International Journal of Computer Applications (0975 – 8887) Volume 64– No.19, February 2013 42 Ontological Description and MAS based Composition Process of SWS's for Recruitment Domain Alankar Srivastava Computer Science & Engg. Department SITM, Barabanki, India Shruti Saxena Computer Science & Engg. Department SRMCEM,Luknow,India Chandrabhan Singh Computer Science & Engg. Department SITM, Barabanki, India ABSTRACT It seems that information is spread over internet and to retrieve useful information easily by mean of intelligent computing is the growing field of research. This need gives birth to semantic web services and various approaches are given for composition of these services. The key problem is to fulfill user’s request. Intelligent Agents are used as a powerful approach for semantic web services selection and composition. Agent can be viewed as a self contained autonomous software component or say codes which is used for fulfillment of users request in any domain. Here in this paper semantic web services and agent capability is used for Man-Power Recruitment System (MPRS). Ontology of domain is used as a complete domain knowledge. Tools like Protégé is used to create agent profile which shows the ontological description of recruitment domain. This paper focuses on how multi-agent system provides a new variety for recruitment system. This paper also spotlights to show an interactive recruiter for Man-Power Recruitment System (MPRS). Here service selection, task decomposition and service composition process of semantic web is exposed on the basis cognitive parameters, predilection and situation for man-power recruitment system. General Terms Semantic web services, SWS Composition, Agent, Ontology and Recruiter. Keywords Man-Power Recruitment System, Recruiter, Man-Power Recruitment (MPR) Domain Ontology, Agent, Semantic Web. 1. INTRODUCTION Semantic web which becomes the extension of current web is not only web of linked pages but it show intelligence also and viewed as a huge collection of data. The emergent use of XML over network prominence on the exploit of XML based indexing as querying language to build web semantic. In recruitment system this semantic attribute can also be used for not only retrieving information over internet but also it can be utilized for training and generating a feasible process of recruitment better for both company and individual. In recruitment process of any company the problem that comes in picture are like where the cost of selection will be minimum i.e. number of eligible candidate will be high, secondly from where we will get candidates which will be suitable in the working environment of the company and various other parameters from the point of view of company. Here in man- power recruitment system the aim is to provide complete compiled information to company or say selectors about college, zone, students, their back log, average percentage, collage reputation and other factors about collage at a single point. It is well-known that semantic web provides a podium for data amalgamation, distribution and reusing from various sources or services. Here the complete domain knowledge is represented in form of ontology’s. Control of semantic web [1] with potentiality and capacity of agent machinery [2, 3] is used to assist man-power recruitment domain. Distributed problem solving approach [4] of multi-agent system is used for semantic web services and then the compositions of these services are done to fulfill the task. A variety of researches are made on SWS and its composition [5, 6]. Data in current web is stored in variety of languages over internet but semantic web provide a unique XML based web description language known as RDF [7]. It has been seen that RDF and OWL [8] are mostly used for depiction of resources and how they are connected on web. This paper focuses on joint approach of multi-agent and SWS’s to assist recruitment system. The combined circumstances of multi-agent and SWS’s can be capitalized in various domains like medical, education planning, intelligent tutoring etc. This paper paying attention on recruitment system. 2. COGNITION BASED AGENT SELECTION Selection of agents is an important task which involves selection of both coordinator as well as service facilitator agents. This selection process is done on the basis of cognitive parameters. These parameters are user’s preferences, quality factors and pre-conditions. In man-power recruitment domain [7] these parameters are average student percentage in the college, previous placement ratio, average backlog per student, state ranking of college, level of aggradations of college, faculty grading etc. Firstly the input request of company is fetched then verification and validation for that input request is done. Then on the basis of various cognitive parameters the QoS of agent is evaluated. Agent having best QoS value is selected as coordinate agent. Selection of facilitator agents is based on decomposition process. When large or say complex input request which is not possible to fulfilled by using single agent then that request is decomposed in various less dependent sub-tasks and for each sub-task a service provider or say facilitator agent is selected. Selection of semantic value can be formulize as QoS of agent. where Pf= placement factor of college Qh= quality of service for hospital C= charge of service D= zonal distance