System and method for improved state identification and prediction in computerized queries

A non-transitory computer-readable storage medium having instructions embodied thereon, the instructions being executable by one or more processors to perform a method for improved state identification and prediction in computerized queries, the method comprising:

• generating a model through word embedding;
• searching the model using a query string to identify one or more semantic matches for the query string;
• in response to searching the model, identifying a plurality of results in the model, wherein the plurality of results comprising semantic matches between the query string and one or more of a step or workflow in the model;
• determining probabilities for the plurality of results that represent semantic strength of the semantic matches;
• predicting one or more additional results that are related to the plurality of results;
• causing a ranked display of (a) the plurality of results comprising semantic matches, (b) the probabilities for the plurality of results, and (c) the one or more additional results, in a graphical user interface (GUI);
• electronically aggregating data from a plurality of distinct workflows, the plurality of distinct workflows corresponding to a plurality of entities;
• and
• creating a probability table using the aggregated data, wherein the probability table includes a plurality of probabilities for a plurality of vectors, wherein the plurality of vectors correspond to preceding steps, current steps, and succeeding steps from the plurality of distinct workflows, and wherein each of the plurality of probabilities corresponds to a calculated likelihood of a preceding step and a succeeding step relative to a current step, for each step within each of the plurality of distinct workflows.

A system for improved state identification and prediction in computerized queries, the system comprising:

• one or more hardware processors that execute computer-readable instructions for a plurality of modules, the plurality of modules including:
• a training module that generates a model through word embedding;
• a query module that:
• searches the model using a query string to identify one or more semantic matches for the query string;
• in response to searching the model, identifies a plurality of results in the model, wherein the plurality of results correspond to semantic matches between the query string and one or more of a step or workflow in the model;
• and
• determines probabilities for the plurality of results that represent semantic strength of the semantic matches;
• a prediction module that predicts one or more additional results that are related to the plurality of results;
• a display module that causes (a) a ranked display of the plurality of results comprising the semantic matches, (b) the probabilities determined for the plurality of results, and (c) the one or more additional results, in a graphical user interface (GUI);
• and
• a module that:
• electronically aggregates data from a plurality of distinct workflows, the plurality of distinct workflows corresponding to a plurality of entities;
• and
• creates a probability table using the aggregated data, wherein the probability table includes a plurality of probabilities for a plurality of vectors, wherein the plurality of vectors correspond to preceding steps, current steps, and succeeding steps from the plurality of distinct workflows, and wherein each of the plurality of probabilities corresponds to a calculated likelihood of a preceding step and a succeeding step relative to a current step, for each step within each of the plurality of distinct workflows.