# US 12,604,061 B2 · Closed captioning summarization > Agents: this data is available as JSON from the patentgrep API and MCP server — see https://patentgrep.com/llms.txt - Number: US 12,604,061 B2 - Document ID: US12604061B2 - Status: Granted - Kind: B2 - Application number: 18/971,864 - Filed: 2024-12-06 - Published: 2025-06-12 - Granted: 2026-04-14 - Earliest claimed: 2023-12-12 - Assignees: CenturyLink Intellectual Property LLC - Applicants: CenturyLink Intellectual Property LLC - Inventors: John R.B. Woodworth; Dean Ballew - Web page: https://patentgrep.com/patent/US12604061B2 ## CPC - H04N 21/4722: for requesting additional data associated with the content - H04N 21/251: {Learning process for intelligent management, e.g. learning user preferences for recommending movies (details of learning user preferences for the retrieval of video data in a video database G06F16/739; computer systems using learning methods G06N3/08)} - H04N 21/4316: {for displaying supplemental content in a region of the screen, e.g. an advertisement in a separate window} - H04N 21/4884: {for displaying subtitles} - H04N 21/8456: {by decomposing the content in the time domain, e.g. in time segments} - H04N 21/8549: Creating video summaries, e.g. movie trailer {(retrieval in video databases by using presentations in form of a video summary G06F16/739)} - H04N: PICTORIAL COMMUNICATION, e.g. TELEVISION ## Abstract Systems and methods for providing closed captioning summarization. Various options may be provided to summarize different portions of a media item (e.g., a last 5 minutes, the last 15 minutes, from the beginning of the show). Upon receiving a request for a summary, a closed captioning summarization service may extract a portion of closed captions of a corresponding closed caption file and generate a prompt for a language model (LM) instructing the language model to summarize the extracted portion. A summary received from the LM may be presented on a display concurrently with the media item to provide a recapitulation of a desired portion of the media item to the user without disrupting play of the media item. ## Claims ### Claim 1 A method for providing closed captioning summarization, comprising: receiving a selection to summarize a preceding part of a media item; determining at least one timestamp corresponding to the selected preceding part; obtaining closed captions corresponding to the at least one timestamp; generating a language-model (LM) prompt including the closed captions and instructions to summarize the closed captions; obtaining information about characters of the media item; including the information about the characters and instructions to generate descriptions of the characters in the LM prompt; providing the LM prompt to an LM; receiving, from the LM, a summary of the closed captions and a character description for each of the characters of the media item; separating the summary into a plurality of portions, where each portion corresponds with a character description for one of the characters of the media item; determining timestamps and position information corresponding to appearances of the characters of the media item in upcoming scenes of the media item; causing a display of the summary on a display screen; and causing a display of each character description portion at a position near the corresponding character based on the timestamps and position information of the appearances of the characters in upcoming scenes. ### Claim 2 (depends on 1) The method of claim 1, wherein receiving the selection to summarize the preceding part of the media item comprises receiving a selection to summarize the media item from a starting point, the starting point including one of: a start of the media item; a time between the start and a current time; or a prior scene of the media item. ### Claim 3 (depends on 1) The method of claim 1, wherein: the media item is an in-progress episode included in a plurality of episodes of a parent media item; and receiving the selection to summarize the preceding part of the media item comprises receiving a selection to summarize at least one preceding episode of the in-progress episode. ### Claim 4 (depends on 3) The method of claim 3, further comprising: obtaining closed captions of the at least one preceding episode; and including the obtained closed captions of the at least one preceding episode and instructions to summarize the closed captions of the at least one preceding episode in the LM prompt. ### Claim 5 (depends on 1) The method of claim 1, further comprising: obtaining additional information about the media item; and including, in the LM prompt, the additional information as context for the summary. ### Claim 6 (depends on 1) The method of claim 1, further comprising: causing a display of one or more of the character descriptions of the characters on the display screen. ### Claim 7 (depends on 1) The method of claim 1, wherein causing the display of the summary on the display screen comprises displaying the summary on a same display screen on which the media item is displayed. ### Claim 8 (depends on 1) The method of claim 1, wherein causing the display of the summary on the display screen comprises displaying the summary on a display screen different from a display screen on which the media item is displayed. ### Claim 9 A system for providing closed captioning summarization, comprising: a processing system; and memory storing instructions that, when executed by the processing system, cause the system to: receive a selection to summarize a preceding part of a media item; determine at least one timestamp corresponding to the selected preceding part; obtain a portion of closed captions corresponding to the at least one timestamp; generate summarization instructions to summarize the closed captions portion; include, in a language-model (LM) prompt, the summarization instructions and the closed captions portion; provide the LM prompt to an LM; receive, from the LM, a summary of the closed captions portion; separate the summary into a plurality of portions, where each portion corresponds with a character description; determine timestamps and position information corresponding to appearances of the characters in upcoming scenes of the media item; and cause a display of each character description portion at a position near the corresponding character based on the timestamps and position information of the appearances of the characters in upcoming scenes; and cause a display of the summary on a display screen. ### Claim 10 (depends on 9) The system of claim 9, wherein the closed captions include at least one of: dialog; stage directions; speaker identification; sound descriptions; and position information. ### Claim 11 (depends on 9) The system of claim 9, wherein the selection to summarize the preceding part of the media item comprises a selection to: summarize the media item from a start of the media item; summarize the media item from a time between the start of the media item and a current time; summarize a prior scene of the media item; or summarize a preceding episode of a plurality of episodes of the media item. ### Claim 12 (depends on 9) The system of claim 9, wherein the instructions further cause the system to: obtain additional information about the media item; and include, in the LM prompt, the additional information as context for the summary. ### Claim 13 (depends on 9) The system of claim 9, wherein the instructions further cause the system to: obtain information about characters of the media item; include the information about the characters and instructions to generate descriptions of the characters in the LM prompt; and cause a display of one or more of the descriptions of the characters on the display screen. ### Claim 14 (depends on 9) The system of claim 9, wherein the media item is displayed on a first display screen and the summary is displayed on a second display screen. ### Claim 15 (depends on 9) The system of claim 9, wherein the media item is displayed on the display screen and the summary is displayed in a designated area of the display screen. ### Claim 16 A non-transitory computer readable medium comprising instructions, which when executed by a computer, cause the computer to: receive a selection to summarize a preceding part of a media item; determine at least one timestamp corresponding to the selected preceding part; obtain a portion of closed captions corresponding to the at least one timestamp; generate summarization instructions to summarize the closed captions portion; include, in a language-model (LM) prompt, the summarization instructions and the closed captions portion; obtain information about characters of the media item; include the information about the characters and instructions to generate descriptions of the characters in the LM prompt; provide the LM prompt to an LM; receive, from the LM, a summary of the closed captions portion and descriptions of the characters; separate the summary into a plurality of portions, where each portion corresponds with a character description; determine timestamps and position information corresponding to appearances of the characters in upcoming scenes of the media item; cause a display of each character description portion at a position near the corresponding character based on the timestamps and position information of the appearances of the characters in upcoming scenes; and cause a display of the summary on a display screen. ### Claim 17 (depends on 16) The non-transitory computer readable medium of claim 16, wherein the instructions further cause the computer to: obtain additional information about the media item; and include, in the LM prompt, the additional information as context for the summary.