# US 12,190,880 B2 · Method and apparatus for multi-party dialogue discourse parsing as a sequence generation > Agents: this data is available as JSON from the patentgrep API and MCP server — see https://patentgrep.com/llms.txt - Number: US 12,190,880 B2 - Document ID: US12190880B2 - Status: Granted - Kind: B2 - Application number: 17/900,134 - Filed: 2022-08-31 - Published: 2024-02-29 - Granted: 2025-01-07 - Earliest claimed: 2022-08-31 - Assignees: Tencent America LLC - Applicants: Tencent America LLC - Inventors: Linfeng Song - Web page: https://patentgrep.com/patent/US12190880B2 ## CPC - G06F 40/35: Discourse or dialogue representation - G10L 15/22: Procedures used during a speech recognition process, e.g. man-machine dialogue - G10L 15/063: {Training} - G10L 15/18: using natural language modelling - G06F 40/205: Parsing - G06F 40/289: Phrasal analysis, e.g. finite state techniques or chunking - G10L 2015/227: {of the speaker; Human-factor methodology} - G06F: ELECTRIC DIGITAL DATA PROCESSING (computer systems based on specific computational models G06N) - G10L: SPEECH ANALYSIS TECHNIQUES OR SPEECH SYNTHESIS; SPEECH RECOGNITION; SPEECH OR VOICE PROCESSING TECHNIQUES; SPEECH OR AUDIO CODING OR DECODING ## Abstract A method performed by at least one processor includes retrieving a dialogue history including a plurality of speech utterances, each speech utterance including one or more words. The method further includes encoding the plurality of speech utterances such that each speech utterance is associated with a sequence identifier indicating an order of each speech utterance in the dialogue history. The method further includes decoding the encoded plurality of speech utterances to generate at least one discourse relation triple corresponding to the dialogue history, the at least one discourse relation triple including a first sequence identifier of a first speech utterance from the plurality of speech utterances, a second sequence identifier of a second speech utterance from the plurality of speech utterances, and a dialogue discourse type. ## Claims ### Claim 1 A method performed by at least one processor, the method comprising: retrieving a dialogue history including a plurality of speech utterances, each speech utterance including one or more words; encoding the plurality of speech utterances such that each speech utterance is associated with a sequence identifier indicating an order of each speech utterance in the dialogue history; and decoding the encoded plurality of speech utterances to generate at least one discourse relation triple corresponding to the dialogue history, the at least one discourse relation triple including a first sequence identifier of a first speech utterance from the plurality of speech utterances, a second sequence identifier of a second speech utterance from the plurality of speech utterances, and a dialogue discourse type. ### Claim 2 (depends on 1) The method of claim 1, wherein the second speech utterance occurs after the first speech utterance such that the second speech utterance is a response to the first speech utterance. ### Claim 3 (depends on 2) The method of claim 2, wherein the decoding the encoded plurality of speech utterances is performed such that the second speech utterance is the last speech utterance in the dialogue history. ### Claim 4 (depends on 1) The method of claim 1, wherein the decoding the encoded plurality of speech utterances is performed such that a discourse relation triple is generated for each speech utterance included in the plurality of speech utterances. ### Claim 5 (depends on 1) The method of claim 1, further comprising: calculating a loss operation that compares a current dialogue context with a target token sequence; and training the encoder and the decoder based on the calculated loss operation. ### Claim 6 (depends on 1) The method of claim 1, wherein the discourse type is selected from one of a question answer pair, a question elaboration, and an acknowledgement. ### Claim 7 (depends on 1) The method of claim 1, wherein the encoder is a T5 encoder and the decoder is a T5 decoder. ### Claim 8 (depends on 1) The method of claim 1, wherein the dialogue history includes at least three or more speakers. ### Claim 9 An apparatus comprising: at least one memory configured to store computer program code; and at least one processor configured to access said at least one memory and operate as instructed by said computer program code, said computer program code including: retrieving code configured to cause the at least one processor to retrieve a dialogue history including a plurality of speech utterances, each speech utterance including one or more words, encoding code configured to cause the at least one processor to encode the plurality of speech utterances such that each speech utterance is associated with a sequence identifier indicating an order of each speech utterance in the dialogue history, and decoding code configured to cause the at least one processor to decode the encoded plurality of speech utterances to generate at least one discourse relation triple corresponding to the dialogue history, the at least one discourse relation triple including a first sequence identifier of a first speech utterance from the plurality of speech utterances, a second sequence identifier of a second speech utterance from the plurality of speech utterances, and a dialogue discourse type. ### Claim 10 (depends on 9) The apparatus of claim 9, wherein the second speech utterance occurs after the first speech utterance such that the second speech utterance is a response to the first speech utterance. ### Claim 11 (depends on 10) The apparatus of claim 10, wherein the decoding code is further configured to cause the at least one processor to decode the encoded plurality of speech utterances such that the second speech utterance is the last speech utterance in the dialogue history. ### Claim 12 (depends on 9) The apparatus of claim 9, wherein the decoding code is further configured to cause the at least one processor to decode the encoded plurality of speech utterances such that a discourse relation triple is generated for each speech utterance included in the plurality of speech utterances. ### Claim 13 (depends on 9) The apparatus of claim 9, wherein said computer program code further includes: calculating code configured to cause the at least one processor to calculate a loss operation that compares a current dialogue context with a target token sequence; and training code configured to cause the at least one processor to train the encoder and the decoder based on the calculated loss operation. ### Claim 14 (depends on 9) The apparatus of claim 9, wherein the discourse type is selected from one of a question answer pair, a question elaboration, and an acknowledgement. ### Claim 15 (depends on 9) The apparatus of claim 9, wherein the encoder is a T5 encoder and the decoder is a T5 decoder. ### Claim 16 (depends on 9) The apparatus of claim 9, wherein the dialogue history includes at least three or more speakers. ### Claim 17 A non-transitory computer readable medium having instructions stored therein, which when executed a processor cause the processor to execute a method comprising: retrieving a dialogue history including a plurality of speech utterances, each speech utterance including one or more words; encoding the plurality of speech utterances such that each speech utterance is associated with a sequence identifier indicating an order of each speech utterance in the dialogue history; and decoding the encoded plurality of speech utterances to generate at least one discourse relation triple corresponding to the dialogue history, the at least one discourse relation triple including a first sequence identifier of a first speech utterance from the plurality of speech utterances, a second sequence identifier of a second speech utterance from the plurality of speech utterances, and a dialogue discourse type. ### Claim 18 (depends on 17) The non-transitory computer readable medium of claim 17, wherein the second speech utterance occurs after the first speech utterance such that the second speech utterance is a response to the first speech utterance. ### Claim 19 (depends on 18) The non-transitory computer readable medium of claim 18, wherein the decoding the encoded plurality of speech utterances is performed such that the second speech utterance is the last speech utterance in the dialogue history. ### Claim 20 (depends on 17) The non-transitory computer readable medium of claim 17, wherein the decoding the encoded plurality of speech utterances is performed such that a discourse relation triple is generated for each speech utterance included in the plurality of speech utterances.