<p>The management of data-intensive applications that need a lot of storage and powerful computational resources for representation and execution is best handled by scientific workflows, which are dependable solutions. Since a workflow’s execution tasks may involve one or more datasets and necessitate expensive data movement between dispersed cloud data centers, the issue of data placement strategy in scientific cloud workflow applications still needs to be improved. Similarity measures have been widely employed in classification and data optimization applications and could be used to determine the level of dependency between workflow datasets. To minimize data movements, this paper proposes a novel data placement method based on interval type-2 fuzzy similarity measures. Workflow datasets are created as interval type-2 fuzzy sets, and similarity calculations are done between the various datasets. The similar and close datasets are grouped in the same clusters, influencing the amount of transferred data. The results of the experiments indicate that, compared to conventional approaches, the advanced approach reduces the volume of data transfers between data centers.</p>