The developed system is job recommendationsystem for campus recruitment which helps college placement office to matchcompany’s profiles and student’s profiles with higher precision and lower cost. For profile matching, two matchingmethods are used: semantic matching, tree-based knowledge matching and querymatching. These methods are integrated according to representations ofattributes of students and companies, and then the profile similarity degree isacquired. Based on profile similarity degree, preference lists of companies andstudents are generated. Also students can perform keyword based search for jobprofiles from various job recruitment sites (e.g. Naukari.com,indeed.com). Forobtaining data from online recruitment sites system uses web crawling. Withloop matching, matching results would be further optimized and provide moreeffective guidance for recommendation.