Generating programs using context-free compositions and probability of determined transformation rules

摘要

There is provided a method and system for generating a program. The method includes detecting a number of steps for performing a task on a computing device and detecting an example relating to each of the steps, wherein the example includes input data and corresponding output data relating to the step. The method also includes, for each example, determining a rule that transforms the input data to the corresponding output data based on cues including textual features within the input data and the corresponding output data. The method further includes generating a program for performing the task based on the rules.

主权项

1. A method for generating a program, comprising:
detecting a plurality of steps for performing a task on a computing device and actual input data for the task; detecting an example relating to each of the plurality of steps, wherein the example comprises input data and corresponding output data relating to the step; and for each example, determining a plurality of rules that transform the input data to the corresponding output data based on cues comprising textual features within the input data and cues comprising textual features within the corresponding output data; for each example, ranking the plurality of rules according to a probability that each rule explains a transformation from the input data to the corresponding output data; for each example, specifying a rule to be used for performing a step represented by the example based on the ranking of the plurality of rules; and generating a program for performing the task based on the specified rules.