Machine learning for automatic smart extraction and workflow assignment of action items

A computer-implemented method for automatic extraction and assignment of action items, the method comprising:

• extracting, by at least a processor, a set of candidate action items from text of a construction project manual;
• applying, by at least the processor, one or more static rules to each of the candidate action items to distinguish valid and invalid candidate action items in the set;
• evaluating, by at least the processor, each valid candidate action item with a first machine learning model to label the valid candidate action item as one of either (i) a true action item or (ii) not a true action item, wherein the first machine learning model performs one class classification of the valid candidate action items to generate the label;
• evaluating, by at least the processor, each true action item with a second machine learning model to allocate each of the true action items to one of a set of construction workflow classes wherein the second machine learning model performs multiple class classification of the true action items to make the allocation;
• and
• transmitting, by at least the processor, the set of true action items to a submittal exchange system to populate one or more workflows.

A computing system for automatic extraction and assignment of action items, the system comprising:

• a processor;
• a memory operably connected to the processor;
• a non-transitory computer-readable medium operably connected to the processor and memory and storing computer-executable instructions that when executed by at least the processor cause the computing system to:
• extract a set of candidate action items from text of a construction project manual;
• apply one or more static rules to each of the candidate action items to distinguish valid and invalid candidate action items in the set;
• evaluate each valid candidate action item with a first machine learning model to label the valid candidate action item as one of either (i) a true action item or (ii) not a true action item, wherein the first machine learning model performs one class classification of the valid candidate action items to generate the label;
• evaluate each true action item with a second machine learning model to allocate each of the true action items to one of a set of construction workflow classes wherein the second machine learning model performs multiple class classification of the true action items to make the allocation;
• and
• transmit the set of true action items to a submittal exchange system to populate one or more workflows.