Design of a Mobile Agent Software for Application in Data Mining.

Date of Submission

December 1998

Date of Award

Winter 12-12-1999

Institute Name (Publisher)

Indian Statistical Institute

Document Type

Master's Dissertation

Degree Name

Master of Technology

Subject Name

Computer Science

Department

Computer and Statistical Services Centre (CSSC)

Supervisor

Bagchi, Aditya (CSSC-Kolkata; ISI)

Abstract (Summary of the Work)

What are Agents?Agents in general are entities who represent a person and carries out certain jobs instead of him. Software Agents are pieces of code, that are Autonomous, that have Social ability, that are Reactive and Proactive.Because of Autonomy, the Agents does not need any intervention from human user and they have some control on their own state. Social ability makes the agents collaborate with other agents or humans and achieve greater tasks which would have been impossible for an agent alone, for example an agent who wants to search databases, but is good at presenting information to human user, can collaborate with another agent which is good at searching databases. Agents sense their environment and respond accordingly, for example arrival of an email can trigger an agent to take particular action. Agents not only react to changes in the environment, they also initiate certain actions, exhibiting goal-driven behaviour, for example As agent which is supposed to fix up an appointment with some person may contact that person by some form of communication.In everyday life, one has used one form of agent or other. For example, mail reader or news reader applications are agents in a way, they are customised with server name and when to fetch mail and in what way to sort the mail/news items. Once it is customised, these applications at proper time fetch the mail/news and sort them in the specified order. Certain applications even tries to find out which mails may be urgent ones by searching for words like 'urgent' or 'emergency' etc. and put those mails on top of the mailbox.Agents fit in the present day network centric world well. For example, if a person wants to search information about a topic of particular interest, the search facilities provided by present generation search engines produce voluminous amount of information, from which required information is to be filtered. If the network bandwidth is very low and the connectivity is also not proper, for example, in case of a home user, then the person has to spend a lot of time before the terminal, waiting for the information to load in his browser and then sift through them and refine the search on and on, till he finds what he needs. If in between the network connection is broken then the search is to be carried out from the start, unless previous search results were stored locally. Agent with the logic for filtering the information needed, embedded in it can download the information when the connectivity is good and off-line can search through them and report the final filtered result, thus freeing its user from this tiresome work.Agent can also use existing Artificial Intelligence techniques to imitate human beings or playing as a good interface between the computer and normal user. Agents can be classified based on whether they are intelligent or not, whether they are mobile or not, the kind of expertise they possess.

Comments

ProQuest Collection ID: http://gateway.proquest.com/openurl?url_ver=Z39.88-2004&rft_val_fmt=info:ofi/fmt:kev:mtx:dissertation&res_dat=xri:pqm&rft_dat=xri:pqdiss:28843710

Control Number

ISI-DISS-1998-56

Creative Commons License

Creative Commons Attribution 4.0 International License
This work is licensed under a Creative Commons Attribution 4.0 International License.

DOI

http://dspace.isical.ac.in:8080/jspui/handle/10263/6229

This document is currently not available here.

Share

COinS