Materials informatics (MI), which accelerates the process of materials development and optimization by utilizing vast amounts of accumulated experimental data, has attracted much attention. In MI, it is necessary to efficiently search and extract desired experimental data from all experimental data obtained by a number of trial-and-error experiments. For this purpose, we developed a system to manage experimental flows in chemical materials development, which can efficiently search for the experimental flows that are similar with a specific experimental flow. In the proposed system, all existing experimental flows are integrated into a virtual experimental flow with operation grouping by taking into account possible orders of task sequence in experimental flows. Two types of meta-data are introduced to reduce the cost in searching for similar experimental flows. We investigated the proposed search method and clarified the scenarios in which the meta-data works effectively.