtered hydrogenated amorphous SiC films have been studied in the temperature range of 350–130 K. High temperature data yield activation energy gap which increases significantly with hydrogen content, while the low temperature conductivity is explained by hopping between localized states.
This paper proposes a number of explicit and implicit ratings in product recommendation system fo... more This paper proposes a number of explicit and implicit ratings in product recommendation system for Business-to-customer ecommerce purposes. The system recommends the products to a new user. It depends on the purchase pattern of previous users whose purchase pattern is close to that of a user who asks for a recommendation. The system is based on weighted cosine similarity measure to find out the closest user profile among the profiles of all users in database. It also implements Association rule mining rule in recommending the products. Also, this product recommendation system takes into consideration the time of transaction of purchasing the items, thus eliminating sequence recognition problem. Experimental result shows for implicit rating, the proposed method gives acceptable performance in recommending the products. It also shows introduction of association rule improves the performance measure of recommendation system.
International Journal of Intelligent Information and Database Systems, 2009
ABSTRACT An important component of any web search engine is its crawler, which is also known as r... more ABSTRACT An important component of any web search engine is its crawler, which is also known as robot or spider. An efficient set of crawlers make any search engine more powerful, apart from its other measures of performance, such as its ranking algorithm, storage mechanism, indexing techniques, etc. In this paper, we have proposed an extended technique for crawling over the World Wide Web (WWW) on behalf of a search engine. This is an approach with multiple crawlers working in parallel combined with the mechanism of focused crawling (Chakrabarti et al., 1999a, 2002; Mukhopadhyay et al., 2006). In this approach, the total structure of any website is divided into several number of levels based on the hyperlink-structure for downloading web pages from that website (Chakrabarti et al., 1999b; Mukhopadhyay and Singh, 2004). The number of crawlers of each level is not fixed, rather dynamic in this context. It is determined at execution time on demand basis using threaded program based on the number of hyperlinks of a specific web page. This paper also proposes a focused hierarchical crawling technique, where crawlers are created dynamically at runtime for different domains to crawl the web pages with the essence of resource sharing.
In this report, We investigated the nonlinear phenomena in the study of flow dynamics of velocity... more In this report, We investigated the nonlinear phenomena in the study of flow dynamics of velocity component. In our studies, we observed nonlinear term in the vertical component of velocity by vittori et al [7] The time series simulation of vertical component which increases with Reynold stress value. We developed direct numerical simulation under two dimensional grid system to study the flow dynamics and vorticity parameter. Flow pattern and flow dynamics near wavy boundary wall in the victinity of ripple bottom was readdressed under direct numerical simualtion(DNS) framework. Both horizontal and vertical component of fluid velocity were studied under pulating force of flow. Vorticity is calculated under complex framework by taking into higher order interaction term. we tried to carry out similar simulation with same particle ejection in the viscous bed using DNS simulation for pulsating flow. Our focus was to observe particle motion using DNS simulation and study the particle phase under vortex structures formed here.
Thin Film Technologies and Special Applications, 1982
ABSTRACT Amorphous Ge-SisH alloys have been prepared by r.f. sputtering in an argon-hydrogen atmo... more ABSTRACT Amorphous Ge-SisH alloys have been prepared by r.f. sputtering in an argon-hydrogen atmosphere of varying hydrogen partial pressure and substrate temperature. The correlated data of the vibrational properties as measured by IR absorption and Raman scattering on these samples reveal several hydrogenic and lattice modes. The hydrogenic modes which grow with increasing hydrogenation can be identified with various Ge (Si)-H type molecular modes. The absorption due to Si-H modes appears stronger than that due to Ge-H modes, indicating a pre-ferential H attachment to Si atoms. Dehydrogenation by annealing at successively high temperatures leads to selective reduction/disappearance of certain modes assigned to Ge (Si)-H vibrations. Also, the observed variation (0.1 to 0.7) in the depolarisation ratio for the Raman spectra of asdeposited samples become stabilised at 0.5 with annealing. These data are interpreted in terms of the film growth process and associated structural inhomogeneities and H- bonding configurations.
A new amorphous semiconductor alloy system Si"C"F~has been prepared by rf sputtering of polycryst... more A new amorphous semiconductor alloy system Si"C"F~has been prepared by rf sputtering of polycrystalline SiC in an Ar+SiF4 atmosphere. Dark conductivity and optical absorption of thin films are measured as functions of F concentration. Infrared spectra indicate a preferential attachment of fluorine to carbon over silicon. The bonded fluorine concentration is estimated to be as high as 40 at. %. The principal reststrahlen band shifts to higher frequencies and appears to sharpen with the increase of fluorine concentration. Fluorinated films are observed to be resistant to high-temperature annealing.
Proceedings - 10th International Conference on Information Technology, ICIT 2007, 2007
In this paper, an approach for storing Markov tree, used in various versions of PPM model while p... more In this paper, an approach for storing Markov tree, used in various versions of PPM model while predicting next Web-page is proposed. Markov tree requires huge amount of memory. This problem is solved using Cellular Automata which is considered as a fast and inexpensive mechanism. The proposed technique utilizes non-linear Single Cycle Multiple Attractor Cellular Automata (SMACA) which replaces Markov tree for minimizing the memory requirement.
Proceedings - 2007 International Symposium on Information Technology Convergence, ISITC 2007, 2007
Abstract Search for any topic or a word using Web search engine depends on its ranking mechanism.... more Abstract Search for any topic or a word using Web search engine depends on its ranking mechanism. Generally, search engine sorts through the millions of Web pages and then present the significant Web-pages that match search topic of the user. These matches will be further ranked, so that the most relevant ones come first. This paper proposes an alternate way to rank hyper-linked Web page through analysis of link structure. This link structure is stored in an efficient way to minimize the storage space using Galois Extension Field GF ( ...
Journal of Convergence Information Technology, 2009
Abstract Markov Models have been widely used for predicting next Web-page from the users&... more Abstract Markov Models have been widely used for predicting next Web-page from the users' navigational behavior recorded in the Web-log. This usage-based technique can be combined with the structural properties of the Web-pages to achieve better prediction accuracy. This paper proposes one of the pre-fetching techniques relying both on Markov Model and Ranking which considers the structural properties of the Web. In this paper, prediction accuracy is realized as a linear function of transition probability of first order ...
Journal of Convergence Information Technology, 2009
Clustering of non-homogenous documents has become an increasing challenge and opportunity with th... more Clustering of non-homogenous documents has become an increasing challenge and opportunity with the huge proliferation of World Wide Web. It has become difficult to retrieve the desired information without proper clustering of Web-page with the increase in information on the WWW. Several new ideas have been proposed in recent years. Among them partitioning approach is still widely used clustering approach for its simplicity. This paper proposes a partitioning approach to cluster the Web-page based on information provided by the hyperlink structure of Web-pages and also by the content of the Web-pages. The proposed approach of Web-page clustering exhibits better result than K-medoid partitioning clustering approach as the centroids are chosen by HITS Algorithm. The partitioning approach like Kmediod, K-means require number of clusters apriori. It has been observed that the performance of these approaches depend on the initial selection centroids of the clusters. These two problems have been solved by the approach proposed in this paper. Experimental result supports our approach as better concept.
Web Search Engine uses forward indexing and inverted indexing as a part of its functional design.... more Web Search Engine uses forward indexing and inverted indexing as a part of its functional design. This indexing mechanism helps retrieving data from the database based on user query. In this paper, an efficient solution to handle the indexing problem is proposed with the introduction of Nonlinear Single Cycle Multiple Attractor Cellular Automata (SMACA). This work simultaneously shows generation of
2008 International Conference on Information Technology, 2008
This paper presents an alternative product recommendation system for Business-to-customer ecommer... more This paper presents an alternative product recommendation system for Business-to-customer ecommerce purposes. The system recommends the products to a new user. It depends on the purchase pattern of previous users whose purchase pattern are close to that of new user. The system is based on vector space model to find out the closest user profile among the profiles of all users in database. It also implements Association rule mining based recommendation system, taking into consideration the order of purchase, in recommending more than one product. To make the association rule memoryefficient, cellular automata is used.
9th International Conference on Information Technology (ICIT'06), 2006
Search for a Web-page using Crawler based Search Engine depends on its Ranking Mechanism. Search ... more Search for a Web-page using Crawler based Search Engine depends on its Ranking Mechanism. Search Engine sorts through millions of Web-pages and present the significant Web-pages through Ranking method. A new approach is proposed here adopting Cellular Automata.
Radiation Effects and Defects in Solids - RADIAT EFF DEFECT SOLID, 1983
Thin films of RF sputtered amorphous germanium, silicon and silicon carbide have been irradiated ... more Thin films of RF sputtered amorphous germanium, silicon and silicon carbide have been irradiated in a reactor with fast-neutron fluence of up to 3.6 × 10 n/cm at a temperature not exceeding 50°C. The irradiation induced changes in electrical and structural properties have been studied. The effect of irradiation on dark conductivity is appreciable though the activation energy gap and phonon spectra do not seem to undergo much change. The results could be explained as a combined effect of two processes (i) the rearrangement of the atoms of the amorphous medium in the presence of neutron-fluence to produce a more complete amorphous network, and (ii) the production of defect clusters in certain regions of the amorphous medium, giving rise to isolated zones with high concentration of dangling bonds. Our results, however, indicate that in the amorphous thin films studied, the first process is predominant throughout the volume.
ABSTRACT The d.c. conductivity of RF sputtered hydrogenated amorphous SiC films have been studied... more ABSTRACT The d.c. conductivity of RF sputtered hydrogenated amorphous SiC films have been studied in the temperature range of 350–130 K. High temperature data yield activation energy gap which increases significantly with hydrogen content, while the low temperature conductivity is explained by hopping between localized states.
Uploads
Papers by Ruma Dutta