# US 12,008,580 B2 · Natural language processing machine learning to convert service recommendations to actionable objects > Agents: this data is available as JSON from the patentgrep API and MCP server — see https://patentgrep.com/llms.txt - Number: US 12,008,580 B2 - Document ID: US12008580B2 - Status: Granted - Kind: B2 - Application number: 17/537,003 - Filed: 2021-11-29 - Published: 2023-06-01 - Granted: 2024-06-11 - Earliest claimed: 2021-11-29 - Assignees: Sap SE - Applicants: Sap SE - Inventors: Roman Rapp; Sunil Kumar Panda; Vinay Sheel; Jatin Kochhar - Web page: https://patentgrep.com/patent/US12008580B2 ## CPC - G06Q 30/016: After-sales - G06F 40/205: Parsing - G06F 40/211: Syntactic parsing, e.g. based on context-free grammar [CFG] or unification grammars - G06F 40/253: Grammatical analysis; Style critique - G06F 40/284: Lexical analysis, e.g. tokenisation or collocates - G06F 40/30: Semantic analysis - G06F 40/56: Natural language generation - G06Q: INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR - G06F: ELECTRIC DIGITAL DATA PROCESSING (computer systems based on specific computational models G06N) ## Abstract In an example embodiment, a natural language processing (NLP) machine learning and rules-based text extraction and analysis approach is used to convert a textual service recommendation document into customer-tailored actions considering the specific context-based executable script. This creates end-to-end automation in implementing suggestions provided in textual documents. Actions mentioned in the document can be processed automatically whenever possible, or at least transformed into a semi-automated action with system support. The solution can be configured to automate end-to-end converting of documents into personalized technical scripts and implementing these scripts at the customer-side. ## Claims ### Claim 1 A system comprising: at least one hardware processor; and a non-transitory computer-readable medium storing instructions that, when executed by the at least one hardware processor, cause the at least one hardware processor to perform operations comprising: accessing a service report, the service report having text indicating one or more technical issues detected during running of an online software service and, for each of the one or more technical issues, one or more recommendations on how to address a corresponding technical issue; segmenting the text in the service report into one or more sentences; performing natural language processing on the one or more sentences to identify one or more structured sentence objects of each of the one or more sentences, each structured sentence object indicating a grammatical type of a corresponding word or article of punctuation in a corresponding sentence; passing the one or more structured sentence objects into a machine-learned model trained by a machine learning algorithm to map commands in the one or more sentences to commands available in one or more software tools; generating an actionable objects list including the mapped commands available in the one or more software tools; and sending the actionable objects list to an action executor on a computer system on which the online software service was running, the action executor designed to automatically execute the mapped commands in the actionable objects list to address the one or more technical issues. ### Claim 2 (depends on 1) The system of claim 1, wherein the actionable objects list further includes, for each of the mapped commands, a weightage indicative of criticality of the corresponding mapped command. ### Claim 3 (depends on 1) The system of claim 1, wherein the machine learning algorithm is a neural network. ### Claim 4 (depends on 1) The system of claim 1, wherein the operations further comprise, prior to performing natural language processing, replacing one or more technical terms in the one or more sentences with terms identified as safe for natural language processing, based on a mapping between technical terms and safe terms. ### Claim 5 (depends on 4) The system of claim 4, wherein there is a different mapping for each of a plurality of different industries. ### Claim 6 (depends on 1) The system of claim 1 wherein the operations further comprise identifying one or more commands in the one or more sentences that are dependent upon one another. ### Claim 7 (depends on 1) The system of claim 1, wherein the service report contains a reference to another service report, and the operations further comprise, in response to detection of the reference, retrieving the another service report and performing the segmenting, performing natural language processing, passing, and generating on both the service report and the another service report at the same time. ### Claim 8 A method comprising: accessing a service report, the service report having text indicating one or more technical issues detected during running of an online software service and, for each of the one or more technical issues, one or more recommendations on how to address a corresponding technical issue; segmenting the text in the service report into one or more sentences; performing natural language processing on the one or more sentences to identify one or more structured sentence objects of each of the one or more sentences, each structured sentence object indicating a grammatical type of a corresponding word or article of punctuation in a corresponding sentence; passing the one or more structured sentence objects into a machine-learned model trained by a machine learning algorithm to map commands in the one or more sentences to commands available in one or more software tools; generating an actionable objects list including the mapped commands available in the one or more software tools; and sending the actionable objects list to an action executor on a computer system on which the online software service was running, the action executor designed to automatically execute the mapped commands in the actionable objects list to address the one or more technical issues. ### Claim 9 (depends on 8) The method of claim 8, wherein the actionable objects list further includes, for each of the mapped commands, a weightage indicative of criticality of the corresponding mapped command. ### Claim 10 (depends on 8) The method of claim 8, wherein the machine learning algorithm is a neural network. ### Claim 11 (depends on 8) The method of claim 8, further comprising, prior to performing natural language processing, replacing one or more technical terms in the one or more sentences with terms identified as safe for natural language processing, based on a mapping between technical terms and safe terms. ### Claim 12 (depends on 11) The method of claim 11, wherein there is a different mapping for each of a plurality of different industries. ### Claim 13 (depends on 8) The method of claim 8, further comprising identifying one or more commands in the one or more sentences that are dependent upon on another. ### Claim 14 (depends on 8) The method of claim 8, wherein the service report contains a reference to another service report, and the method further comprises, in response to detection of the reference, retrieving the another service report and performing the segmenting, performing natural language processing, passing, and generating on both the service report and the another service report at the same time. ### Claim 15 A non-transitory machine-readable medium storing instructions which, when executed by one or more processors, cause the one or more processors to perform operations comprising: accessing a service report, the service report having text indicating one or more technical issues detected during running of an online software service and, for each of the one or more technical issues, one or more recommendations on how to address a corresponding technical issue; segmenting the text in the service report into one or more sentences; performing natural language processing on the one or more sentences to identify one or more structured sentence objects of each of the one or more sentences, each structured sentence object indicating a grammatical type of a corresponding word or article of punctuation in a corresponding sentence; passing the one or more structured sentence objects into a machine-learned model trained by a machine learning algorithm to map commands in the one or more sentences to commands available in one or more software tools; generating an actionable objects list including the mapped commands available in the one or more software tools; and sending the actionable objects list to an action executor on a computer system on which the online software service was running, the action executor designed to automatically execute the mapped commands in the actionable objects list to address the one or more technical issues. ### Claim 16 (depends on 15) The non-transitory machine-readable medium of claim 15, wherein the actionable objects list further includes, for each of the mapped commands, a weightage indicative of criticality of the corresponding mapped command. ### Claim 17 (depends on 15) The non-transitory machine-readable medium of claim 15, wherein the machine learning algorithm is a neural network.