Recently, data mining of uncertain data especially frequent pattern mining has become an interesting research topic. On the other hand, in many applications, interactive mining is a need where user changes minimum support threshold to find proper frequent patterns. Thus, in this paper, an efficient tree is proposed called UDFP-tree based on our previous model [9] for interactive mining from uncertain data. By using the proposed tree, the mining model construction is separated from the mining process, and it is possible to build mining model once and use it many times. To evaluate the proposed tree, it was compared with UFP-tree algorithm which is the best algorithms for static frequent pattern mining from uncertain data. The experiments results show that although the runtime of UFP-tree for static mining was less than UDFP- tree, in interactive mining, the UFP-tree must be reconstructed and the number of changes in minimum support threshold increases its cost. Therefore, after UDFP-tree is constructed it can be frequently mined by different minimum support threshold.