In large-scale trade shows, there are two kinds of connections dependent on interests of attendees: connections between attendees and booths, and that between booths and booths. The large-scale-ness, however, prevents the establishment of these connections because each attendee cannot obtain the entire information on all of his interesting booths.??In this paper, we propose a personal navigation system to promote establishment of these connections by providing the information of unvisited booths which are interesting to each attendee. The proposed system numerically estimates each user's interest from the history of visited booths and recommends unvisited booths according to his estimated interest through a mobile device and RFID tags installed in booths. By recommending booths with a floor map, their locations, his current position, and quantitative bases of recommendation, each attendee can select and visit one of recommended booths according to these satisfaction levels and locations.