A Machine Learning Approach for Automated Filling of Categorical Fields in Data Entry Forms

Users frequently interact with software systems through data entry forms. However, form filling is time-consuming and error-prone. Although several techniques have been proposed to auto-complete or pre-fill fields in the forms, they provide limited support help users fill categorical fields, i.e., that require choose right value among a large set of options. In this paper, we propose LAFF, learning-based automated approach for LAFF first builds Bayesian Network models by learning field dependencies from historical input instances, representing values filled past. To improve its ability, uses local modeling effectively mine cluster instances. During phase, such predict possible target field, based on already-filled their dependencies; predicted (endorsed prediction confidence) are then provided end-user as list suggestions. We evaluated assessing effectiveness efficiency two datasets, one them proprietary banking domain. Experimental results show able accurate suggestions Mean Reciprocal Rank above 0.73. Furthermore, efficient, requiring at most 317 ms per suggestion.