Input recommendation method, input recommendation device, mobile terminal, and readable storage medium

Abstract

The application discloses an input recommendation method, an input recommendation device, a mobile terminal and a readable storage medium. The input recommendation method comprises the following steps: obtaining input content input by a user and a source type of the input content; wherein the source type comprises having been in an input box and having copied content sent by an opposite party; obtaining at least one piece of recommended content based on the input content and the source type; wherein, in the case that the source type is the having been in the input box, the recommended content represents candidate continuation content of the input content; in the case that the source type is the having copied the content sent by the opposite party, the recommended content represents candidate reply content of the input content; and the at least one piece of recommended content is displayed on an input interface. The above scheme can improve the accuracy and input efficiency of input recommendation.

Description

Technical Field

[0001] The present application relates to the technical field of data processing, and in particular, to an input recommendation method, an input recommendation apparatus, a mobile terminal, and a readable storage medium. Background Art

[0002] With the continuous development of mobile Internet, instant messaging has been widely used in work and daily life, and users are increasingly pursuing more efficient input methods. To improve input efficiency, existing input recommendation methods provide an input association function during data processing. However, current input association functions usually provide users with subsequent associated characters based on the characters already input by the user. For example, when the user inputs "hao", characters such as "de", "ba", and "ya" will be associated and recommended for the user to select. Such a recommendation method has limited improvement on input efficiency, and cannot meet users' demand for higher-precision input in different scenarios such as active input or replying to the other party. In view of this, how to improve the precision of input recommendation and input efficiency has become an urgent problem to be solved. Summary of the Invention

[0003] The main technical problem solved by the present application is to provide an input recommendation method, an input recommendation apparatus, a mobile terminal, and a readable storage medium, which can improve the precision of input recommendation and input efficiency.

[0004] To solve the above technical problem, a first aspect of the present application provides an input recommendation method, comprising: acquiring input content input by a user and a source type of the input content; wherein the source type includes content already in an input box and content sent by a counterpart that has been copied; acquiring at least one piece of recommendation content obtained based on the input content and the source type; wherein when the source type is the content already in the input box, the recommendation content represents candidate continuation content of the input content, and when the source type is the content sent by the counterpart that has been copied, the recommendation content represents candidate reply content of the input content; displaying the at least one piece of recommendation content on an input interface.

[0005] To address the aforementioned technical problems, a second aspect of this application provides an input recommendation device, comprising: an input module, an acquisition module, and a display module. The input module acquires user-inputted content and the source type of the input content; wherein the source type includes content already in an input box and content copied from another party; the acquisition module acquires at least one recommended content based on the input content and the source type; wherein, when the source type is "already in an input box," the recommended content represents candidate continuation content of the input content, and when the source type is "content copied from another party," the recommended content represents candidate response content to the input content; the display module displays the at least one recommended content on the input interface.

[0006] To address the aforementioned technical problems, a third aspect of this application provides a mobile terminal, including a display screen, a memory, and a processor, wherein the display screen and the memory are respectively coupled to the processor, the display screen is used at least to display content to a user and to allow the user to input or copy content, the memory stores program instructions, and the processor is used to execute the program instructions to implement the input recommendation method described in the first aspect.

[0007] To address the aforementioned technical problems, a fourth aspect of this application provides a computer-readable storage medium having program instructions stored thereon, which, when executed by a processor, implement the input recommendation method described in the first aspect.

[0008] The above solution obtains the user's input content and the corresponding source type. The source type includes content already in the input box and content copied from another party. When the source type is already in the input box, at least one recommended content is obtained based on the input content and the source type. This recommended content is a candidate continuation of the input content. Thus, when the input box contains input content, candidate continuation content is recommended to the user, which can improve input efficiency and adapt to scenarios where the user inputs themselves. When the source type is content copied from another party, at least one recommended content is obtained based on the input content and the source type. This recommended content is a candidate reply to the input content, which can improve reply efficiency and adapt to application scenarios where the user replies to the other party. The obtained at least one recommended content is displayed on the input interface so that the user can select or directly obtain recommended content that matches the scenario, thereby improving the accuracy and efficiency of input recommendations. Attached Figure Description

[0009] To more clearly illustrate the technical solutions in the embodiments of this application, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort. Wherein:

[0010] Figure 1 This is a flowchart illustrating one implementation method of the input recommendation method in this application;

[0011] Figure 2 This is a schematic diagram illustrating an application scenario of one implementation method of the input recommendation method in this application;

[0012] Figure 3 This is a schematic diagram illustrating an application scenario of another implementation of the input recommendation method of this application;

[0013] Figure 4 This is a schematic diagram illustrating an application scenario of another implementation of the input recommendation method in this application;

[0014] Figure 5 This is a flowchart illustrating another implementation of the input recommendation method of this application;

[0015] Figure 6 This is a flowchart illustrating another implementation of the input recommendation method in this application;

[0016] Figure 7 This is a schematic diagram illustrating an application scenario of another implementation of the input recommendation method of this application;

[0017] Figure 8 This is a schematic diagram of one embodiment of the input recommendation device of this application;

[0018] Figure 9 This is a schematic diagram of the structure of one embodiment of the mobile terminal of this application;

[0019] Figure 10 This is a schematic diagram of one embodiment of the computer-readable storage medium of this application. Detailed Implementation

[0020] The technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this application, and not all embodiments. Based on the embodiments of this application, all other embodiments obtained by those of ordinary skill in the art without creative effort are within the scope of protection of this application.

[0021] If the technical solution of this application involves personal information, the product using this technical solution has clearly informed the user of the personal information processing rules and obtained the user's voluntary consent before processing the personal information. If the technical solution of this application involves sensitive personal information, the product using this technical solution has obtained the user's separate consent before processing the sensitive personal information, and also meets the requirement of "express consent". For example, on the device for processing personal information or the device for collecting personal information, the personal information processing rules are clearly indicated / informed, or the user's authorization is obtained through pop-up information or by asking the user to upload their personal information; wherein, the personal information processing rules may include information such as the personal information processor, the purpose of personal information processing, the processing method, and the types of personal information processed.

[0022] In this paper, the terms "system" and "network" are often used interchangeably. The term "and / or" describes the relationship between related objects, indicating that three relationships can exist. For example, A and / or B can represent: A alone, A and B simultaneously, or B alone. Additionally, the character " / " generally indicates that the preceding and following related objects have an "or" relationship. Furthermore, "many" in this paper means two or more.

[0023] Please see Figure 1 , Figure 1 This is a flowchart illustrating an implementation method of the input recommendation method of this application, which includes:

[0024] S101: Obtain the user's input content and the source type of the input content, wherein the source type includes content already in the input box and content copied from the other party.

[0025] Specifically, the system obtains the user's input content and the source type corresponding to the input content. The source type includes content already entered in the input box and content already copied from the other party.

[0026] In one application method, the user's input content and the source type corresponding to the input content are obtained from the input box or clipboard. When it is detected that the input box contains already entered content, the input content is obtained from the input box, and the source type of the input content is determined to be already in the input box. When it is detected that the clipboard contains copied content that has not been pasted into the input box, the input content is obtained from the clipboard, and the source type of the input content is determined to be copied content sent by the other party.

[0027] In another application, in response to the input or pasting of content into the input box, the input content is retrieved from the input box, and the source type of the input content is determined to be already in the input box. In response to the clipboard containing copied content, it is determined whether the copied content is consistent with the content sent by the other party in the current application where the input interface is located. If they are consistent, the input content is retrieved from the clipboard, and the source type of the input content is determined to be copied content sent by the other party.

[0028] Optionally, the above steps are performed based on the input method or the smart terminal where the input method is located. When the input method includes a clipboard function, the copied content is obtained based on the clipboard built into the input method. When the input method does not include a clipboard function, the copied content is obtained based on the clipboard on the smart terminal where the input method is running.

[0029] In one implementation scenario, the input method includes a clipboard function. When a user enters or pastes content into the input box, the input content is retrieved from the input box, and the source type of the input content in the input box is determined to be already in the input box. This is to adapt to the user entering any content into the input box and then suggest candidate continuation content.

[0030] In another implementation scenario, when a user copies some content to the clipboard, if it is detected that the user has not pasted the copied content into the input box within the first interval, the copied content is retrieved from the clipboard as the input content, and the source type of the input content is determined to be copied content sent by the other party. Thus, the source type of the content copied by the user but not pasted into the input box is defined as copied content sent by the other party. This is to adapt to the situation where the user switches between different applications and still uses the copied content as the input content, so that the candidate reply content of the input content can be obtained based on the input content and the source type.

[0031] In another implementation scenario, when a user copies some content to the clipboard, if it is detected that the user has not pasted the copied content into the input box within the first interval, the copied content is retrieved from the clipboard and compared with at least one piece of content in the current application that has the closest timestamp to the content sent by the other party. If the overlap rate exceeds a preset ratio, the copied content on the clipboard is used as the input content, and the source type of the input content is defined as copied content sent by the other party. This is to adapt to the scenario where the user is having a real-time conversation in the current application, reduce the probability of excessive recommendations in non-reply scenarios, and improve the user experience.

[0032] S102: Obtain at least one recommended content based on the input content and the source type, wherein, if the source type is already in the input box, the recommended content represents the candidate continuation content of the input content, and if the source type is copied from the other party's sent content, the recommended content represents the candidate reply content of the input content.

[0033] Specifically, when the source type is "already in the input box", at least one recommended content is obtained based on the input content and the source type. This recommended content is a candidate continuation of the input content. Thus, when the input box includes the input content, the system recommends candidate continuation of the input content to the user, which can improve input efficiency and adapt to scenarios where the user inputs the content themselves. When the source type is "content sent by the other party has been copied", at least one recommended content is obtained based on the input content and the source type. This recommended content is a candidate reply to the input content, which can improve reply efficiency and adapt to application scenarios where the user replies to the other party.

[0034] In one implementation scenario, when the source type is already in the input box, multiple candidate successive content items based on the input content and source type are obtained as multiple recommended content items for the input content, and at least some of the recommended content items have different styles or semantics, thereby providing users with more choices.

[0035] In another implementation scenario, when the source type is copied content sent by the other party, multiple candidate response contents based on the input content and source type are obtained as multiple recommended contents for the input content, and at least some of the recommended contents have different styles or semantics, thereby providing users with more choices.

[0036] In another implementation scenario, when the source type is "copying content sent by the other party", a candidate reply content based on the input content and the source type is obtained as a recommended content for the input content. The style of the recommended content matches the user's expression habits, thereby improving the accuracy of the recommended content.

[0037] In a specific implementation scenario, the recommended content is output by the Large Language Model (LLM), and the input content and its corresponding source type jointly generate the prompts of the Large Language Model.

[0038] It should be noted that all types of content, including input content and recommended content, must include at least one of the following: text, symbols, and emojis.

[0039] S103: Display at least one recommended item on the input interface.

[0040] Specifically, at least one recommended content is displayed on the input interface so that users can select or directly obtain recommended content that matches the scenario, thereby improving the accuracy and efficiency of input recommendations.

[0041] Furthermore, the input interface includes at least the display interface of the input method and the display interface of the application in which the input method is located.

[0042] In one implementation scenario, when the source type is already in the input box, multiple candidate continuation contents for input content are obtained. These candidate continuation contents are displayed sequentially in the candidate area of ​​the input interface for the user to select. The user can swipe to view all candidate continuation contents in the candidate area and select one by clicking to input it into the input box. After the candidate continuation contents are entered into the input box, the system returns to the step of obtaining the user's input content and the source type of the input content for continuous recommendation, thereby improving the user's input efficiency.

[0043] In another implementation scenario, when the source type is "copying content sent by the other party", multiple candidate replies to the input content are obtained. At least one candidate reply to the input content is displayed in the candidate area of ​​the input interface for the user to view, increasing the user's choices. The user can click to select a candidate reply to enter into the input box. After the candidate reply is entered into the input box, the system returns to the step of obtaining the user's input content and the source type of the input content for continuous recommendation, thereby improving the user's input efficiency.

[0044] In another implementation scenario, when the source type is "copying content sent by the other party", a candidate reply is obtained from the input content. The candidate reply is directly added to the input box and displayed to the user, improving reply efficiency. After the candidate reply is entered into the input box, the process can return to the step of obtaining the user's input content and the source type of the input content for continuous recommendation, thereby improving the user's input efficiency.

[0045] In a specific implementation scenario, after a user has sent multiple messages, the semantic information of the user's sent messages is obtained. If the source type is already in the input box, the candidate continuation messages of the multiple input messages are displayed in the candidate area of ​​the input interface in order of their matching degree with the semantic information of the sent messages. The input interface is the display interface of the input method.

[0046] In another specific implementation scenario, after a user has sent multiple messages, the semantic information of the messages sent by the user is obtained. If the source type is already in the input box, the candidate continuation messages of the multiple input messages are displayed in the candidate area of ​​the input interface in order of their matching degree with the semantic information of the sent messages. The input interface is the display interface of the application where the input method is located.

[0047] In another specific implementation scenario, after a user has interacted with the other party at least once, the semantic information of the interacted content is obtained. When the source type is copied content sent by the other party, the candidate reply content with the highest semantic matching degree with the interacted content is selected from the candidate reply content of multiple input content and displayed in the candidate area of ​​the input interface.

[0048] In another specific implementation scenario, after a user has interacted with the other party at least once, the semantic information of the interacted content is obtained. If the source type is copied content sent by the other party, the candidate reply content with the highest semantic matching degree with the interacted content is selected from the candidate reply content of multiple input content, and added to the input box of the input method to be displayed to the user.

[0049] In another specific implementation scenario, when a user is interacting with another party, if the source type is "copying content sent by the other party", a candidate reply is obtained from the input content, and the corresponding candidate reply is added to the input box of the input method and displayed to the user.

[0050] For a specific application scenario, please refer to Figure 2 , Figure 2 This is a schematic diagram of an application scenario for an implementation method of the input recommendation method of this application. When the source type is "content sent by the other party has been copied", at least one recommended content is obtained based on the input content and the source type, including: obtaining a candidate reply content based on the input content and the source type; displaying at least one recommended content on the input interface, including: adding the corresponding candidate reply content to the input box for display.

[0051] Specifically, when input content is obtained and the source type of the input content is "copied content sent by the other party", a candidate reply content is obtained based on the input content and the source type. The candidate reply content is then directly added to the input box for display. After the user selects to send, the candidate reply content can be sent to the other party. Furthermore, after the candidate reply content is added to the input box, the input content in the input box can be continuously associated and recommended, thereby obtaining the candidate follow-up content corresponding to the input content in the input box, thus improving the efficiency of the user when replying.

[0052] What is understandable is, such as Figure 2 The image shows the input interface of the input method, in the sub-... Figure 2 In version .1, the copied input includes: "I'm fine, what are you busy with?". When the system detects that the user has selected a smart reply option (such as "child") on the input method's interface, the system will respond accordingly. Figure 2 When the dashed box in .1 is shown, the smart reply selection area displays the processing progress (such as sub-processes). Figure 2 As shown in the dashed box in .2), one of the obtained candidate replies includes: "While replying to your message, I'm enjoying a coffee." In addition to the text, the candidate reply also includes an emoji corresponding to a coffee cup. The candidate reply is directly displayed in the input box (e.g., in the text box). Figure 2(As shown in the dashed box in .3), after the candidate reply content is added to the input box, the input content in the input box can be continuously associated and recommended to obtain the candidate continuation content corresponding to the input content in the input box. The candidate continuation content includes: "Come over here!" (e.g., Zi) Figure 2 (As shown in the dashed box in .4), and the candidate response content includes not only the text mentioned above, but also the corresponding facial expression.

[0053] In another specific application scenario, please refer to Figure 3 , Figure 3 This is an application scenario illustration of another implementation of the input recommendation method of this application. When the source type is "content sent by the other party has been copied", obtaining at least one recommended content based on the input content and the source type includes: obtaining multiple candidate reply contents based on the input content and the source type; displaying at least one recommended content on the input interface includes: displaying one of the candidate reply contents in the candidate area of ​​the input interface until the user selects the corresponding candidate reply content in the candidate area to add to the input box, or the user adds content in the input box, or the user selects an extension option on the input interface; in response to detecting that the user selects an extension option on the input interface, displaying all candidate reply contents in the candidate extension area of ​​the input interface until the user selects any candidate reply content in the candidate area to add to the input box, or the user selects to exit the candidate extension area on the input interface.

[0054] Specifically, when input content is obtained and the source type of the input content is "copied content sent by the other party", multiple candidate reply contents are obtained based on the input content and source type. Then, one of the candidate reply contents is displayed in the candidate area. At this time, the user can directly click on the candidate reply content in the candidate area to add the candidate reply content to the input box. The user can also directly add content, such as typing or pasting content. If the user is not satisfied with the candidate reply content in the candidate area at this time, the user can select the expansion option in the input interface.

[0055] Furthermore, when it is detected that the user selects an expanded option on the input interface, all the acquired candidate responses are displayed in the candidate expanded area of ​​the input interface. The candidate expanded area is the area outside the input box in the input interface. Figure 3 The location of the dashed box on the right side of the image allows users to select a candidate reply from the candidate expansion area and add it to the input box, or they can select the back button on the input interface to exit the candidate expansion area.

[0056] It is understandable that, such as Figure 3The image shows the input interface of the input method. The user copied the input content, which included: "What do you want to eat tonight?". Multiple candidate responses were obtained based on the input content and source type. One of these candidate responses included: "Scallion pancakes with cola tonight." In addition to the text, the candidate responses also included emojis for scallion pancakes and cola. The user could then select this candidate response to add to the input box, or they could directly add content to the input box. In this case, the candidate responses in the candidate area would be replaced by the candidate continuation content corresponding to the input content in the input box.

[0057] Furthermore, if it is detected that the user selects an extended option (such as a sub-option) on the input method's input interface... Figure 3 The dashed box in .1 indicates the candidate expansion area of ​​the input interface (such as the sub-area). Figure 3 As shown in the dashed box in .2), all candidate reply content is displayed in the candidate expansion area. The content displayed will not be described in detail here. Users can select any candidate reply content in the candidate expansion area and add it to the input box, or they can add content directly in the input box. When it is detected that the user selects the back button in the input method interface, the candidate expansion area is exited. In this way, the candidate expansion area provides users with multiple reply options and increases the probability that the candidate reply content matches the user's expected content.

[0058] In another specific application scenario, please refer to Figure 4 , Figure 4 This is an application scenario illustration of another implementation of the input recommendation method of this application. When the source type is already in the input box, at least one recommended content is obtained based on the input content and the source type, including: obtaining multiple candidate continuation content based on the input content and the source type; displaying at least one recommended content on the input interface, including: displaying multiple candidate continuation content in the candidate area of ​​the input interface until the user selects any candidate continuation content in the candidate area to add to the input box, or the user adds content in the input box.

[0059] Specifically, when input content is obtained and its source type is already in the input box, multiple candidate continuation content are obtained based on the input content and source type. These multiple candidate continuation content are then displayed sequentially in the candidate area of ​​the input interface. Users can swipe to view all candidate continuation content in the candidate area. At this time, users can select any candidate continuation content and add it to the input box. Users can also add content directly to the input box. When new content is added to the input box, the input content in the input box is continuously suggested to adapt to the user's active input scenario and improve input efficiency.

[0060] It is understandable that, such as Figure 4The image shows the input interface of the input method. The user's initial input in the input box includes: "How have you been lately?". Multiple candidate continuation messages are retrieved based on the input content and source type, including: "Haven't been in touch for a long time," "How are you?", "Brother," and other candidate continuation messages that can be viewed by swiping (such as sub-continuation messages). Figure 4 (As shown in the dashed box in .1), the user can select a candidate continuation content to add to the input box (such as a sub-box). Figure 4 (As shown in the dashed box in .2), the user's selected candidate continuation content includes: "Haven't been in touch for a long time." Returning to the point where the user's input content and its source type are obtained, new candidate continuation content is acquired, including: "How have you been lately?", "How are you?", and other candidate continuation content that can be swiped to view (such as sub-content). Figure 4 (As shown in the dashed box in .3), the entire input process can continue for multiple rounds (such as sub-rounds). Figure 4 (As shown in the dashed box in .4), this improves the user's input efficiency, and users can also directly add content in the input box.

[0061] The above solution obtains the user's input content and the corresponding source type. The source type includes content already in the input box and content copied from another party. When the source type is already in the input box, at least one recommended content is obtained based on the input content and the source type. This recommended content is a candidate continuation of the input content. Thus, when the input box contains input content, candidate continuation content is recommended to the user, which can improve input efficiency and adapt to scenarios where the user inputs themselves. When the source type is content copied from another party, at least one recommended content is obtained based on the input content and the source type. This recommended content is a candidate reply to the input content, which can improve reply efficiency and adapt to application scenarios where the user replies to the other party. The obtained at least one recommended content is displayed on the input interface so that the user can select or directly obtain recommended content that matches the scenario, thereby improving the accuracy and efficiency of input recommendations.

[0062] Please see Figure 5 , Figure 5 This is a flowchart illustrating another embodiment of the input recommendation method of this application, the method including:

[0063] S501: Obtain the user's input content and the source type of the input content, wherein the source type includes content already in the input box and content copied from the other party.

[0064] Specifically, step S501 is similar to step S101 in the above embodiments. For details, please refer to the corresponding embodiments. This application will not repeat them here.

[0065] S502: Get the application type corresponding to the current application where the input interface is located.

[0066] Specifically, determine the current application in which the input interface is located, and determine the application type corresponding to the current application. The application type is at least related to the purpose of the application.

[0067] Optionally, the application types should include at least communication, finance, shopping, office, and leisure.

[0068] S503: Obtain at least one recommended content based on the input content and source type.

[0069] Specifically, it retrieves at least one recommended content item that matches the style of the application type, based on the input content, source type, and application type. Therefore, users can perceive that they are receiving recommended content that better matches the style of the application type during interaction.

[0070] In one implementation scenario, a first prompt is generated based on the input content, source type, and application type. The first prompt is provided to the large language model, and at least one recommended content is received from the large language model after responding to the first prompt. The recommended content matches the style of the application type.

[0071] Specifically, the input content, source type, and application type are combined to generate a first prompt, which is then provided to the large language model. The large language model responds to the first prompt and outputs at least one recommended content after responding to the first prompt, so that the recommended content matches the style of the application type. This allows users to perceive that when they input content on different types of applications, the candidate continuation content or candidate response content matches the style of the application type, thereby improving the user experience.

[0072] Optionally, the large language model is an autoregressive pre-training (GPT) model, which generates a first suggestion by combining the input content, source type, and application type, provides the first suggestion to the large language model, and requires the large language model to provide at least one recommended content.

[0073] In a specific implementation scenario, the autoregressive model is the ChatGPT model. The first prompt is input into the ChatGPT model, where the first prompt includes the number of recommended content items. For example, the first prompt is: please output at least ten possible subsequent input items that match the style of the following input items that match the application type, or please output at least one candidate content that matches the style of the application type and can respond to the following input items.

[0074] In another implementation scenario, based on the input content, source type, and application type, recommended content matching the style of the application type is obtained from a recommendation library that matches the style of the application type.

[0075] S504: Display at least one recommended item on the input interface.

[0076] Specifically, step S504 is similar to step S103 in the above embodiments. For details, please refer to the corresponding embodiments. This application will not repeat them here.

[0077] Unlike the previous embodiments, by obtaining the application type of the application, at least one recommended content that matches the style of the application type is obtained, thereby improving recommendation accuracy and user experience.

[0078] Please see Figure 6 , Figure 6 This is a flowchart illustrating another implementation of the input recommendation method of this application, the method comprising:

[0079] S601: Obtain the user's input content and the source type of the input content, wherein the source type includes content already in the input box and content copied from the other party.

[0080] Specifically, step S601 is similar to step S101 in the above embodiments. For details, please refer to the corresponding embodiments. This application will not repeat them here.

[0081] S602: Obtain reference content for the current application where the input interface is located.

[0082] Specifically, when it is detected that the input interface has switched from another application to the current application, and the switching time is within a preset time interval before the input content, if the application type of the application before the switch is the same as the application type of the application after the switch, then the reference content includes at least part of the dialogue content in the application before the switch; otherwise, the reference content represents at least part of the dialogue content in the current application.

[0083] It should be noted that the application type is at least related to the purpose of the application. For example, the application type should at least include communication (e.g., WeChat and QQ), finance (e.g., Alipay and various bank apps), shopping (e.g., Taobao, JD.com and Pinduoduo), office (e.g., DingTalk and WeChat Work) and leisure (e.g., Meituan and Dianping).

[0084] In one application mode, after obtaining the input content, the current timestamp corresponding to the input content is determined, and it is determined whether the input interface has switched applications within a preset time interval between the current timestamps. If it has, it is determined whether the application type corresponding to the application before the switch is the same as the application type corresponding to the application after the switch. If they are the same, the reference content includes at least part of the dialogue content in the application before the switch. If they are different, the reference content only includes at least part of the dialogue content in the current application.

[0085] Understandably, when the application type before switching is different from the application type after switching, and the current application does not include the content already displayed, the reference content is in a state of no content until dialogue content is detected in the current application, at least part of the dialogue content will be used as the reference content.

[0086] In one implementation scenario, when a user switches from another application to the current application and adds or copies content to the input box within a preset time interval, the system detects whether the application type before and after the switch is the same. If the two applications belong to the same type, at least a portion of the dialogue content from the previous application is used as reference content for the current application. This allows the user to better access reference content when switching between applications of the same type, ensuring that the recommended content obtained after switching between applications of the same type semantically matches the dialogue content in the previous application, thus improving the user experience. The preset time interval can be any customizable time such as 30 seconds, 45 seconds, or 60 seconds; this application does not impose specific limitations on it.

[0087] In a specific implementation scenario, when a user switches from another application to the current application and adds or copies content to the input box to the clipboard within a preset time interval, and the application type of the application before the switch is the same as the application type after the switch, then all dialogue content in the applications before and after the switch that is within the time threshold of the current input content is used as reference content.

[0088] In another specific implementation scenario, when a user switches from another application to the current application and adds or copies content to the input box to the clipboard within a preset time interval, and the application type of the application before the switch is the same as the application type after the switch, then all dialogue content in the applications before and after the switch that is separated from the current input content by a preset number of steps will be used as reference content.

[0089] In another specific implementation scenario, when a user switches from another application to the current application and adds or copies content to the input box to the clipboard within a preset time interval, and the application type before the switch is different from the application type after the switch, then all dialogue content in the current application that is within the time threshold of the current input content is used as reference content.

[0090] In another specific implementation scenario, when a user switches from another application to the current application and adds or copies content to the input box to the clipboard within a preset time interval, and the application type before the switch is different from the application type after the switch, then all dialogue content in the current application that is a preset number of times away from the current input content is used as reference content.

[0091] Optionally, before obtaining at least one recommended content that semantically matches the reference content based on the input content, source type, and reference content, the method further includes: obtaining the current timestamp corresponding to the input content, and deleting dialogue content in the reference content that exceeds a time threshold with respect to the current timestamp; wherein the length of the time threshold is greater than the length of a preset time interval.

[0092] Specifically, when the input interface is detected to switch from other applications to the current application, the application type of the current application is determined, an input interface matching the application type of the current application is obtained, and then the user is shown an input interface matching the application type of the current application in the current application.

[0093] Furthermore, when the application type changes, the user should clearly perceive the difference between the application types before and after the switch. In this case, the reference content should not include the dialogue content in the application before the switch. When the application type does not change, the user should clearly perceive that the application types before and after the switch are the same. In this case, the reference content may include the dialogue content in the application before the switch, making the user's perception more intuitive.

[0094] In one implementation scenario, changes to the input interface are displayed through changes in either color or pattern.

[0095] S603: Obtain at least one recommended content based on the input content and source type.

[0096] Specifically, it retrieves at least one recommended content that semantically matches the reference content, based on the input content, source type, and reference content. Therefore, users can perceive during interaction that they are receiving recommended content that more semantically matches the reference content.

[0097] In one implementation scenario, a second prompt is generated based on the input content, source type, and reference content, and the second prompt is provided to the large language model; at least one recommended content output by the large language model after responding to the second prompt is received, wherein the recommended content semantically matches the reference content.

[0098] Specifically, the input content, source type, and reference content are combined to generate a second prompt, which is then provided to the large language model. The large language model responds to the second prompt and outputs a result. The system receives at least one recommended content output by the large language model after responding to the second prompt, so that the recommended content matches the semantics of the reference content. This allows users to perceive that when they input content in the current application, the obtained candidate continuation content or candidate response content matches the semantics of the reference content, thereby improving the user experience.

[0099] In one implementation scenario, the large language model is an autoregressive pre-training (GPT) model. It generates a second suggestion by combining the input content, source type, and application type. The second suggestion is then provided to the large language model, which is required to provide at least one recommended content.

[0100] In a specific implementation scenario, the autoregressive model is the ChatGPT model. The second prompt is input into the ChatGPT model, where the second prompt includes the number of recommended content items. For example, the first prompt is to output at least ten possible subsequent input items that match the semantics of the reference content, or to output at least one candidate content that matches the semantics of the reference content and can respond to the following input content.

[0101] In another implementation scenario, based on the input content, source type, and reference content, recommended content that matches the semantics of the reference content is obtained from a recommendation library that matches the semantics of the reference content.

[0102] Optionally, before obtaining at least one recommended content that semantically matches the reference content based on the input content, source type, and reference content, the method further includes: obtaining the current timestamp corresponding to the input content, and deleting dialogue content in the reference content that exceeds a time threshold with respect to the current timestamp; wherein the length of the time threshold is greater than the length of a preset time interval.

[0103] Specifically, the current timestamp corresponding to the input content is obtained, and the dialogue content in the reference content is updated. If the reference content includes dialogue content whose time difference with the current timestamp exceeds a time threshold, the dialogue content that has exceeded the time threshold is deleted. This improves the rationality and reference value of the reference content, so that users can perceive that after a long period of no input, they can get at least one recommended content corresponding to the input content again, and the recommended content at this time is no longer related to dialogue content whose time difference with the current timestamp exceeds a time threshold.

[0104] It should be noted that the length of the time threshold is greater than the preset time interval. The preset time interval is used to determine whether it is necessary to detect changes in the application type before and after the switch, while the time threshold is used to determine whether the dialogue content that is furthest from the current timestamp in the reference content needs to be deleted.

[0105] In a specific implementation scenario, the preset time interval is 60 seconds and the time threshold is 300 seconds. When a user switches the input interface from another application to the current application, for example, from Taobao to JD.com (JD.com is the current application), and the input method receives new input content within 60 seconds of switching, the input method uses the current timestamp of the input content to retrieve the conversation content from the previous application within 300 seconds of the current timestamp as reference content. Since the user switches between shopping applications and receives new input content in a short time, the user is likely selecting similar products in different applications. Therefore, using at least part of the conversation content from the previous application as reference content allows the user to perceive that the recommended content is still relevant to the product they want to buy when shopping. However, if the input method does not receive new input content within 60 seconds of switching to the current application, it is determined that although the user has switched between applications of the same type, they do not have a need to purchase similar products. Therefore, the conversation content from the previous application is no longer used as reference content, improving the user experience. The specific values ​​of the preset time interval and time threshold can be customized, and this application does not impose specific restrictions on them.

[0106] S604: Display at least one recommended item on the input interface.

[0107] Specifically, step S604 is similar to step S103 in the above embodiments. For details, please refer to the corresponding embodiments. This application will not repeat them here.

[0108] It should be noted that the above implementation methods based on application type and reference content can also be combined. Recommended content can be obtained by combining application type, reference content, input content and its corresponding source type, so that the recommended content matches the style of the application type and the semantics of the reference content when obtaining recommended content.

[0109] Unlike the previous embodiments, this method obtains reference content from the current application, taking application switching into account when obtaining the reference content, and obtains at least one recommended content that semantically matches the reference content, thereby improving recommendation accuracy and user experience.

[0110] Optionally, in any of the above embodiments, in response to the source type of the input content being already in the input box and the user selecting an adjustment mode on the input interface, at least one piece of adjustment content based on the input content is obtained, each piece of adjustment content is used as recommended content of the input content, and the process directly proceeds to the step of displaying at least one piece of recommended content on the input interface.

[0111] Specifically, when the input content is obtained and its source type is already in the input box, if it is detected that the user selects an adjustment mode on the input interface, at least one adjustment content based on the input content is obtained, and each adjustment content is used as recommended content for the input content. This directly leads to the step of displaying at least one recommended content on the input interface. This is to adapt to the situation where the user needs to adjust the input content when actively inputting, and to provide the user with at least one richer adjustment content as recommended content for the input content, thereby improving the user experience.

[0112] Optionally, the adjustment of the content includes at least one of adding, deleting, and modifying the input content, thereby enabling at least one of the functions of expanding, shortening, and polishing the input content.

[0113] Furthermore, displaying at least one recommended item on the input interface includes: displaying one of the adjustment items in the candidate area of ​​the input interface until the user selects the corresponding adjustment item in the candidate area, replacing the input content in the input box with the selected adjustment item, or detecting that the user selects a change option on the input interface; in response to detecting that the user selects a change option on the input interface, displaying all adjustment items in the candidate extended area of ​​the input interface until the user selects any adjustment item in the candidate area, replacing the input content in the input box with the selected adjustment item, or detecting that the user selects to exit the candidate extended area on the input interface.

[0114] Specifically, please refer to Figure 7 , Figure 7 This is a schematic diagram illustrating an application scenario of another implementation of the input recommendation method of this application. Specifically, when multiple adjustment items are obtained, one of the adjustment items is displayed in the candidate area of ​​the input interface. At this time, the user can directly click on the adjustment item in the candidate area to replace the input item in the input box with the selected adjustment item. Alternatively, if the user is not satisfied with the adjustment item in the candidate area, the user can select a replacement option on the input interface. Figure 7 The "Change" option is shown in the image.

[0115] Furthermore, when it is detected that the user selects a change option in the input interface, all the obtained adjustment content is displayed in the candidate expansion area of ​​the input interface. The candidate expansion area is the area in the input interface other than the input box. At this time, the user can select an adjustment content in the candidate expansion area to add to the input box, or the user can select the back button on the input interface to exit the candidate expansion area.

[0116] It is understandable that, such as Figure 7 The image shows the input interface of the input method. The input box contains the input content, including "Forever Young," and the user can select an adjustment mode (such as sub-mode). Figure 7The chat mode (shown by the dashed box in .1) retrieves multiple adjustment options based on the input, one of which includes: "Age is just a number; having a youthful spirit is the most important thing." In addition to the above text, candidate replies also include star-shaped emojis. Users can then select this adjustment option to replace the content in the input box, improving input efficiency. Alternatively, users can choose to change the options (such as...). Figure 7 (as shown in the dashed box in .2).

[0117] Furthermore, if it is detected that the user selects the "Refresh" option on the input method's input interface, the candidate expansion area of ​​the input interface is displayed, and all adjustment content is shown in the candidate expansion area (such as sub-options). Figure 7 As shown in the dashed box in .3), the content displayed will not be elaborated further in this application. Users can select any adjustment content in the candidate expansion area to replace the content in the input box. As shown in the final input box, it includes: "We will all grow old, but never lose the desire for knowledge and the courage to explore the unknown." In addition to the above text, the final content in the input box also includes the emoticon corresponding to the fist (such as a fist). Figure 7 As shown in the dashed box in .4, users can also select the back button to exit the candidate expansion area in the input interface. When the user selects the back button in the input interface, the candidate expansion area is exited. In this way, the candidate expansion area provides users with a variety of adjustment content and increases the probability that the candidate expansion area matches the user's expected content.

[0118] In a specific implementation scenario, before obtaining at least one adjustment content based on the input content, it also includes: obtaining the user's expression habits.

[0119] Specifically, after a user logs into their account on the input method and authorizes the input method to collect their expression habits, the user's expression habits are obtained in the adjustment mode.

[0120] Furthermore, obtaining at least one adjustment based on the input content includes: obtaining at least one adjustment based on the input content and expression habits; wherein at least part of the adjustment matches the style of the expression habits.

[0121] Specifically, after obtaining the user's expression habits, at least one adjustment content that matches the style of the expression habits is obtained based on the input content and expression habits.

[0122] Optionally, multiple adjustment content items are obtained based on the input content and expression habits. Among them, at least some of the adjustment content items match the style of the expression habits and are displayed first, while the other adjustment content items are regular adjustment content items, and the regular adjustment content items do not refer to the user's expression habits.

[0123] In one implementation scenario, a third prompt is generated based on the input content and expression habits, and the third prompt is provided to the large language model; at least one adjustment content is received from the output of the large language model after responding to the third prompt; wherein the adjustment content matches the style of the expression habits.

[0124] Specifically, the input content and the user's expression habits are used to generate a third prompt, which is then provided to the large language model so that the large language model responds to the third prompt and outputs a result. The system receives at least one adjustment content output by the large language model after responding to the third prompt, so that the adjustment content matches the user's expression habits. This allows the user to perceive that the adjustment content obtained in the adjustment mode matches the style of their personal expression habits during interaction, thereby improving the user experience.

[0125] In another implementation scenario, based on the input content and expression habits, recommended content that matches the style of the expression habits is obtained from a recommendation library that matches the style of the expression habits.

[0126] It is understandable that in other specific implementation scenarios, user tags assigned by the user to the other party can also be obtained, thereby obtaining the adjustment content based on the input content and user tags. This application does not impose specific restrictions on this.

[0127] Please see Figure 8 , Figure 8 This is a schematic diagram of an embodiment of the input recommendation device of this application. The input recommendation device 80 includes an input module 800, an acquisition module 802, and a display module 804. The input module 800 is used to acquire the input content and the source type of the input content from the user. The source type includes content already in the input box and content copied from the other party. The acquisition module 802 is used to acquire at least one recommended content based on the input content and the source type. When the source type is "already in the input box", the recommended content represents the candidate continuation content of the input content. When the source type is "content copied from the other party", the recommended content represents the candidate reply content of the input content. The display module 804 is used to display at least one recommended content on the input interface.

[0128] In the above scheme, the input module 800 obtains the user's input content and the corresponding source type. The source type includes content already in the input box and content copied from the other party. When the source type is already in the input box, the acquisition module 802 obtains at least one recommended content based on the input content and the source type. This recommended content is a candidate continuation of the input content. Thus, when the input box contains input content, it recommends candidate continuation of the input content to the user, which can improve input efficiency and adapt to the scenario of user self-input. When the source type is content copied from the other party, the acquisition module 802 obtains at least one recommended content based on the input content and the source type. This recommended content is a candidate reply to the input content, which can improve reply efficiency and adapt to the application scenario of user replying to the other party. The display module 804 displays the at least one recommended content obtained on the input interface, so that the user can select or directly obtain recommended content that matches the scenario, thereby improving the accuracy of input recommendation and input efficiency.

[0129] Optionally, when the source type is "copy the content sent by the other party", the acquisition module 802 is also used to acquire a candidate reply based on the input content and the source type; the display module 804 is also used to add the corresponding candidate reply to the input box for display.

[0130] Optionally, when the source type is "copy the content sent by the other party", the acquisition module 802 is further used to acquire multiple candidate reply contents based on the input content and the source type; the display module 804 is further used to display one of the candidate reply contents in the candidate area of ​​the input interface until the user selects the corresponding candidate reply content in the candidate area to add to the input box, or the user adds content in the input box, or the user selects an extension option in the input interface; in response to detecting that the user selects an extension option in the input interface, all candidate reply contents are displayed in the candidate extension area of ​​the input interface until the user selects any candidate reply content in the candidate area to add to the input box, or the user selects to exit the candidate extension area in the input interface.

[0131] Optionally, when the source type is already in the input box, the acquisition module 802 is further used to acquire multiple candidate continuation contents based on the input content and the source type; the display module 804 is further used to display the multiple candidate continuation contents in the candidate area of ​​the input interface until the user selects any candidate continuation contents in the candidate area to add to the input box, or the user adds content in the input box.

[0132] Optionally, the acquisition module 802 is also used to acquire the application type corresponding to the current application where the input interface is located; and to acquire at least one recommended content that matches the style of the application type based on the input content, source type and application type.

[0133] Optionally, the acquisition module 802 is further configured to acquire reference content of the current application where the input interface is located; acquire at least one recommended content that semantically matches the reference content based on the input content, source type, and reference content; wherein, when it is detected that the input interface switches from other applications to the current application, and the switching time is within a preset time interval before the input content, if the application type corresponding to the application before the switch is the same as the application type corresponding to the application after the switch, then the reference content includes at least part of the dialogue content in the application before the switch; otherwise, the reference content represents at least part of the dialogue content in the current application.

[0134] Optionally, the acquisition module 802 is further configured to acquire the current timestamp corresponding to the input content and delete dialogue content in the reference content that exceeds a time threshold with respect to the current timestamp; wherein, the length of the time threshold is greater than the length of a preset time interval.

[0135] Optionally, the display module 804 is also configured to, in response to detecting that the input interface has switched from another application to the current application, obtain an input interface that matches the application type corresponding to the current application; and display the input interface that matches the application type corresponding to the current application in the current application.

[0136] Optionally, the acquisition module 802 is further configured to, in response to the input content source type being already in the input box and detecting that the user selects an adjustment mode on the input interface, acquire at least one adjustment content based on the input content, and use each adjustment content as recommended content for the input content; the display module 804 is further configured to display one of the adjustment contents in the candidate area of ​​the input interface until the user selects the corresponding adjustment content in the candidate area, and uses the selected adjustment content to replace the input content in the input box, or detects that the user selects a change option on the input interface; in response to detecting that the user selects a change option on the input interface, display all adjustment contents in the candidate expansion area of ​​the input interface until the user selects any adjustment content in the candidate area, and uses the selected adjustment content to replace the input content in the input box, or detects that the user selects to exit the candidate expansion area on the input interface.

[0137] Optionally, the acquisition module 802 is also used to acquire the user's expression habits; acquire at least one adjustment content based on the input content and expression habits; wherein at least part of the adjustment content matches the style of the expression habits.

[0138] Please see Figure 9 , Figure 9This is a schematic diagram of the structure of a mobile terminal according to an embodiment of this application. The mobile terminal 90 includes a display screen 900, a memory 901, and a processor 902. The display screen 900 and the memory 901 are respectively coupled to the processor 902. The display screen 900 is used at least to display content to the user and to allow the user to input or copy content. The memory 901 stores program instructions (not identified). The processor 902 is used to execute the program instructions to implement the input recommendation method described in any of the above embodiments. For related explanations, please refer to the detailed description of the above method embodiments, which will not be repeated here.

[0139] The above solution can improve the accuracy and efficiency of input recommendations.

[0140] Please see Figure 10 , Figure 10 This is a schematic diagram of one embodiment of the computer-readable storage medium of this application. The computer-readable storage medium 100 stores program instructions 1000, which, when executed by a processor, implement the input recommendation method described in any of the above embodiments. For further details on related content, please refer to the detailed description of the above method embodiments; it will not be repeated here.

[0141] The above solution can improve the accuracy and efficiency of input recommendations.

[0142] It should be noted that the units described as separate components may or may not be physically separate, and the components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the units can be selected to achieve the purpose of this embodiment according to actual needs.

[0143] Furthermore, the functional units in the various embodiments of this application can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit. The integrated unit can be implemented in hardware or as a software functional unit.

[0144] If the integrated unit is implemented as a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of this application, in essence, or the part that contributes to the prior art, or all or part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) or processor to execute all or part of the steps of the methods of various embodiments of this application. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0145] The above description is merely an embodiment of this application and does not limit the patent scope of this application. Any equivalent structural or procedural transformations made using the content of this application's specification and drawings, or direct or indirect applications in other related technical fields, are similarly included within the patent protection scope of this application.

Claims

1. An input recommendation method, characterized in that, The input recommendation method includes: Obtain the user's input content and the source type of the input content; wherein, the source type includes content already in the input box and content copied from the other party; Obtain at least one recommended content based on the input content and the source type; wherein, when the source type is "already in the input box", the recommended content represents the candidate continuation content of the input content, and when the source type is "content sent by the other party has been copied", the recommended content represents the candidate reply content of the input content; Display at least one recommended item on the input interface; Before obtaining at least one recommended content based on the input content and the source type, the method further includes: obtaining reference content of the current application where the input interface is located; wherein, when it is detected that the input interface is switched from another application to the current application, and the switching time is within a preset time interval before the input content, if the application type corresponding to the application before the switch is the same as the application type corresponding to the application after the switch, then the reference content includes at least part of the dialogue content in the application before the switch; otherwise, the reference content represents at least part of the dialogue content in the current application. The step of obtaining at least one recommended content based on the input content and the source type includes: obtaining at least one recommended content that semantically matches the reference content, based on the input content, the source type, and the reference content.

2. The method according to claim 1, characterized in that, When the source type is "content copied from another party", obtaining at least one recommended content based on the input content and the source type includes: Obtain a candidate response based on the input content and the source type; Displaying at least one recommended item on the input interface includes: The corresponding candidate response content is added to the input box for display.

3. The method according to claim 1, characterized in that, When the source type is "content copied from another party", obtaining at least one recommended content based on the input content and the source type includes: Obtain multiple candidate response contents based on the input content and the source type; Displaying at least one recommended item on the input interface includes: One of the candidate responses is displayed in the candidate area of ​​the input interface until the user selects the corresponding candidate response in the candidate area and adds it to the input box, or the user adds content in the input box, or the user selects an extension option in the input interface. In response to detecting that the user selects an expansion option on the input interface, all candidate reply content is displayed in the candidate expansion area of ​​the input interface until the user selects any candidate reply content in the candidate area to add to the input box, or the user selects to exit the candidate expansion area on the input interface.

4. The method according to claim 1, characterized in that, When the source type is already in the input box, obtaining at least one recommended content based on the input content and the source type includes: Obtain multiple candidate continuation contents based on the input content and the source type; Displaying at least one recommended item on the input interface includes: The multiple candidate continuation contents are displayed in the candidate area of ​​the input interface until the user selects any of the candidate continuation contents in the candidate area and adds it to the input box, or the user adds content in the input box.

5. The method according to claim 1, characterized in that, Before obtaining at least one recommended content based on the input content and the source type, the method further includes: Obtain the application type corresponding to the current application where the input interface is located; Obtaining at least one recommended content based on the input content and the source type includes: Obtain at least one recommended content that matches the style of the application type, based on the input content, the source type, and the application type.

6. The method according to claim 1, characterized in that, Before obtaining at least one recommended content that semantically matches the reference content, based on the input content, the source type, and the reference content, the method further includes: Obtain the current timestamp corresponding to the input content, and delete the dialogue content in the reference content that exceeds the time threshold of the current timestamp; wherein, the length of the time threshold is greater than the length of the preset time interval.

7. The method according to claim 1, characterized in that, Before obtaining the reference content of the current application where the input interface is located, the method further includes: In response to detecting that the input interface has switched from another application to the current application, an input interface matching the application type corresponding to the current application is obtained; The input interface that matches the application type corresponding to the current application is displayed in the current application.

8. The method according to claim 1, characterized in that, After obtaining the user input content and the source type of the input content, the method further includes: In response to the fact that the source type of the input content is already in the input box, and the user selects an adjustment mode on the input interface, at least one adjustment content is obtained based on the input content, each of the adjustment contents is used as the recommended content of the input content, and the process directly proceeds to the step of displaying the at least one recommended content on the input interface; Displaying at least one recommended item on the input interface includes: One of the adjustment options is displayed in the candidate area of ​​the input interface until the user selects the corresponding adjustment option in the candidate area. The selected adjustment option replaces the input content in the input box, or the user is detected to have selected a change option in the input interface. In response to detecting that the user selects a change option on the input interface, all the adjustment content is displayed in the candidate expansion area of ​​the input interface until the user selects any of the adjustment content in the candidate area and replaces the input content in the input box with the selected adjustment content, or the user selects to exit the candidate expansion area on the input interface.

9. The method according to claim 8, characterized in that, Before obtaining at least one adjustment content based on the input content, the method further includes: Obtain users' expression habits; The step of obtaining at least one adjustment content based on the input content includes: Obtain at least one adjusted content based on the input content and the expression habits; wherein at least a portion of the adjusted content matches the style of the expression habits.

10. An input recommendation device, characterized in that, The input recommendation device includes: An input module is used to obtain the input content entered by the user and the source type of the input content; wherein, the source type includes content already in the input box and content already copied from the other party; The acquisition module is used to acquire at least one recommended content based on the input content and the source type; wherein, when the source type is "already in the input box", the recommended content represents the candidate continuation content of the input content, and when the source type is "content sent by the other party has been copied", the recommended content represents the candidate reply content of the input content; The display module is used to display at least one piece of recommended content on the input interface; Before obtaining at least one recommended content based on the input content and the source type, the method further includes: obtaining reference content of the current application where the input interface is located; wherein, when it is detected that the input interface is switched from another application to the current application, and the switching time is within a preset time interval before the input content, if the application type corresponding to the application before the switch is the same as the application type corresponding to the application after the switch, then the reference content includes at least part of the dialogue content in the application before the switch; otherwise, the reference content represents at least part of the dialogue content in the current application. The step of obtaining at least one recommended content based on the input content and the source type includes: obtaining at least one recommended content that semantically matches the reference content, based on the input content, the source type, and the reference content.

11. A mobile terminal, characterized in that, include: The device includes a display screen, a memory, and a processor, wherein the display screen and the memory are respectively coupled to the processor, the display screen is used at least to display content to a user and to allow the user to input or copy content, the memory stores program instructions, and the processor is used to execute the program instructions to implement the input recommendation method as described in any one of claims 1-9.

12. A computer-readable storage medium having program instructions stored thereon, characterized in that, When the program instructions are executed by the processor, they implement the input recommendation method as described in any one of claims 1-9.

Images