Predictive service request system and methods

A hardware system failure detection system, comprising:

• a processor;
• and
• a memory coupled with and readable by the processor and storing therein a set of instructions which, when executed by the processor, causes the processor to:
• access a plurality of service request outcome models, each of the plurality of service request outcome models usable for generating a predicted outcome corresponding to information that addresses one or more service requests;
• receive a user input comprising natural language input that characterizes one or more abnormal operations performed by a hardware system;
• tokenize the user input into a plurality of tokens;
• access service request data from the hardware system that performed the one or more abnormal operations;
• process the service request data to identify a set of categorical features and a set of non-categorical features associated with the hardware system;
• generate, based on the plurality of tokens from the user input and the set of non-categorical features identified from the service request data, a first bag-of-words binary matrix that includes an entry for each of the plurality of tokens and each non-categorical feature of the set of non-categorical features;
• generate a second categorical binary matrix including an entry for each categorical feature of the set of categorical features;
• combine the first bag-of-words binary matrix and the second categorical binary matrix to create a third binary matrix;
• generate, for each of the plurality of service request outcome models, a predicted outcome that corresponds to the third binary matrix;
• select, based on the predicted outcomes that correspond to the third binary matrix, a service request outcome model among the plurality of service request outcome models;
• identify one or more of a predicted service request category, a predicted service request resolution, or a predicted service request diagnostic based on the selected service request outcome model;
• and
• transmit the one or more of the predicted service request category, the predicted service request resolution, or the predicted service request diagnostic to an automated service request resolution system to resolve the service request.

A non-transitory computer-readable memory device storing code representing instruction to be executed by a processor, the code comprising code to cause the processor to:

• access a plurality of service request outcome models, each of the plurality of service request outcome models usable for generating a predicted outcome corresponding to information that addresses one or more service requests;
• receive a user input comprising natural language input that characterizes one or more abnormal operations performed by a hardware system;
• tokenize the user input into a plurality of tokens;
• access service request data from the hardware system indicated that performed the one or more abnormal operations;
• process the service request data to identify a set of categorical features and a set of non-categorical features associated with the hardware system;
• generate, based on the plurality of tokens from the user input and the set of non-categorical features identified from the service request data, a first bag-of-words binary matrix that includes an entry for each of the plurality of tokens and each non-categorical feature of the set of non-categorical features;
• generate a second categorical binary matrix including an entry for each categorical feature of the set of categorical features;
• combine the first bag-of-words binary matrix and the second categorical binary matrix to create a third binary matrix;
• generate, for each of the plurality of service request outcome models, a predicted outcome that corresponds to the third binary matrix;
• select, based on the predicted outcomes that correspond to the third binary matrix, a service request outcome model among the plurality of service request outcome models;
• identify one or more of a predicted service request category, a predicted service request resolution, or a predicted service request diagnostic based on the selected service request outcome model;
• and
• transmit the one or more of the predicted service request category, the predicted service request resolution, or the predicted service request diagnostic to an automated service request resolution system to resolve the service request.