With the advent of information technology, it has attracted much attention from the organisations and individuals on managing information effectively. High quality information search and retrieval has subsequently been a topical issue in IS research as well as one of the major objectives in developing computer-based information systems. The introduction of computer search engine which was designed to address the related issues is a milestone in the IS development history. Nevertheless, its limited solutions only provide quality returned information when the users input the keywords precisely and accurately. Our research aims to tackle with the deficiencies of current search engine which sometimes lead to irrelevant information output. A personal document recommendation system is then designed by semantic ontology technology to catch users' profile and their preferences for information retrieval. In specific, Formal Concept Analysis (FCA) algorithm is applied to be the core ontology construction technology in this study. This study adopts an experimental approach, by which nature we have treated two different FCA based recommendation systems as the experimental group and a traditional keyword based recommendation system as the control group. Users' satisfaction on the quality of returned information and systems usage are then evaluated. The results indicate that users have higher satisfaction on the Scoring-FCA based recommendation systems than the traditional keyword based systems. In short, this study contributes to the knowledge through the design and implement of a personal document recommendation system by semantic ontology and can therefore enhance individual knowledge management.