In the field of quality control, the PDCA cycle has been traditionally used in order to improve quality and it has demonstrated its utility. PDCA cycles mean a cycle consisting of Plan, Do, Check, and Act, a method to establish an efficient process to achieve an objective. It means setting an objective and a process to achieve it (Plan), carrying out the process as planned (Do), checking the result (Check), and analyzing problems in the process that caused failure, if we did not achieve the objective and correcting it (Act). Today when more speed is required, the CAPD cycle that begins with close checking of the phenomena has been shown to increase total time efficiency in some cases.Today, in education problems, students are strongly demanding learning efficiency. For example, when they set their own learning goals such as speaking English more fluently and improving their mathematical abilities, they are very interested in discovering learning methods to improve their learning efficiency. From the perspective of learning quality improvement, this is a very interesting problem. Tsubaki, Kakuta and Sun (2005) defined the learning-type PDCA cycle and the CAPD cycle. Then, this paper analyzes text data of 105 actual students' reports on carrying out learning improvement activities with one of the above cycles by self-organizing map. Tsubaki, Kakuta and Sun (2005) analysed the text data of students'reports on carring out learning improvement activities with one of these cycles by correspondence analysis. Correspondence analysis is useful for analysing their text data. However, this paper shows effectiveness of the learning-type PDCA cycleand the CAPD cycle by analysing the same text data(reports of 49 students using the PDCA cycle and 56 students using the CAPD cycle) using self-organizing map. This study examines features of the PDCA and the CAPD cycles and shows the methodology of analysis and its achievement.This paper shows difference of the features of these two cycles. Then, this study clarifies the features of each factor of the learning-type PDCA cycle and CAPD cycle by showing the results of analysis of each factor: Plan, Do, Check, and Act. Moreover, examining the individual difference of students' learning methods, this study examines what kind of reform is necessary to improve learning efficiency depending on student type. As a result of analysis of individual differences, about 40% of students perform well by effectively following the learning-type PDCA cycle and the CAPD cycle while the rest of the students have some room left for improvement somewhere in the steps. This result indicates that by analyzing sentence data of the learning process, each student can clearly show the orientation of improvement.Tsubaki,M., Kakuta,M. and Sun,J.(2005) : "A Study on Learning Typed PDCA and CAPD Cycles for Quality Improvement, " Proceedings of The 9 th World Multi-conference on Systemics, Cybernetics andInformatics, pp.129-136. Kohonen,T. (1995): Self-Organizing Maps, Spromger. Greenacre, M.J. (1984) : Theory and applications of correspondence analysis. Academic Press.We intend to submit our paper to international journal.