Best Online Journal of Computer Science and Technology - IRA-JOURNAL

International Research & Analysis Journal facebook twitter youtube

IRAJ VOLUMES

IRAJ Volume 4 Issue 1, 2015-01-16 To 2015-04-15

  • pages (16-23)

    Reads (197)

    Topic Name: High Performance Side Information Mining using Enhanced Quotes Algorithm

    Author (s) : Anshuta Mishra, Asst. Prof. Prateek Gupta


    Quick abstract

    in text documents lot of important information is available as side information. Example of side-information are document provenance information, the links in the document, user-access behavior from web logs, or other non-textual attributes which are embedded into the text document. This information can be used in document clustering. This information may be noisy; hence it is not always easier to use it efficiently. Therefore it is required to use this information in document clustering with proper algorithms to avoid addition of noise in the extracted clusters. Therefore, we need a principled way to perform the mining process, so as to maximize the advantages from using this side information. In this paper, we enhance the design of an existing algorithm which combines classical partitioning algorithms with probabilistic models in order to create an effective clustering approach. We provide experimental results on a various real data sets to illustrate the advantages of using such an approach.


  • pages (9-15)

    Reads (191)

    Topic Name: Reliable Architecture Application Migration over the Cloud

    Author (s) : Yogita Kastor, Prof Deepak Choubey


    Quick abstract

    Cloud computing is spreading round the world and inflicting the researchers to target it. These are 1st, creating it doable to speak between two or additional clouds and second, security of communication. This work proposes to boost the dependability primarily based style optimization by considering the on top of issues. It'll use priorities of the applications supported the historical data for deciding the Ranking of the System in situ of evaluating it solely dynamically. It'll be appropriate for all kinds of applications.


  • pages (1-8)

    Reads (209)

    Topic Name: High Performance & Secured ZigBee Network Using Mixed Key

    Author (s) : Vinod Saxena, Asst. Prof. Sanjay Gupta


    Quick abstract

    THE IEEE 802.15.4/ZIGBEE IS A CUSTOMARY PROTOCOL WHICH PROVIDES LOW RATE, LOW VALUE AND LOWER POWER. IT IS A WIRELESS TECHNOLOGY HAS BEEN THOUGHT OF AS A VERY IMPORTANT TECHNOLOGY FOR WSNS. ORGANIZATIONS USE ZIGBEE NETWORKS TO EFFECTIVELY DELIVER SOLUTIONS FOR A WIDE AREAS ALONG WITH MERCHANT DEVICE MANAGEMENT, ENERGY MANAGEMENT AND LOW LATENCY, HOME AND BUSINESS BUILDING AUTOMATION SIMILARLY AS WORKS MANAGEMENT. ZIGBEE TECHNOLOGY IS BEING EMBEDDED IN AN EXCEEDINGLY BIG SELECTION OF PRODUCTS AND APPLICATIONS ACROSS MEDICAL, COMMERCIAL, CONSUMER, INDUSTRIAL AND GOVERNMENT MARKETS WORLDWIDE. ORGANIZATIONS HAVE A STRAIGHTFORWARD, RELIABLE, AFFORDABLE AND LOW-POWER STANDARDS-BASED WIRELESS TECHNOLOGY ZIGBEE, OPTIMIZED FOR THE DISTINCTIVE DESIRES OF REMOTE OBSERVANCE AND MANAGEMENT APPLICATIONS.


  • pages (24-30)

    Reads (196)

    Topic Name: Web Data Classification Using Partitioning & Side Information

    Author (s) : Nazia Khan, Asst. Prof. Brajesh Patel


    Quick abstract

    with the spread of internet and increasing data over the internet, it is becoming mandatory to research for data filtering and clustering over the Internet as per the requirement of the users of Internet. Search Engine Software and Search Engine Optimization Techniques are helping but they are either proprietary or not up to the mark. Internet is consisting of lot of data and normal SEO techniques mislead the users to irrelevant search results. Actual contents are not searched by using these techniques. Researchers suggest that major relevant contents are enclosed and identified using side information within the contents. Side information includes headings, alternate text, Meta text, bold and strong elements etc. This work proposes to cluster web pages on the basis of the side information as stated above. For clustering web pages various steps has been suggested including data filtering, frequency based clustering etc.


  • pages (31-36)

    Reads (197)

    Topic Name: Reliable Architecture Application Migration over the Cloud

    Author (s) : Swati Yadav, Prof. Randeep Kaur


    Quick abstract

    Recent technological advancements have crystal rectifier to a deluge of knowledge from distinctive domains (e.g., health care and sciatic sensors, user-generated knowledge, web and financial firms, and provide chain systems) over the past twenty years. The term huge knowledge was coined to capture which means of this rising trend. Additionally to its sheer volume, huge knowledge additionally exhibits different distinctive characteristics as compared with ancient knowledge. For example, huge knowledge is usually unstructured and need additional period analysis. This development incorporates new system architectures for knowledge acquisition, transmission, storage, and large-scale processing mechanisms. MapReduce could be a wide used parallel computing framework for big scale processing. The two major performance metrics in MapReduce are job execution time and cluster outturn. This paper can apply the quick knowledge retrieval victimization combination of sorting strategies over MapReduce. The mix sort can have less time quality in respect of the traditional sorting over MapReduce. Keywords-MapReduce; huge Data; Hadoop; Cloud Computing, Combo Sort


top