Call summary processing method and apparatus, electronic device, and storage medium

Abstract

The present application discloses a call summary processing method and apparatus, an electronic device, and a storage medium. The method comprises: displaying a call summary corresponding to target call data on a specified interface, wherein the call summary comprises a specified keyword, and the specified keyword represents key detail information in the target call data; and in response to a touch operation on the specified keyword, navigating to a corresponding position of the specified keyword in the target call data. The method improves the user experience.

Description

Call summary processing method and device, electronic equipment and storage medium

[0001] Cross-reference to Related Applications

[0002] This application claims priority to Chinese application No. 202410427693.6, filed on April 9, 2024, which is hereby incorporated by reference in its entirety for all purposes. TECHNICAL FIELD

[0003] The present application relates to the technical field of natural language processing, and more particularly, to a call summary processing method and device, an electronic equipment and a storage medium. BACKGROUND

[0004] With the rapid development of mobile communication technology, more and more users communicate through voice calls in the form of work and life, greatly improving the convenience of communication. For example, if the user is communicating important conference, some personnel who are unable to participate due to special reasons will generally hope to record the conference communication content. However, even if the complete call content can be recorded, the user needs to listen to the playback slowly or convert the recorded content into text to view in sequence to know the important content of the conference, which makes the user unable to quickly view the key information of the communication content, reducing the user experience. SUMMARY

[0005] The present application provides a call summary processing method and device, an electronic equipment and a storage medium to improve the above problems.

[0006] In a first aspect, the present application provides a call summary processing method, which includes: displaying a call summary corresponding to target call data on a specified interface, the call summary including a specified keyword, the specified keyword representing key detail information in the target call data; in response to a touch operation acting on the specified keyword, jumping to a corresponding position of the specified keyword in the target call data.

[0007] In a second aspect, the present application provides a call summary processing device, which includes: a call summary display module configured to display a call summary corresponding to target call data on a specified interface, the call summary including a specified keyword, the specified keyword representing key detail information in the target call data; a call summary processing module configured to, in response to a touch operation acting on the specified keyword, jump to a corresponding position of the specified keyword in the target call data.

[0008] In a third aspect, the present application provides an electronic device, comprising one or more processors and a memory; one or more programs stored in the memory and configured to be executed by the one or more processors, the one or more programs configured to perform the method of the first aspect.

[0009] In a fourth aspect, the present application provides a computer-readable storage medium, having program codes stored therein, wherein the program codes perform the method of the first aspect when executed. BRIEF DESCRIPTION OF DRAWINGS

[0010] In order to more clearly illustrate the technical solutions in the embodiments of the present application, the drawings needed to be used in the embodiment description will be briefly introduced. Obviously, the drawings in the following description are only some embodiments of the present application, and other drawings can be obtained by those skilled in the art without any creative effort on the basis of these drawings.

[0011] FIG. 1 shows a flowchart of a call summary processing method according to an embodiment of the present application.

[0012] FIG. 2 shows an example diagram of a generated call summary according to an embodiment of the present application.

[0013] FIG. 3 shows an effect diagram after adding content identifier and position identifier to a specified keyword in the to-be-processed call data according to an embodiment of the present application.

[0014] FIG. 4 shows a flowchart of a call summary processing method according to another embodiment of the present application.

[0015] FIG. 5 shows an example diagram of a display interface in which a specified keyword is popped up with additional function options corresponding to the specified keyword according to an embodiment of the present application.

[0016] FIG. 6 shows an example diagram of displaying a specified keyword in a call summary in a specified format according to an embodiment of the present application.

[0017] FIG. 7 shows a flowchart of a call summary processing method according to yet another embodiment of the present application.

[0018] FIG. 8 shows a flowchart of the method of step S310 in FIG. 7.

[0019] FIG. 9 shows an example diagram of a display of a call summary generation control according to an embodiment of the present application.

[0020] FIG. 10 shows a flowchart of the method of step S330 in FIG. 7.

[0021] FIG. 11 shows a structural block diagram of a call summary processing apparatus according to an embodiment of the present application.

[0022] FIG. 12 shows a structural block diagram of an electronic device according to an embodiment of the present application.

[0023] FIG. 13 shows a storage unit for storing or carrying program codes for implementing a call summary processing method according to an embodiment of the present application. DETAILED DESCRIPTION

[0024] In order to make the personnel in the technical field better understand the scheme of the present application, the technical scheme in the embodiments of the present application will be described clearly and completely below in conjunction with the drawings in the embodiments of the present application.

[0025] With the rapid development of mobile communication technology, more and more users communicate through the form of voice calls in work and life, greatly improving the convenience of communication.

[0026] The inventor found that in order to help record the key information in the call process, the content recorded by manual operation of the user is taken as a call note in the related art. However, this recording method is easy to miss important call information, resulting in incomplete call tags, and cannot meet the user's demand for viewing the key information of the call content, affecting the user experience.

[0027] The inventor found through long-term research that the call summary corresponding to the target call data can be displayed on a specified interface, the call summary includes a specified keyword, and the specified keyword represents key detail information in the target call data; then, in response to a touch operation acting on the specified keyword, jump to the corresponding position of the specified keyword in the target call data. Thus, by displaying the specified keyword corresponding to the key detail information in the target call data in the call summary corresponding to the target call data, the user can jump to the original position of the specified keyword to view the key detail information by touching the specified keyword, greatly meeting the user's demand for checking the key information in the call data, and improving the user experience.

[0028] Therefore, in order to improve the above problems, the inventor proposes a call summary processing method, device, electronic device and computer readable storage medium provided by the present application, which can meet the user's demand for checking the key information in the call data and improve the user experience.

[0029] In order to better understand the scheme described in the embodiments of the present application, the terms involved in the description process of the embodiments of the present application are briefly explained as follows:

[0030] AndesGPT: It is a generative large language model based on a hybrid cloud architecture, which integrates continuous learning of Chinese dialogue data, instruction fine-tuning, human feedback reinforcement learning, and knowledge enhancement. AndesGPT has strong Chinese reasoning and coding capabilities, and can perform well on complex Chinese problems and database query tasks.

[0031] Prompt: As an instruction or question (also known as a prompt word or phrase), it can guide ChatGPT to generate a corresponding response. It tells ChatGPT what we want, and ChatGPT generates a corresponding response based on the prompt. This interactive way is intuitive and natural, making ChatGPT easy to use. However, a correct and effective prompt is the key to obtaining high-quality output. A good prompt can make ChatGPT's output more accurate, in-depth, and creative, while an inappropriate prompt may cause the output to deviate from the topic or have low quality.

[0032] The embodiments of the present application will be described in detail below with reference to the accompanying drawings.

[0033] Please refer to FIG. 1, which shows a flowchart of a call summary processing method according to an embodiment of the present application. The embodiment provides a call summary processing method, which can be applied to an electronic device, such as a mobile phone, a computer, a tablet, or a wearable electronic device. The type of electronic device is not limited. The method includes:

[0034] Step S110: Display the call summary corresponding to the target call data on the specified interface, wherein the call summary includes a specified keyword, and the specified keyword represents the key detail information in the target call data.

[0035] The target call data represents the voice data in the call process, i.e., the call voice data. The call summary represents the summary of the call content in the voice call process, and includes a specified keyword representing the key detail information in the target call data. The specified interface can be understood as an interface in the electronic device for displaying the call summary.

[0036] Optionally, the specified interface can include one or more pages that are different from the call interface. When the content length of the call summary is short, the specified interface can include one page; when the content length of the call summary is long, the specified interface can include multiple pages, which can be displayed in the form of page turning or other forms, without limitation.

[0037] The electronic device in the embodiments of the present application is configured with an instant messaging application program, and the specific type of the instant messaging application program can not be limited, and the user can make voice call (or video call) with other users through the instant messaging application program. Optionally, the instant messaging application program can include a call summary generation function, which can be in the default open state when the instant messaging application program is in the open state; or the call summary generation function can be triggered to be opened when the user makes voice (or video) call (i.e. in the call state) through the instant messaging application program. Among them, the user can be the calling party of the voice (or video) call, or the answering party of the voice (or video) call.

[0038] Specifically, the call summary generation function can be automatically triggered to be opened when the duration of the voice call made by the user through the instant messaging application program reaches a preset duration (i.e. the duration of the call is greater than or equal to the preset duration, and the specific value of the preset duration can not be limited, for example, the preset duration can be half an hour), so as to prevent the user from forgetting the call content when the call time is long; the call summary generation function can be automatically triggered to be opened when the frequency of the voice call made by the user through the instant messaging application program is greater than or equal to a preset frequency (the specific value can not be limited, for example, five times a day), so as to intelligently help the user record the call content of the frequently used communication application program; the call summary generation function can be automatically triggered to be opened when the object of the voice call made by the user through the instant messaging application program is greater than a specified number (for example, 2 or the like, and the specific value is not limited), or the call summary generation function can be automatically triggered to be opened when the object of the voice call made by the user through the instant messaging application program is a specified user (for example, a user followed by the user, a leader, a superior, a teacher, etc.), so as to intelligently generate a call summary.

[0039] Optionally, the call summary generation function of the instant messaging application program can be in the default open state in this mode. In this way, a specified interface can be called when the call is completed, and the call summary corresponding to the target call data is displayed on the specified interface; or the specified interface can be called in the process of the call, and the call summary corresponding to the target call data is dynamically displayed on the specified interface (i.e. the content of the call summary is adjusted constantly along with the progress of the call process, until the complete call summary is obtained when the call is completed).

[0040] Optionally, if the call summary generation function of the instant messaging application program is in the default open state, the voice data in the whole call process can be taken as the target call data; and if the user needs to trigger the call summary generation function of the instant messaging application program to be opened, the time when the call summary generation function is triggered to be opened and the call voice data thereafter can be taken as the target call data.

[0041] Optionally, the target call data can be derived from voice call data or video call data (containing voice call data).

[0042] Step S120: in response to the touch operation acting on the specified keyword, jumping to the corresponding position of the specified keyword in the target call data.

[0043] The call summary in the embodiments of the present application can include an automatic jumping function, which specifically can include: if a touch operation acting on a specified keyword is detected, the automatic jumping function can automatically jump to the corresponding position of the specified keyword in the target call data, so as to jump to the key information of the detailed viewing requirement. The corresponding positions of different specified keywords in the target call data are different, i.e., the corresponding positions of each specified keyword in the target call data are unique, so as to avoid the phenomenon that accurate jumping cannot be achieved.

[0044] It should be noted that the specific type of the touch operation can not be limited, for example, the touch operation can be clicking (including single clicking, double clicking or multiple clicking, etc.), long pressing, pressing, etc.

[0045] In the embodiments of the present application, in order to achieve accurate jumping, the specified keyword can include a content identifier and a position identifier, wherein the content identifier can be used to highlight the importance of the specified keyword, and the position identifier can be used to quickly find the detailed information mentioned by the specified keyword. In this way, in response to the touch operation acting on the specified keyword, the corresponding position of the specified keyword in the target call data can be jumped to based on the content identifier and the position identifier.

[0046] It should be noted that the contents of different specified keywords can be the same, but the position identifiers of different specified keywords are different, i.e., the corresponding positions of each specified keyword in the target call data are unique, so as to avoid the phenomenon that accurate jumping cannot be achieved.

[0047] It should be noted that the specific type of the touch operation can not be limited, for example, the touch operation can be clicking (including single clicking, double clicking or multiple clicking, etc.), long pressing, pressing, etc. <e>Designated keywords< / e> The form of adding the content identifier to the specified keyword can not be limited, for example, the content identifier can be added to the specified keyword in the form of <e1>Designated keywords< / e1> to highlight the importance of the specified keyword. In order to facilitate distinction, the position identifier can be added to the content identifier, and the position identifier is unique. For example, the unique position identifier can be added to the content identifier of the specified keyword in the form of

[0048] In a specific application scenario, please refer to FIG. 2, which shows an example diagram of a call summary (i.e., the telephone summary shown in FIG. 2) displayed on a specified interface according to the embodiments of the present application. The specified keyword in the call summary is marked with <ex>Designated keywords< / ex>, X = 1, 2, 3, …, 9" form. It should be noted that for the same specified keyword, there are content identifier and location identifier in the call record data and the telephone summary (call summary), as shown in FIG. 3, which shows an example diagram of the call record data corresponding to the call summary shown in FIG. 2 according to the embodiment of the present application. Then in this way, if the touch operation acting on the specified keyword in the call summary is detected, the target call data (i.e. the call record data) corresponding to the specified keyword can be jumped to in response to the touch operation acting on the specified keyword. For example, if the user touches the specified keyword "Zhi Fei Technology" in the telephone summary shown in FIG. 2, it will jump to the place shown in FIG. 3. <e3>ZhiFei Technology< / e3>

[0049] The call summary processing method provided by the embodiment can make the user jump to the original text position of the specified keyword to view the key detail information only by touching the specified keyword, greatly meeting the user's demand for reviewing the key information in the call data and improving the user experience.

[0050] By adding content identifier and location identifier to the key detail information in the call data, and generating a call summary including the key detail information in the call data, and the content identifier and location identifier of the key detail information, the important detail information in the call record can be retained, so as to facilitate meeting the user's demand for reviewing the key information in the call data and improving the user experience.

[0051] Meanwhile, the automatic jump to the corresponding position of the specified keyword in the target call data is triggered in a touch manner, so that the key information of the detailed review requirement can be jumped to.

[0052] Please refer to FIG. 4, which shows a flow chart of a call summary processing method according to another embodiment of the present application. The call summary processing method can be applied to an electronic device, which can be a mobile phone, a computer, a tablet or a wearable electronic device, and the type of the electronic device can not be limited. The method comprises:

[0053] Step S210: displaying a call summary corresponding to the target call data in a specified interface, wherein the call summary includes a specified keyword, and the specified keyword represents the key detail information in the target call data.

[0054] The specific implementation of step S210 can refer to the related description of step S110 in the foregoing embodiments, which will not be repeated here.

[0055] ​Step S220: in response to the touch operation acting on the specified keyword, jumping to the corresponding position of the specified keyword in the target call data.

[0056] The implementation of step S220 can refer to the related description of step S120 in the foregoing embodiments, which will not be repeated here.

[0057] In this embodiment, in order to improve the convenience and intelligence of information inquiry, the specified keyword in the generated call summary can be associated with the corresponding additional function. In this way, if a user touches a specified keyword, the corresponding additional function option of the specified keyword can be popped up, and the user can select any option and perform subsequent related operations. If the specified keyword is an address, the corresponding additional function can be a navigation function. If the specified keyword is a phone number, the corresponding additional function can be a function of making a call, sending a message, saving a number, copying a number, etc.

[0058] In a specific application scenario, please refer to FIG. 5, which shows an example of a display interface provided by an embodiment of the present application, in which touching a specified keyword pops up the corresponding additional function option of the specified keyword. As shown in FIG. 5, if a user touches the specified keyword (here, a phone number) "138****1171", the display interface of the corresponding additional function option of the phone number is popped up, which includes functions of making a call, sending a message, saving a number, copying a number, etc. The display interface can be superimposed on the call summary display interface or can be suspended on the call summary display interface. The specific display manner can not be limited.

[0059] It should be noted that, in order to avoid misoperation, the touch operation triggering the jumping function and the touch operation triggering the display of the additional function can be different. For example, the touch operation triggering the jumping function can be a single-click operation, and the touch operation triggering the display of the additional function can be a double-click operation. Alternatively, the touch operation triggering the jumping function can be a single-click operation, and the touch operation triggering the display of the additional function can be a long-press operation. Alternatively, the type of the touch operation triggering the jumping function and the type of the touch operation triggering the display of the additional function can not be limited.

[0060] Step S230: displaying the specified keyword in the call summary in a specified format.

[0061] In this embodiment, in order to help the user quickly identify important details in the call summary, the specified keywords in the call summary can be displayed in a specified format. The specific form of the specified format can not be limited, for example, the specified keywords in the call summary can be displayed by bolding the font, the specified keywords in the call summary can be displayed by highlighting, the specified keywords in the call summary can be displayed by adding specified characters, or the specified keywords in the call summary can be displayed by any combination of the above display methods.

[0062] Optionally, the call summary in this embodiment can be displayed in a new page, for example, a new page can be popped up to display the generated call summary when the call ends. Optionally, in order to facilitate ensuring the review validity period of the call summary, the call summary can be saved as a "note" or other file storage format.

[0063] In a specific application scenario, please refer to FIG. 6, which shows an example of displaying specified keywords in the call summary in a specified format according to an embodiment of the present application. As shown in FIG. 6, for the call summary shown in FIG. 2, the font of the specified keywords in the call summary can be bolded and enlarged to emphasize the importance of the details information indicated by the specified keywords.

[0064] The call summary processing method provided in this embodiment enables the user to jump to the original text position of the specified keywords by touching the specified keywords, so as to view the key details information, greatly meeting the user's review needs for the key information in the call data and improving the user experience.

[0065] Meanwhile, by displaying the specified keywords in the call summary in a specified format, the user can quickly identify important details in the call summary, improving the user's review efficiency. By triggering the pop-up of the additional function options corresponding to the specified keywords in a touch manner, the convenience and intelligence of information review can be improved.

[0066] Please refer to FIG. 7, which shows a flowchart of a call summary processing method according to another embodiment of the present application. This embodiment provides a call summary processing method, which can be applied to an electronic device, which can be a mobile phone, a computer, a tablet computer or a wearable electronic device, and the specific type of the electronic device can not be limited. The method comprises:

[0067] Step S310: obtaining specified keywords in the to-be-processed call data.

[0068] Optionally, the to-be-processed call data can be derived from voice call data or video call data (including voice call data).

[0069] The to-be-processed call data in the embodiment can include coarse-grained keywords and fine-grained keywords. Compared with the coarse-grained keywords, the fine-grained keywords can be understood as smaller and more specific detailed information, and the coarse-grained keywords can be understood as more general and wider range information. For example, assuming that a piece of call voice data is "go to Xiaozhao's house for dinner tomorrow afternoon 3 o'clock, Xiaoli and Xiaoma will also go", then the keywords "tomorrow afternoon 3 o'clock", "go", "Xiaozhao's house", and "dinner" can be coarse-grained keywords thereof, and the keywords "Xiaoli", "Xiaozhao", and "also go" can be fine-grained keywords thereof. Optionally, the contents of the coarse-grained keywords and the fine-grained keywords can be different.

[0070] The specified keyword can be understood as the above-mentioned fine-grained keyword, and the specified keyword represents a key detailed information in the to-be-processed call data. The specific content form of the specified keyword can not be limited, for example, the specified keyword can be a character, a number, a word, or a plurality of words.

[0071] In order to facilitate the generation of a call summary with more complete content, the acquisition of the specified keyword in the to-be-processed call data can be triggered when it is detected that the electronic device is in a call state. The specific acquisition process is described below.

[0072] As an implementation manner, referring to FIG. 8, step S310 can include:

[0073] Step S311: performing voice recognition on the to-be-processed call data to obtain to-be-processed text data.

[0074] In the process of acquiring the specified keyword in the to-be-processed call data, as an implementation manner, the to-be-processed call data can be first subjected to voice recognition, that is, the to-be-processed call data is converted from voice data to text data to obtain call text data as to-be-processed text data.

[0075] The voice recognition on the to-be-processed call data can be performed by using a voice recognition model. For example, the to-be-processed call data can be subjected to voice recognition by using an ASR (Automatic Speech Recognition) technology. Optionally, the specific type of the voice recognition model is not limited herein.

[0076] Step S312: performing content recognition on the to-be-processed text data according to a preset keyword library to obtain the specified keyword.

[0077] In this embodiment, a plurality of different conversation scenarios can be defined in advance, and the fine-grained keywords required or possibly required by each of the conversation scenarios can be stored in a preset keyword library. For example, the preset keyword library can be divided into a work conversation scenario keyword library and a life conversation scenario keyword library, and the work conversation scenario keyword library can be further divided into keywords of a financial industry conversation scenario, keywords of an Internet industry conversation scenario, and keywords of a service industry conversation scenario.

[0078] As an implementation, the content of the to-be-processed text data can be recognized according to the preset keyword library to obtain the specified keywords. Specifically, the to-be-processed text data can be compared with the preset keyword library. If the to-be-processed text data contains the same keywords as the preset keyword library, the same keywords can be extracted from the to-be-processed text data as the specified keywords.

[0079] Optionally, in a specific conversation scenario, the specified keywords can be information such as time, detailed address, name, company, institution, train number, and telephone number.

[0080] As another implementation, the instant messaging application can be configured with a conversation summary generation function. The conversation interface of the instant messaging application can display a conversation summary generation control corresponding to the conversation summary generation function. The user can start or stop the conversation summary generation function by touching the conversation summary generation control.

[0081] In this way, if the conversation summary generation function of the instant messaging application is not in a default start state, or if the conversation summary generation function is not automatically started with the start of the instant messaging application, the conversation summary generation function can be adjusted to the start state in response to the touch operation on the conversation summary generation control when the electronic device is detected to be in the conversation state, and the specified keywords in the to-be-processed conversation data are triggered to be acquired. The specified keywords represent the key details in the to-be-processed conversation data.

[0082] In a specific application scenario, refer to FIG. 9, which shows a display example of the call summary generation control provided by the embodiments of the present application. As shown in FIG. 9, the current user is dialing a call to classmate A, which can be regarded as being in a call state. At this time, the call summary generation control can be displayed in any position (or any blank position) of the call interface, and the call summary generation control at this time is in a closed state by default. As an implementation manner, if the user touches the call summary generation control, the call summary generation function can be triggered to be adjusted to an open state, so as to generate a call summary in real time during the call process, and help the user record the key information and key detail information in the call process. It should be noted that the style of the call summary generation control is only used as an example for illustration, and does not constitute a limitation on the present solution. The specific display position of the call summary generation control can not be limited, and the display mode can also be fixed (not limited to floating).

[0083] As another implementation manner, the specified keyword in the to-be-processed call data can be obtained by specifying a language model. The specified language model is a pre-trained large language model, and the specified language model has the ability to output a call summary with a specified keyword with content identification and position identification according to call data with content identification and position identification.

[0084] As an implementation manner, the sample call data can be obtained first, the sample call data includes synthetic call data and real call data, then the target keyword in the sample call data can be obtained, and the content identification and the position identification are added to the target keyword to obtain the specified call data, and the call content summary including the target keyword, the content identification and the position identification is generated based on the specified call data as the call summary of the sample call data. Then, the specified call data and the call summary of the sample call data can be input into the language model to be trained to train the language model to be trained, and the language model when the loss function of the language model to be trained converges is used as the specified language model.

[0085] The sample call data is used to train the language model. In order to save the construction cost of the sample call data, the sample call data can include synthetic call data and real call data. The synthetic call data can be understood as call record data generated by a large model according to a set call topic. Optionally, the call topic of the synthetic call data and the call topic of the real call data can be the same or different.

[0086] To reduce the labor cost, the large model can be used to obtain the target keyword (and the specified keyword meaning is the same as the foregoing, the specific content is different) in the sample call data, and the target keyword is added with the content identifier and the position identifier as described in the foregoing embodiments to obtain the specified call data. Further, the large model is used to generate the call content summary including the target keyword, the content identifier of the target keyword, and the position identifier based on the specified call data, as the call summary of the sample call data.

[0087] Specifically, a suitable prompt can be written for the large model to require the large model to generate the call summary of the sample call data according to the prompt.

[0088] Optionally, the large model at this time does not have the ability to output the call summary of the specified keyword with the content identifier and the position identifier based on the call data with the content identifier and the position identifier. Therefore, the call summary generated by the large model can be understood as a pseudo-call summary. In order to improve the accuracy of the language model training data generation, the correctness of the call summary generated by the large model can be checked by manual participation, and the call summary corrected by manual participation can be used as the final call summary of the sample call data.

[0089] As an implementation manner, the specified call data and the final call summary of the sample call data can be input into the language model to be trained to train the language model to be trained (herein, supervised training), and the language model when the loss function of the language model to be trained converges is used as the specified language model.

[0090] The language model to be trained can generate a reference call summary based on the sample call data fused with the content identifier and the position identifier, and compare the reference call summary with the foregoing final call summary to calculate the difference between the two to obtain the loss function. The language model to be trained can be continuously fine-tuned until the loss function of the language model to be trained converges.

[0091] The specific type of the specified language model can not be limited. For example, the specified language model in the embodiment can be an AndesGPT model (a large generative language model based on a hybrid cloud architecture) or a ChatGPT language model.

[0092] The specified keyword in the call data to be processed is obtained by the specified language model, which can reduce the labor cost and improve the generation efficiency of the call summary.

[0093] Meanwhile, by taking the synthesized conversation data and the real conversation data as sample conversation data, the cost can be reduced; by combining the large model and the artificially generated language model training process conversation summary comparison data, the accuracy of the language model training data generation can be improved, and thus the accuracy of the conversation summary generation can be improved.

[0094] Step S320: adding content identification and position identification to the specified keyword to obtain target conversation data.

[0095] In order to emphasize the importance of the specified keyword for the generated conversation summary, and to avoid missing important details of the to-be-processed conversation data and failing to generate an information-complete conversation summary, as an implementation, content identification can be added to the specified keyword extracted from the to-be-processed text data. As described above, the specific form of the content identification can not be limited, for example, the content identification can be added to the specified keyword in the form of <e>Designated keywords< / e> to highlight the importance of the specified keyword.

[0096] Further, in order to facilitate the user to quickly find the details information mentioned by the specified keyword, the position identification can be added to the specified keyword, specifically, the unique position identification can be added to the content identifier of the specified keyword, so that the user can jump directly from the specified keyword in the conversation summary to the original text based on the unique position identification, to realize detailed viewing of the details information mentioned by the specified keyword. For example, taking the above example as an example, the unique position identification can be added to the content identifier of the specified keyword in the form of <e1>Designated keywords< / e1> , wherein "1" represents the position identification.

[0097] In this embodiment, the to-be-processed conversation data to which the content identification and the position identification are added to the specified keyword is taken as the target conversation data. The target conversation data is the same as the to-be-processed conversation data in content, except that the specified keyword in the target conversation data is added with the content identification and the position identification.

[0098] In one specific application scenario, as shown in FIG. 3, an effect diagram of the to-be-processed conversation data to which the content identification and the position identification are added to the specified keyword is shown. As shown in FIG. 3, the to-be-processed conversation data (the conversation record shown in FIG. 3) is the conversation data between the other party and the party, and the to-be-processed conversation data includes the specified keywords such as "Li Zong, tomorrow, Zhi Fei Technology, Huang Jun, tomorrow afternoon 3 o'clock, Qianhai Tea Room, Chen Secretary, 138****1177". Each specified keyword is added with content identification and position identification. In order to emphasize the importance of the specified keyword for the generated conversation summary through the content identification, and to quickly find the details information mentioned by the specified keyword through the position identification.

[0099] Step S330: generating a call content summary including the specified keyword, the content identifier and the location identifier based on the target call data, as a call summary of the target call data.

[0100] In the case that the target call data including the specified keyword added with the content identifier and the location identifier is obtained, a call content summary including the specified keyword, the content identifier and the location identifier of the specified keyword can be generated based on the target call data, as a call summary of the to-be-processed call data. That is, the generated call summary includes not only the specified keyword, but also the content identifier and the location identifier of the specified keyword. The specific generation process of the call summary is described as follows:

[0101] Referring to FIG. 10, as an implementation manner, step S330 can include:

[0102] Step S331: generating a call content summary statement based on the target call data.

[0103] In order to improve the accuracy of generating the call summary, the generation process of the call summary can be divided into three parts. As an implementation manner, a call content summary statement can be generated based on the target call data first. Specifically, the target call data can be summarized in one sentence first, that is, the all topics and contents of the target call data are highly summarized in one sentence. The one-sentence summary is a high-level summary of the target call data, which can facilitate the user to quickly understand the call information and save time.

[0104] Optionally, in order to improve the readability of the call summary, the number of words of the call content summary statement can be less than / greater than a first threshold value. The specific value of the first threshold value can not be limited, and the number of words of the call content summary statement corresponding to the number of words of the target call data can be different, that is, the value of the first threshold value can be adjusted. The value of the first threshold value in the embodiment can be 80.

[0105] Step S332: adding the specified keyword including the content identifier and the location identifier to the call content summary statement to obtain a to-be-processed call content summary.

[0106] In the case that the call content summary statement is generated, the generated call content summary statement can be expanded to enrich the detailed information of the generated call summary. Specifically, the specified keyword including the content identifier and the location identifier can be added to the call content summary statement to obtain a to-be-processed call content summary. At this time, the to-be-processed call content summary contains fine-grained call information, which can meet the user's demand for viewing the detailed information of the call content.

[0107] Step S333: point-by-point record is performed on the to-be-processed call content summary, and a call summary of the target call data is obtained.

[0108] Further, in order to make the generated call summary clear and orderly to show to the user, point-by-point record can be performed on the to-be-processed call content summary, that is, the core events in the call data are point-by-point recorded, and then the obtained content can be taken as the call summary of the to-be-processed call data.

[0109] In a specific application scenario, as shown in FIG. 2, an example diagram of the generated call summary provided by the embodiment of the present application is shown. As shown in FIG. 2, the call summary (the telephone summary shown in FIG. 2) includes two types of result titles, that is, “smart summary” and “to-do list”, the smart summary includes a one-sentence summary and point-by-point recorded core events. The one-sentence summary in FIG. 2 is “Boss of Chen knows and confirms the contact information of the contact”. <e1>Tomorrow< / e1> And <e3>ZhiFei Technology< / e3> the <e4>Huang Jun< / e4> boss knows and confirms the contact information of the contact <e8>Secretary Chen< / e8> ”.The point-by-point recorded core events include “1. Chen asks if he is free, and says that he wants to introduce <e1>Tomorrow< / e1> the <e3>ZhiFei Technology< / e3> boss to know each other. 2. It is agreed that <e4>Huang Jun< / e4> to meet and <e5>Tomorrow afternoon 3 o'clock< / e5> to chat <e7>Qianhai Tea Room< / e7> and <e4>Huang Jun< / e4> to discuss issues, and I need to make the plan and appeal clear. 3. The phone number of the boss's contact, Chen's secretary, is confirmed <e8>Material exclusivity period< / e8> ”. <e9>Supply cycle< / e9> <e4>Huang Jun< / e4> <e6> 138****1177< / e6>

[0110] The to-do list can be understood as a future schedule. In order to improve the aesthetic and readability of the summary, the total number of characters of the point-by-point recorded core events and the to-do list can be less than a second threshold value, and the specific value of the second threshold value can not be limited. Optionally, the second threshold value in the embodiment can be 350 characters.

[0111] Step S340: displaying the call summary corresponding to the target call data on a specified interface, the call summary including a specified keyword, the specified keyword representing key detailed information in the target call data.

[0112] The specific implementation of step S340 can refer to the related description of step S110 in the foregoing embodiment, and will not be described here.

[0113] Step S350: in response to a touch operation acting on the specified keyword, jumping to a corresponding position of the specified keyword in the target call data.

[0114] ​​The specific implementation of step S350 can refer to the related description of step S120 in the foregoing embodiments, and details are not described herein.

[0115] Step S360: Displaying the specified keyword in the call summary in a specified format.

[0116] The specific implementation of step S360 can refer to the related description of step S230 in the foregoing embodiments, and details are not described herein.

[0117] The call summary processing method provided in this embodiment can add content identifiers and position identifiers to the key detail information in the call data, and generate a call summary including the key detail information in the call data and the content identifiers and position identifiers of the key detail information, so that important detail information in the call record can be retained, the user's demand for checking the key information in the call data can be met, and the user experience is improved.

[0118] Meanwhile, the content identifiers can be used to emphasize the importance of the specified keyword for the generated call summary, and the position identifiers can be used to quickly find the detail information mentioned by the specified keyword.

[0119] By dividing the generation process of the call summary into three parts, the core events and detail information can be retained while summarizing the call content, and the generated to-do list can prevent the user from forgetting important arrangements, and the user call experience can be improved. By limiting the number of words of the generated call summary, the readability of the call summary can be improved.

[0120] When the call summary generation control is touched, the call summary generation function is triggered to be adjusted to an enabled state, so that the call summary can be generated in real time according to the user's needs during the call process, and the user can be helped to record the key information and key detail information in the call process.

[0121] Please refer to FIG. 11, which is a structural block diagram of a call summary processing apparatus provided in an embodiment of the present application. The call summary processing apparatus 400 provided in this embodiment can run on an electronic device, and the apparatus 400 includes a call summary display module 410 and a call summary processing module 420.

[0122] The call summary display module 410 is configured to display a call summary corresponding to target call data on a specified interface, where the call summary includes a specified keyword, and the specified keyword represents key detail information in the target call data.

[0123] The call summary processing module 420 is configured to respond to a touch operation on the specified keyword, and jump to a corresponding position of the specified keyword in the target call data.

[0124] In the embodiments of the present application, the specified keyword includes a content identifier and a position identifier. In this way, the call summary processing module 420 can be configured to jump to the corresponding position of the specified keyword in the target call data in response to the touch operation on the specified keyword based on the content identifier and the position identifier.

[0125] Optionally, the apparatus 400 can further include a display module configured to display the specified keyword in the call summary in a specified format.

[0126] Optionally, the apparatus 400 can further include a call summary generation module configured to: obtain a specified keyword in the to-be-processed call data before displaying the call summary corresponding to the target call data on a specified interface; add a content identifier and a position identifier to the specified keyword to obtain the target call data; and generate a call content summary including the specified keyword, the content identifier and the position identifier based on the target call data as the call summary of the target call data.

[0127] Optionally, the apparatus 400 can further include a call summary generation module configured to: obtain a specified keyword in the to-be-processed call data before displaying the call summary corresponding to the target call data on a specified interface; add a content identifier and a position identifier to the specified keyword to obtain the target call data; and generate a call content summary including the specified keyword, the content identifier and the position identifier based on the target call data as the call summary of the target call data.

[0128] Optionally, the apparatus 400 can further include a call summary generation module configured to: obtain a specified keyword in the to-be-processed call data before displaying the call summary corresponding to the target call data on a specified interface; add a content identifier and a position identifier to the specified keyword to obtain the target call data; and generate a call content summary including the specified keyword, the content identifier and the position identifier based on the target call data as the call summary of the target call data.

[0129] Optionally, the apparatus 400 can further include a call summary generation module configured to: obtain a specified keyword in the to-be-processed call data before displaying the call summary corresponding to the target call data on a specified interface; add a content identifier and a position identifier to the specified keyword to obtain the target call data; and generate a call content summary including the specified keyword, the content identifier and the position identifier based on the target call data as the call summary of the target call data.

[0130] Optionally, the apparatus 400 can further include a call summary generation module configured to: obtain a specified keyword in the to-be-processed call data before displaying the call summary corresponding to the target call data on a specified interface; add a content identifier and a position identifier to the specified keyword to obtain the target call data; and generate a call content summary including the specified keyword, the content identifier and the position identifier based on the target call data as the call summary of the target call data.

[0131] Optionally, the apparatus 400 can further include a model training module configured to obtain sample conversation data, the sample conversation data including synthesized conversation data and real conversation data; obtain a target keyword in the sample conversation data, and add a content identifier and a position identifier to the target keyword to obtain specified conversation data, and generate a conversation content summary including the target keyword, the content identifier and the position identifier based on the specified conversation data as a conversation summary of the sample conversation data; and input the specified conversation data and the conversation summary of the sample conversation data into a language model to be trained to train the language model to be trained, and use a language model when a loss function of the language model to be trained converges as the specified language model.

[0132] Those skilled in the art can clearly understand that, for the convenience and brevity of description, the specific working process of the above-described apparatus and modules can refer to the corresponding process in the foregoing method embodiments, which will not be described here.

[0133] In several embodiments provided in the present application, the coupling between the modules can be electrical, mechanical or other forms of coupling.

[0134] In addition, each functional module in each embodiment of the present application can be integrated in one processing module, or each module can exist physically alone, or two or more modules can be integrated in one module. The integrated module can be realized in the form of hardware or in the form of a software functional module.

[0135] Please refer to FIG. 12, based on the above conversation summary processing method and apparatus, the present embodiment further provides an electronic device 100 which can execute the foregoing conversation summary processing method. The electronic device 100 includes a memory 102 and one or more (only one is shown in the figure) processors 104 coupled with each other, and the communication line connection between the memory 102 and the processor 104. The memory 102 stores programs that can execute the contents in the foregoing embodiments, and the processor 104 can execute the programs stored in the memory 102.

[0136] The processor 104 can include one or more processing cores. The processor 104 connects various parts within the entire electronic device 100 by various interfaces and lines, performs various functions of the electronic device 100 and processes data by running or executing instructions, programs, code sets or instruction sets stored in the memory 102, and calling data stored in the memory 102. Optionally, the processor 104 can be implemented in at least one of a hardware form of a digital signal processing (DSP), a field-programmable gate array (FPGA), and a programmable logic array (PLA). The processor 104 can be integrated with a combination of one or more of a central processing unit (CPU), a graphics processing unit (GPU), and a modem. Among them, the CPU mainly processes operating systems, user interfaces, and application programs; the GPU is responsible for rendering and drawing display content; and the modem is used for processing wireless communication. It can be understood that the above-mentioned modem can also not be integrated into the processor 104, but can be implemented by a separate communication chip.

[0137] The memory 102 can include a random access memory (RAM) and can also include a read-only memory (ROM). The memory 102 can be used to store instructions, programs, codes, code sets or instruction sets. The memory 102 can include a program storage area and a data storage area, wherein the program storage area can store instructions for implementing an operating system, instructions for implementing at least one function (such as a touch function, a sound playing function, an image playing function, etc.), instructions for implementing the above-mentioned various embodiments, etc. For example, the memory 102 can store an apparatus. The apparatus can be the apparatus 400 mentioned above. The data storage area can also store data created by the electronic device 100 in use (such as a phone book, audio and video data, chat record data), etc.

[0138] Please refer to FIG. 13, which shows a structural block diagram of a computer readable storage medium provided by an embodiment of the present application. The computer readable medium 500 stores program codes therein, and the program codes can be called and executed by a processor to perform the methods described in the above method embodiments.

[0139] The computer-readable storage medium 500 can be an electronic memory such as a flash memory, EEPROM (electrically erasable programmable read-only memory), EPROM, hard disk or ROM. Optionally, the computer-readable storage medium 500 comprises a non-transitory computer-readable medium. The computer-readable storage medium 500 has storage space for program code 510 that performs any of the method steps of the above-described methods. These program codes can be read from or written to one or more computer program products. The program code 510 can be compressed, for example, in a suitable form.

[0140] To sum up, the embodiment of the present application provides a call summary processing method and device, electronic equipment and storage medium. The method can be applied to electronic equipment. The method displays a call summary corresponding to target call data on a specified interface. The call summary includes a specified keyword, which represents key detail information in the target call data. Then, in response to a touch operation on the specified keyword, the method jumps to a corresponding position of the specified keyword in the target call data. Thus, by displaying the specified keyword corresponding to the key detail information in the target call data in the call summary corresponding to the target call data, the user can jump to the original position of the specified keyword to check the key detail information by touching the specified keyword. This greatly meets the user's demand for checking key information in call data and improves the user experience.

[0141] Finally, it should be noted that: the above embodiments are only used to illustrate the technical solutions of the present application, but not to limit them; although the present application has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that they can still modify the technical solutions recorded in the foregoing embodiments, or make equivalent replacement for part of the technical features; and these modifications or replacements do not drive the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present application.

Claims

1. A call summary processing method, wherein: The method comprises: Displaying a call summary corresponding to the target call data on a designated interface, the call summary including designated keywords representing key detail information in the target call data; In response to a touch operation on the designated keyword, the user jumps to a corresponding position of the designated keyword in the target call data.

2. The method according to claim 1, wherein The designated keyword includes a content identifier and a location identifier, and the response to the touch operation on the designated keyword jumps to a corresponding location of the designated keyword in the target call data, including: In response to a touch operation on the designated keyword, the user jumps to a corresponding position of the designated keyword in the target call data based on the content identifier and the position identifier.

3. The method according to claim 1 or 2, wherein: The method further comprises: Display specified keywords in the call summary in a specified format.

4. The method according to claim 3, wherein: Before displaying the call summary corresponding to the target call data on the designated interface, the method further includes: Get the specified keywords in the call data to be processed; Adding a content identifier and a location identifier to the designated keyword to obtain target call data; A call content summary including the designated keyword, the content identifier, and the location identifier is generated based on the target call data as a call summary of the target call data.

5. The method according to claim 4, wherein The generating, based on the target call data, a call content summary including the designated keyword, the content identifier, and the location identifier as a call summary of the target call data includes: generating a call content summary statement based on the target call data; Adding the designated keywords including the content identifier and the location identifier to the call content summary statement to obtain a summary of the call content to be processed; The summary of the call content to be processed is recorded point by point to obtain a call summary of the target call data.

6. The method according to claim 4, wherein: The step of obtaining the specified keyword in the call data to be processed includes: Performing speech recognition on the call data to be processed to obtain text data to be processed; The content of the text data to be processed is identified according to a preset keyword library to obtain specified keywords.

7. The method according to claim 4, wherein: The step of obtaining the specified keyword in the call data to be processed includes: In response to a touch operation on a call summary generation control, a specified keyword in the call data to be processed is obtained.

8. The method according to any one of claims 1 to 7, wherein: The method further comprises: In response to a touch operation on the designated keyword, displaying additional function options corresponding to the designated keyword; An operation corresponding to a target function option is executed, where the target function option is an option selected by the user from the displayed additional function options.

9. The method according to claim 8, wherein The touch operation for triggering the display of the additional function option is different from the touch operation for triggering the execution of the jump.

10. The method according to claim 9, wherein: The touch operation for triggering the display of the additional function option includes a single-click operation, and the touch operation for triggering the execution of the jump includes a double-click operation; Alternatively, the touch operation for triggering the display of the additional function option includes a single-click operation, and the touch operation for triggering the execution of the jump includes a long-press operation.

11. The method according to any one of claims 1 to 7, wherein: The call summary is generated when the call summary generation function is in the activated state. The call summary generation function is activated or deactivated by a call summary generation control, and the call summary generation control is displayed in the call interface.

12. The method according to any one of claims 1 to 7, wherein: The designated interface is displayed when the call is completed, or the designated interface is displayed during the call.

13. The method according to any one of claims 1 to 7, wherein: The specified keyword is displayed in one or more of a variety of display modes, including: Display the specified keywords in the call summary in bold font; Display specified keywords in the call summary by highlighting them; Displays the specified keywords in the call summary by adding the specified characters.

14. A call summary processing device, wherein: The device comprises: A call summary display module, configured to display a call summary corresponding to target call data on a designated interface, wherein the call summary includes designated keywords representing key detail information in the target call data; The call summary processing module is used to respond to a touch operation on the designated keyword and jump to a corresponding position of the designated keyword in the target call data.

15. An electronic device, wherein: including one or more processors and memory; One or more programs are stored in the memory and configured to be executed by the one or more processors, and the one or more programs are configured to execute the method according to any one of claims 1 to 13.

16. A computer-readable storage medium, wherein: The computer-readable storage medium stores program code, wherein when the program code is executed by a processor, the method according to any one of claims 1 to 13 is executed.

Images