# US 12,001,785 B2 · Mining multi-party collaboration platforms to create triaging trees and playbooks > Agents: this data is available as JSON from the patentgrep API and MCP server — see https://patentgrep.com/llms.txt - Number: US 12,001,785 B2 - Document ID: US12001785B2 - Status: Granted - Kind: B2 - Application number: 17/017,202 - Filed: 2020-09-10 - Published: 2022-03-10 - Granted: 2024-06-04 - Earliest claimed: 2020-09-10 - Assignees: International Business Machines Corporation - Applicants: International Business Machines Corporation - Inventors: Amitkumar Manoharrao Paradkar; Amar P. Azad; Ajay Gupta; Suranjana Samanta; Prateeti Mohapatra; Harshit Kumar; Eyal Shnarch - Web page: https://patentgrep.com/patent/US12001785B2 ## CPC - G06F 40/20: Natural language analysis (semantic analysis of natural language G06F40/30) - G06F 40/35: Discourse or dialogue representation - G06F 16/2228: {Indexing structures} - G06F 16/285: {Clustering or classification} - G06F 16/3329: Natural language query formulation - G06F 16/345: {Summarisation for human users} - G06F 16/9027: {Trees} - G06N 3/04: Architecture, e.g. interconnection topology - G06N 3/044: Recurrent networks, e.g. Hopfield networks - G06N 3/0442: characterised by memory or gating, e.g. long short-term memory [LSTM] or gated recurrent units [GRU] - G06N 3/09: Supervised learning - G06N 7/01: Probabilistic graphical models, e.g. probabilistic networks - G06F: ELECTRIC DIGITAL DATA PROCESSING (computer systems based on specific computational models G06N) - G06N: COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS ## Abstract Methods, systems, and computer program products for mining multi-party collaboration platforms to create triaging trees and playbooks are provided herein. A computer-implemented method includes obtaining, from a multi-user platform, conversations related to at least one technical issue; generating a plurality of triaging trees by analyzing the conversations, wherein each of the triaging trees stores information corresponding to temporal sequences of steps related to diagnosing and resolving said at least one technical issue; and deriving a playbook for resolving said at least one technical issue at least in part by combining two or more of the plurality of triaging trees. ## Claims ### Claim 1 A computer-implemented method, the method comprising: obtaining, from a multi-user platform, conversations related to one or more technical issues; analyzing the conversations, wherein, for a given one of the conversations, said analyzing comprises assigning respective ones of a plurality of labels to corresponding parts of the given conversation and identifying one or more entities within each labeled part of the given conversation; generating a plurality of tree data structures based on the analyzing, wherein each of the tree data structures comprises a set of nodes for storing information corresponding to temporal sequences of steps related to diagnosing and resolving at least one of the one or more technical issues, wherein the set of nodes comprises a root node corresponding to the at least one of the technical issues, one or more symptom nodes, and one or more action nodes associated with one or more of the symptom nodes, and wherein the tree data structure corresponding to the given one of the conversations is populated based at least in part on the assigned labels and dependency information determined for the assigned labels by an attention model; combining at least a first tree data structure and a second tree data structure of the plurality of tree data structures in response to determining that the first and the second tree data structures comprise respective root nodes corresponding to a same one of the one or more technical issues, wherein the combining comprises merging respective symptom nodes of the first and second tree data structures that correspond to a same at least one symptom and merging respective actions nodes of the first and second tree data structures that correspond to a same at least one action; and deriving a sequence of steps for resolving said at least one technical issue based at least in part on the result of the combining; wherein the method is carried out by at least one computing device. ### Claim 2 (depends on 1) The method of claim 1, comprising: generating a multi-granular summary for at least one of the tree data structures, wherein the multi-granular summary comprises text summarizing the at least one tree data structure in multiple levels of detail; and generating an output for at least one user, wherein the output is controlled based at least in part on at least one of said multiple levels of detail of said multi-granular summary and one or more inputs of the least one user. ### Claim 3 (depends on 2) The method of claim 2, wherein generating the multi-granular summary for at least one of the tree data structures comprises at least one of: populating a template using the information stored in the at least one tree data structure; and applying a trained neural network that maps the at least one tree data structure to the text. ### Claim 4 (depends on 1) The method of claim 1, wherein the plurality of labels comprises labels for at least the following items: the at least one technical issue; one or more symptoms of the least one technical issue; one or more attempted actions to resolve the at least one technical issue; and one or more outcomes of the attempted actions. ### Claim 5 (depends on 4) The method of claim 4, wherein the attention model comprises a bi-directional long short-term memory with a conditional random field layer. ### Claim 6 (depends on 1) The method of claim 1, comprising: outputting at least a portion of the generated tree data structures for verification by a user; and retraining the attention model based on feedback received from said user in response to said outputting. ### Claim 7 (depends on 1) The method of claim 1, wherein each of the conversations is between two or more users of a multi-user platform, and wherein the method comprises: identifying one or more further conversations on said multi-user platform; and updating the derived sequence of steps based at least in part on said further conversations. ### Claim 8 (depends on 1) The computer-implemented method of claim 1, wherein software is provided as a service in a cloud environment. ### Claim 9 A computer program product comprising a computer readable storage medium having program instructions embodied therewith, the program instructions executable by a computing device to cause the computing device to: obtain, from a multi-user platform, conversations related to one or more technical issues; analyze the conversations, wherein, for a given one of the conversations, said analyzing comprises assigning respective ones of a plurality of labels to corresponding parts of the given conversation and identifying one or more entities within each labeled part of the given conversation; generate a plurality of tree data structures based on the analyzing, wherein each of the tree data structures comprises a set of nodes for storing information corresponding to temporal sequences of steps related to diagnosing and resolving at least one of the one or more technical issues, wherein the set of nodes comprises a root node corresponding to the at least one of the technical issues, one or more symptom nodes, and one or more action nodes associated with one or more of the symptom nodes, and wherein the tree data structure corresponding to the given one of the conversations is populated based at least in part on the assigned labels and dependency information determined for the assigned labels by an attention model; combining at least a first tree data structure and a second tree data structure of the plurality of tree data structures in response to determining that the first and the second tree data structures comprise respective root nodes corresponding to a same one of the one or more technical issues, wherein the combining comprises merging respective symptom nodes of the first and second tree data structures that correspond to a same at least one symptom and merging respective actions nodes of the first and second tree data structures that correspond to a same at least one action; and derive a sequence of steps for resolving said at least one technical issue based at least in part on the result of the combining. ### Claim 10 (depends on 9) The computer program product of claim 9, wherein the program instructions executable by a computing device further cause the computing device to: generate a multi-granular summary for at least one of the tree data structures, wherein the multi-granular summary comprises text summarizing the at least one tree data structure in multiple levels of detail; and generate an output for at least one user, wherein the output is controlled based at least in part on at least one of said multiple levels of detail of said multi-granular summary and one or more inputs of the least one user. ### Claim 11 (depends on 9) The computer program product of claim 9, wherein the plurality of labels comprises labels for at least the following items: the at least one technical issue; one or more symptoms of the least one technical issue; one or more attempted actions to resolve the at least one technical issue; and one or more outcomes of the attempted actions. ### Claim 12 (depends on 11) The computer program product of claim 11, wherein the attention model comprises a bi-directional long short-term memory with a conditional random field layer. ### Claim 13 (depends on 9) The computer program product of claim 9, wherein the program instructions executable by a computing device further cause the computing device to: output at least a portion of the generated tree data structures for verification by a user; and retrain the attention model based on feedback received from said user in response to said outputting. ### Claim 14 (depends on 9) The computer program product of claim 9, wherein each of the conversations is between two or more users of a multi-user platform, and wherein the program instructions further cause the computing device to: identify one or more further conversations on said multi-user platform; and update the derived sequence of steps based at least in part on said further conversations. ### Claim 15 A system comprising: a memory; and at least one processor operably coupled to the memory and configured for: obtaining, from a multi-user platform, conversations related to one or more technical issues; analyzing the conversations, wherein, for a given one of the conversations, said analyzing comprises assigning respective ones of a plurality of labels to corresponding parts of the given conversation and identifying one or more entities within each labeled part of the given conversation; generating a plurality of tree data structures based on the analyzing, wherein each of the tree data structures comprises a set of nodes for storing information corresponding to temporal sequences of steps related to diagnosing and resolving at least one of the one or more technical issues, wherein the set of nodes comprises a root node corresponding to the at least one of the technical issues, one or more symptom nodes, and one or more action nodes associated with one or more of the symptom nodes, and wherein the tree data structure corresponding to the given one of the conversations is populated based at least in part on the assigned labels and dependency information determined for the assigned labels by an attention model; combining at least a first tree data structure and a second tree data structure of the plurality of tree data structures in response to determining that the first and the second tree data structures comprise respective root nodes corresponding to a same one of the one or more technical issues, wherein the combining comprises merging respective symptom nodes of the first and second tree data structures that correspond to a same at least one symptom and merging respective actions nodes of the first and second tree data structures that correspond to a same at least one action; and deriving a sequence of steps for resolving said at least one technical issue based at least in part on the result of the combining. ### Claim 16 (depends on 15) The system of claim 15, wherein the at least one processor is configured for: generating a multi-granular summary for at least one of the tree data structures, wherein the multi-granular summary comprises text summarizing the at least one tree data structure in multiple levels of detail; and generating an output for at least one user, wherein the output is controlled based at least in part on at least one of said multiple levels of detail of said multi-granular summary and one or more inputs of the least one user. ### Claim 17 (depends on 16) The system of claim 16, wherein generating the multi-granular summary for at least one of the tree data structures comprises at least one of: populating a template using the information stored in the at least one tree data structure; and applying a trained neural network that maps the at least one tree data structure to the text. ### Claim 18 (depends on 15) The system of claim 15, wherein the plurality of labels comprises labels for at least the following items: the at least one technical issue; one or more symptoms of the least one technical issue; one or more attempted actions to resolve the at least one technical issue; and one or more outcomes of the attempted actions. ### Claim 19 (depends on 18) The system of claim 18, wherein the attention model comprises a bi-directional long short-term memory with a conditional random field layer. ### Claim 20 (depends on 15) The system of claim 15, wherein the at least one processor is configured for: outputting at least a portion of the generated tree data structures for verification by a user; and retraining the attention model based on feedback received from said user in response to said outputting.