Intelligent question and answer method and device, electronic equipment and storage medium

Abstract

The invention provides an intelligent questioning and answering method and device, electronic equipment and a storage medium, and belongs to the field of intelligent questioning and answering. The method comprises the steps that a target questioning and answering mode is determined from multiple questioning and answering modes based on the information amount of a target question, and the numbers of feature types needing to be combined with all the questioning and answering modes are different; determining a target answer text corresponding to the target question based on the target question-answering mode; and adjusting the target answer text based on the broadcast feature, determining a target answer corresponding to the target question, and outputting the target answer. According to the intelligent question-answering method and device, the electronic equipment and the storage medium, the intelligent degree of intelligent question-answering can be improved, and the user experience is improved.

Description

Technical Field

[0001] The present disclosure belongs to the field of intelligent question answering, and more specifically, relates to an intelligent question answering method, device, electronic device and storage medium. Background Art

[0002] With the development of science and technology and the improvement of quality of life, people's demand for intelligent question-answering has become higher and more diverse. The intelligent question-answering system is not restricted by time, holidays and other factors, and can provide help to users at any time. For common and simple questions, intelligent question-answering can quickly give answers, saving users a lot of time and avoiding users from looking for answers in lengthy web pages or customer service queues.

[0003] Nowadays, intelligent question-answering systems use a relatively fixed question-answering model. These systems are usually based on a single algorithm or rule and treat all questions equally. They are unable to give users satisfactory answers and require users to keep asking questions, which makes it difficult to meet user needs. Summary of the invention

[0004] The purpose of the present disclosure is to provide an intelligent question-answering method, device, electronic device and storage medium to improve the intelligence of intelligent question-answering and enhance user experience.

[0005] A first aspect of the embodiments of the present disclosure provides an intelligent question-answering method, including: Determining a target question-answering model from a plurality of question-answering models based on the amount of information of the target question, wherein each of the question-answering models needs to combine a different number of feature types; Determining a target answer text corresponding to the target question based on the target question-answering mode; The target answer text is adjusted based on the broadcast characteristics, the target answer corresponding to the target question is determined, and the target answer is output.

[0006] A second aspect of the embodiments of the present disclosure provides an intelligent question-answering device, including: A mode selection module, used for determining a target question-answering mode from multiple question-answering modes based on the amount of information of the target question, wherein each question-answering mode needs to combine a different number of feature types; A text output module, used to determine a target answer text corresponding to the target question based on the target question-answering mode; The answer module is used to adjust the target answer text based on the broadcast characteristics, determine the target answer corresponding to the target question, and output the target answer.

[0007] According to a third aspect of an embodiment of the present disclosure, an electronic device is provided, comprising a memory, a processor, and a computer program stored in the memory and running on the processor, wherein the processor implements the steps of the above-mentioned intelligent question-answering method when executing the computer program.

[0008] According to a fourth aspect of an embodiment of the present disclosure, a computer-readable storage medium is provided, wherein the computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, the steps of the above-mentioned intelligent question-answering method are implemented.

[0009] The intelligent question-answering method, device, electronic device, and storage medium provided by the embodiments of the present disclosure have the following beneficial effects: The disclosure determines the target question-and-answer mode based on the amount of information of the target question, which can more accurately match the complexity of the question and the number of feature types to be considered, so that the answer is more in line with the actual needs of the question, and improves the accuracy and pertinence of the question-and-answer. Multiple question-and-answer modes cover different types of questions and feature types, and can provide users with more diverse answer options. At the same time, the target answer text is adjusted based on the broadcast feature, so that the answer is more in line with the user's auditory preferences and understanding habits, thereby optimizing the user experience. The disclosure introduces multiple question-and-answer modes and a broadcast feature adjustment mechanism, so that the disclosure can be flexibly adjusted according to different questions and user needs, so that the disclosure can better adapt to different application scenarios and user groups, and improve the practicality and generalization ability of the disclosure. The disclosure selects the most suitable question-and-answer mode based on the amount of information of the target question, which can reduce unnecessary feature processing and calculation, thereby improving the efficiency of question-and-answer. In addition, the adjustment of the broadcast feature also helps to increase the generation and output speed of the answer, so that the user can obtain the required information faster. The intelligence of the intelligent question-and-answer is improved, and the user experience is improved. BRIEF DESCRIPTION OF THE DRAWINGS

[0010] In order to more clearly illustrate the technical solutions in the embodiments of the present disclosure, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present disclosure. For ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work.

[0011] Figure 1 A flowchart of an intelligent question-answering method provided by an embodiment of the present disclosure; Figure 2 A structural block diagram of an intelligent question-answering device provided in one embodiment of the present disclosure; Figure 3 A schematic block diagram of an electronic device provided by an embodiment of the present disclosure. DETAILED DESCRIPTION

[0012] In the following description, specific details such as specific system structures and technologies are provided for the purpose of illustration rather than limitation, so as to provide a thorough understanding of the embodiments of the present disclosure. However, it should be clear to those skilled in the art that the present disclosure may be implemented in other embodiments without these specific details. In other cases, detailed descriptions of well-known systems, devices, circuits, and methods are omitted to avoid obstructing the description of the present disclosure with unnecessary details.

[0013] In order to make the purpose, technical solutions and advantages of the present disclosure more clear, specific embodiments will be described below in conjunction with the accompanying drawings.

[0014] Please refer to Figure 1 , Figure 1 A flowchart of an intelligent question-answering method provided in an embodiment of the present disclosure is provided, and the method includes: S101: Determine a target question-answering model from multiple question-answering models based on the amount of information of the target question, wherein each question-answering model needs to combine a different number of feature types.

[0015] In this embodiment, the target question is a question raised by the user that needs to be answered, such as "How to treat a cold?" The target question can be input into the execution subject by voice.

[0016] The amount of information is an indicator to measure the complexity of the target problem or the amount of content it contains.

[0017] The information content of the target problem can be calculated as follows: Get the information content of the target question based on the text length of the target question; and / or, obtaining the information content of the target question based on the sentence structure complexity evaluation method; and / or, obtaining information about the target question based on a topic model method; And / or, the amount of information of the target problem is obtained based on the information entropy calculation method.

[0018] Multiple question-and-answer modes are different pre-set ways or strategies for answering questions, and each mode considers different factors (feature types) when answering questions.

[0019] The number of feature types refers to the number of different types of factors that each question-answering model combines when answering questions. For example, one question-answering model only considers the question type, while another considers multiple features such as question type, user age, and user emotion.

[0020] For example, there are two question-answering modes: Mode A and Mode B. Mode A only answers questions based on the type of question, while Mode B combines the three features of question type, user's professional background, and current usage scenario. When a user asks the question "What is quantum entanglement?", it is judged that this question has a small amount of information (just a simple conceptual inquiry), and the target question-answering mode is determined to be Mode A; if the user asks "What is the application of quantum entanglement in quantum computing and what impact will it have on future technological development? I am a computer science student and am doing related research", this question has a large amount of information, and the target question-answering mode is determined to be Mode B.

[0021] In this embodiment, before determining the target question-answering mode from multiple question-answering modes based on the information amount of the target question, the method further includes: The information amount of the target question is determined based on the user information and the first information amount of the target question.

[0022] In this embodiment, the amount of information of the target question may also include the amount of information of user information. The user information may include: the user's age, gender, industry, years of experience in the industry, etc. The user information may be obtained through the information provided during user registration.

[0023] In this embodiment, in response to the target question being related to the user information, the information amount of the target question includes the information amount of the user information; In response to the target question being irrelevant to the user information, the information amount of the target question does not include the information amount of the user information.

[0024] Taking into account people's language habits, they usually do not combine too much personal information when asking questions. They usually just give their own questions and needs. Therefore, user information can be taken into account when calculating the amount of information.

[0025] Whether the target question and user information are relevant can be judged based on methods such as keyword matching, knowledge graphs, classification systems, and labels.

[0026] For example, user C's information is that he is 22 years old, male, works in the construction industry, and has been in the industry for 1 year. When user C asks "What's the weather like today?", the target question is irrelevant to the user information, and the information content of the target question does not include the information content of the user information.

[0027] User D's information is that he is 20 years old, male, works in the food industry, and has been in the industry for 5 years. When both user C and user D raise the question "the process of building a house", since user C's industry is the construction industry and user D is a kindergarten student, the amount of information in user C's question can include the amount of information in the user information, while the amount of information in user D's question does not include the amount of information in the user information, that is, for the same question, the amount of information of the target question corresponding to different users is different.

[0028] At this point, the answers to users can be more professional and targeted: "Building a house generally requires the following process: xxx in the preliminary planning; then xxx in the foundation engineering; xxx after the main structure is completed; at the same time xxx is carried out simultaneously, and pipes and lines are reserved; finally xxx. You have just entered the industry, so you should learn more from experienced colleagues in actual operations and pay attention to construction safety and quality standards."

[0029] The answer for user D should be simpler and more visual: "To put it simply, building a house is like playing with a large set of building blocks. First, you need to make a plan, just like you have to think about what kind of house you want to build before you play with the building blocks. This plan includes the size, shape, and function of the house, such as whether it is an ordinary residence for living in, or a shop for opening a store. Then it is time to lay the foundation, which is like when playing with building blocks you have to first stabilize the bottom layer. The foundation of the house must be based on the conditions of the land. If the land is relatively soft, the foundation must be made more solid."

[0030] S102: Determine a target answer text corresponding to the target question based on the target question-answering model.

[0031] In this embodiment, the target answer text is the initial answer content obtained through a retrieval or generation process based on the selected target question-answering model, and is in a text form that has not been adjusted.

[0032] The target answer text can be the information processed by the target question-answering model and input into the intelligent question-answering model. The above-mentioned intelligent question-answering model can be a retrieval-based model, a rule-based model, a deep learning-based model, etc. That is, any model that can output text based on input text or voice can be used as an intelligent question-answering model.

[0033] In this embodiment, different target question-answering modes may output different information.

[0034] For example, user C's information is 22 years old, male, working in the construction industry, and has been in the industry for 1 year; user D's information is 20 years old, male, and working in the food industry. For the same question: the process of building a house, since user C and user D have different user information, the amount of information of their corresponding target questions is also different, so user C's question can be answered in mode B (combined with the type of question, the user's professional background and the current usage scenario), and user D's question can be answered in mode A (combined with the type characteristics of the question). For the same question, user C's question can be searched with keywords such as: construction industry, entry-level technicians, and the process of building a house, while user D's question can be searched with keywords such as: the process of building a house.

[0035] S103: Adjust the target answer text based on the broadcast characteristics, determine the target answer corresponding to the target question, and output the target answer.

[0036] In this embodiment, the broadcast features may include broadcast tone, broadcast speed, and broadcast emotion, etc. These features determine how the answer is presented to the user in the form of voice. The target answer is the final answer content after the broadcast features have been adjusted.

[0037] The target answer text can be adjusted based on the tone, speed and emotion of the announcement.

[0038] For example, if the voice tone in the broadcast feature is a gentle female voice, the speech speed is moderate, and the emotion is friendly. For the target answer text "You can take compound acetaminophen tablets to relieve symptoms of a cold", the converted target answer can be "Dear, if you have a cold, you can take compound acetaminophen tablets to relieve the symptoms~". Then, this target answer is output to the user through speech synthesis technology, or output in text form on a device with a screen display.

[0039] It can be concluded from the above that the present disclosure determines the target question-answering mode based on the amount of information of the target question, which can more accurately match the complexity of the question and the number of feature types to be considered, so that the answer is more in line with the actual needs of the question, and improves the accuracy and pertinence of the question-answering. Multiple question-answering modes cover different types of questions and feature types, and can provide users with more diverse answer options. At the same time, the target answer text is adjusted based on the broadcast feature, so that the answer is more in line with the user's auditory preferences and understanding habits, thereby optimizing the user experience. The present disclosure introduces multiple question-answering modes and broadcast feature adjustment mechanisms, so that the present disclosure can be flexibly adjusted according to different questions and user needs, so that the present disclosure can better adapt to different application scenarios and user groups, and improve the practicality and generalization ability of the present disclosure. The present disclosure selects the most suitable question-answering mode based on the amount of information of the target question, which can reduce unnecessary feature processing and calculation, thereby improving the efficiency of question-answering. In addition, the adjustment of the broadcast feature also helps to increase the generation and output speed of the answer, so that the user can obtain the required information faster. The intelligence of intelligent question-answering is improved, and the user experience is improved.

[0040] In one embodiment of the present disclosure, determining a target question-answering mode from multiple question-answering modes based on the amount of information of the target question includes: In response to the amount of information being greater than or equal to the first amount of information, determining the target question-and-answer mode to be the first question-and-answer mode; In response to the amount of information being less than the first amount of information, determining the target question-answering mode to be the second question-answering mode; Among them, the number of feature types that need to be combined in the first question-and-answer mode is greater than the number of types that need to be combined in the second question-and-answer mode.

[0041] In this embodiment, the first amount of information is a preset threshold value of the amount of information, which is used to measure the amount of information of the target question, so as to determine which question-answering mode to select. For example, it can be determined by analyzing and counting a large number of questions, and if the comprehensive indicators such as the number of question characters, vocabulary richness, and semantic complexity reach a certain level, it is the first amount of information.

[0042] The first question-answering model needs to combine many types of features when generating answers. These features can include the user's age, gender, hobbies, emotional state, type of question, semantic content of the question, the context of the question (time, place, etc.), and other factors. This model is suitable for complex questions with a large amount of information, and its purpose is to answer questions more accurately and comprehensively.

[0043] Compared with the first question-answering mode, the second question-answering mode is a relatively simple strategy for answering questions. It combines fewer types of features when answering questions. It only focuses on the type of question itself and simple semantic content, and less on other information about the user. This mode is used for simple questions with less information and can give answers quickly.

[0044] The amount of information can be calculated based on one or more of the following methods: a calculation method based on text length, a calculation method based on vocabulary, a calculation method based on information entropy and a probability model, and the like.

[0045] For example, after studying a large number of tourism-related questions, if a question contains more than 30 characters, involves at least 3 different tourism elements (such as attractions, transportation, accommodation, etc.), and has a certain semantic complexity (such as asking about the impact of different seasons on the experience of a certain travel route), it is considered to have reached the first information amount.

[0046] When a user asks, "I plan to take my parents to Beijing for a week-long trip in the fall. I hope the itinerary is not too tight. I want them to visit famous attractions such as the Forbidden City and the Summer Palace, experience Beijing's hutong culture, and taste local specialties. At the same time, I have to consider my parents' physical condition and the accommodation should be convenient for travel. Do you have any good travel tips?" This question has a large amount of information (greater than the first amount of information), which meets the condition that the amount of information is greater than or equal to the first amount of information. In this case, the first question-answering mode is used, and multiple features are combined when answering this question.

[0047] In this process, based on the user's information that the user is 30 years old and male, we can recommend some attractions and activities with in-depth cultural experiences based on the user's age of 30 and his interest in history and culture; taking into account the fact that he is traveling with his parents and their physical condition, we choose routes with convenient transportation and less physical exertion when arranging the itinerary; based on the seasonal characteristics of autumn, we recommend Beijing's special foods suitable for tasting in autumn; and considering the requirement of accommodation that should be convenient for travel, we recommend suitable hotels near the attractions.

[0048] When a user asks "What are the opening hours of the Forbidden City in Beijing?", the information content of this question is small (less than the first information content), only involves one tourism element (attraction) and the content is simple. At this time, the target question-answering mode is determined to be the second question-answering mode. When answering, only the opening hours of the Forbidden City need to be retrieved and output, and other characteristics such as the user's age and hobbies do not need to be considered.

[0049] It can be concluded from the above that the present disclosure selects different question-and-answer modes according to the amount of information of the target question, and the method can more accurately match the complexity of the question and the type of features to be considered. For questions with a large amount of information, the first question-and-answer mode can combine multiple features to provide more accurate and comprehensive answers; for questions with a small amount of information, the second question-and-answer mode can quickly and concisely give answers. This way of dynamically selecting the question-and-answer mode significantly improves the accuracy and pertinence of questions and answers. The present disclosure can provide personalized answers based on different question types and user information. The present disclosure pre-sets the information amount threshold and selects the question-and-answer mode according to the amount of information, which can reduce unnecessary feature processing and calculation, thereby improving the efficiency of question-and-answer. For questions with a small amount of information, the second question-and-answer mode can quickly retrieve and output answers, reducing the processing time of the present disclosure and increasing the speed at which users obtain information. The intelligence level of intelligent question-and-answer is improved, and the user experience is improved.

[0050] In one embodiment of the present disclosure, the target question-answering mode is a first question-answering mode; Determine the target answer text corresponding to the target question based on the target question-answering model, including: The target question is retrieved based on user features and question features to determine the target answer text. The user features are the features corresponding to the user who asked the target question, and the question features are the features of the target question.

[0051] In one embodiment of the present disclosure, the target question-answering mode is the second question-answering mode; Determine the target answer text corresponding to the target question based on the target question-answering model, including: The target question is retrieved based on the question features to determine the target answer text, where the question features are the features of the target question.

[0052] In this embodiment, the first question-and-answer mode is suitable for target questions with a large amount of information. In this mode, multiple factors need to be considered when answering questions. The second question-and-answer mode is used for target questions with a small amount of information. In the second question-and-answer mode, the answer is mainly based on the characteristics of the question itself, without considering other additional information of the user.

[0053] User characteristics refer to the relevant attributes of the user who asks the question, including but not limited to age, gender, occupation, interests and hobbies, historical question records, current emotional state, type of device used, geographical location, usage time and other information, which can help the present disclosure better understand user needs and background, and thus give answers that are more suitable for the user.

[0054] Question features are the characteristics presented by the target question itself, such as the type of question (whether it is asking for facts, seeking advice, or requiring an explanation of principles, etc.), the subject area of ​​the question (such as technology, medical care, entertainment, etc.), the complexity of the question (measured by vocabulary, sentence structure, number of concepts involved, etc.), keywords in the question, and the urgency of the question (which can be judged by specific vocabulary or punctuation marks, such as "urgent" and "immediately".

[0055] For example, a user asked, "I felt my heart beating a little faster and I felt a little dizzy after exercising recently. Can I continue exercising if I have high blood pressure? If so, what should I pay attention to?"

[0056] This question belongs to the health consultation type. The subject area is the intersection of medical care and fitness. The question is quite complex. The key words include "high blood pressure", "exercise", "fast heartbeat" and "dizziness". It has a certain urgency (involving physical discomfort).

[0057] This embodiment will combine user features and question features for retrieval. Because the user has a history of hypertension and feels uncomfortable after exercise, and is also a fitness enthusiast, this disclosure may retrieve the following target answer text: "Given that you have a history of hypertension and experience symptoms of increased heart rate and dizziness after exercise, you need to be cautious about exercise. For people with hypertension, moderate exercise is beneficial, but you are currently feeling unwell, so it is recommended that you stop exercising and observe. If symptoms persist or worsen, please seek medical attention immediately. If you want to continue fitness in the future, you can choose some low-intensity, rhythmic aerobic exercises, such as walking, Tai Chi, etc. Be sure to warm up before exercise, and pay attention to keeping your breathing even during exercise, avoid excessive fatigue, and pay close attention to your body's reactions. In addition, measure your blood pressure regularly to ensure that it is within a stable range."

[0058] Or, for example, a user asks "What is the normal blood pressure range for patients with hypertension?" This is a simple factual question with a low complexity and the keywords "patients with hypertension" and "normal blood pressure range". The search is based only on the characteristics of the question, and the target answer text is obtained: "In general, it is ideal for patients with hypertension to control their blood pressure within the range of systolic pressure 90-139 mmHg and diastolic pressure 60-89 mmHg, but the specific range may vary due to individual differences and doctor's advice." In this process, other user characteristics such as the user's age, occupation, hobbies, etc. are not considered, because the question has a small amount of information and can be answered based on the content of the question alone.

[0059] From the above, it can be concluded that, in the first question-answering mode, the present disclosure comprehensively considers user characteristics and question characteristics to determine the target answer text, which can more accurately understand the user's specific needs and background, thereby providing a more personalized and user-friendly answer. In the second question-answering mode, the present disclosure mainly searches based on question characteristics, and does not need to consider other additional information of the user. When processing questions with a small amount of information and a simple structure, it can significantly improve the efficiency of question-answering and reduce unnecessary calculations and resource consumption. By introducing two question-answering modes, the present disclosure can flexibly select according to the complexity of the question and the amount of information, so that the present disclosure can better adapt to different application scenarios and user groups, improve the practicality and generalization ability of the present disclosure, improve the intelligence level of intelligent question-answering, and enhance the user experience.

[0060] In one embodiment of the present disclosure, before adjusting the target answer text based on the broadcast feature and determining the target answer corresponding to the target question, the method further includes: Determine the first broadcast feature based on the user's age, gender, speech speed and emotion; Determining a second broadcast feature based on the target answer text; Obtaining a broadcast feature based on the first broadcast feature and the second broadcast feature; Among them, the feature type of the first broadcast feature is greater than the feature type of the second broadcast feature.

[0061] In this embodiment, the first announcement feature is an announcement-related characteristic determined based on the user's own attributes, and may include announcement timbre, announcement speech speed, and announcement emotion.

[0062] The second broadcast feature is a broadcast-related feature determined based on the content characteristics of the target answer text itself, including the second broadcast speed and the second broadcast emotion. This is to adjust the broadcast method according to the nature of the answer content so that the broadcast and the content are better matched, and can include the broadcast speed and broadcast emotion.

[0063] For example, if the answer is urgent information (such as "Evacuate immediately in case of fire"), the speed will be faster; if it is a complex concept explanation, the speed will be slowed down appropriately. If the answer is comforting ("Don't worry, everything will be fine"), the emotion will be warm and soothing; if it is exciting news ("You won the lottery!"), the emotion will be excited and passionate.

[0064] The announcement tone can be determined according to the user's age and gender. Users of different ages and genders have different preferences or more acceptable types of tone. For example, young female users prefer crisp and sweet tones; older male users are more suitable for calm and gentle tones.

[0065] The broadcast speed can be determined according to the user's speaking speed. If the user speaks quickly, he or she will be used to or more adapted to a slightly faster broadcast speed; conversely, if the user speaks slowly, the corresponding broadcast speed will also be set slower, which can make the interaction feel more natural and coordinated.

[0066] The broadcast emotion can be determined by the user's mood. For example, if the user is detected to be in a positive mood, the broadcast emotion will also present a positive style, such as enthusiasm and cheerfulness; if the user is in a low mood, the broadcast emotion will be adjusted to a more soothing and gentle style.

[0067] In this embodiment, when the broadcasting speech speed in the first broadcasting feature and the broadcasting speech speed in the second broadcasting feature are different, weights can be assigned to the broadcasting speech speed in the first broadcasting feature and the broadcasting speech speed in the second broadcasting feature respectively according to the type of question, and the weighted average value is calculated as the final broadcasting speech speed. For example, for ordinary information questions, the broadcasting speech speed weight can be set to 0.6 and the second broadcasting speech speed weight can be set to 0.4; for emergency questions, the first broadcasting speech speed weight can be set to 0.2 and the second broadcasting speech speed weight can be set to 0.8.

[0068] Alternatively, if the user believes that the personalization and stability of the experience are more important, then the user-based first broadcast feature can be set to a higher priority. Because the first broadcast feature is determined based on relatively fixed personal attributes such as the user's age, gender, speech speed, and emotions, giving priority to these features can maintain a consistent broadcast style for the user.

[0069] In this embodiment, when the announcement emotion in the first announcement feature and the announcement emotion in the second announcement feature are different, the announcement emotion in the first announcement feature and the announcement emotion in the second announcement feature can be merged based on the emotion vector mixing method.

[0070] The user's speaking speed can be determined by recording the duration of the user's voice when asking a question and the number of words (or syllables) in the question text, and calculating the number of words (or syllables) per unit time.

[0071] The user's emotions can be judged by building an emotional keyword dictionary that contains positive, negative and neutral words, and by counting the number of each type of words in the user's question text.

[0072] The intonation of speech can also be analyzed. For example, high pitch, large pitch variation and rising intonation indicate positive or excited emotions; low pitch, small pitch variation and falling intonation indicate negative emotions; and steady intonation indicates neutral emotions.

[0073] For example, user E is a 30-year-old female who speaks fast and is in a positive mood. Her question is: travel recommendations. At this time, a young, crisp and sweet voice can be selected for her, similar to the voice of a young female radio host. Because the user speaks fast, the broadcast speed in the first broadcast feature is set to a faster level, such as about 200 words per minute. Because the user is in a positive mood, the broadcast emotion in the first broadcast feature is set to a warm and cheerful style.

[0074] The target answer text may be "You can go to Bali! There is xxxx there."

[0075] The text is a light-hearted travel recommendation content, and the speech speed in the second broadcast feature is set to be moderately fast, such as about 180 words per minute, to reflect the feeling of excitement and anticipation.

[0076] The broadcast emotion in the second broadcast feature can be an excited and joyful style, which matches the positive atmosphere of the travel recommendation.

[0077] The final broadcast tone can be young, crisp and sweet. Taking into account the first broadcast speed and the second broadcast speed, and considering the user habits and content nature, the final speed is determined to be about 190 words per minute. Combining the broadcast emotions in the enthusiastic and cheerful first broadcast feature and the broadcast emotions in the excited and joyful second broadcast feature, the broadcast is carried out in a positive and energetic mood, vividly conveying the travel recommendation content to user E.

[0078] From the above, it can be concluded that the present disclosure determines the first broadcast feature based on the user's personal attributes such as age, gender, speech speed and emotion, and the method can provide users with more personalized broadcast services. The present disclosure combines the broadcast emotion in the first broadcast feature with the broadcast emotion in the second broadcast feature, and the method can convey the broadcast content in a more natural and smooth way, so that users can more easily accept and understand the information. The present disclosure establishes an emotional keyword dictionary and analyzes the changes in the intonation of the voice, and the method can accurately judge the emotional state of the user, providing a strong basis for adjusting the broadcast emotion. These measures have improved the efficiency and accuracy of the broadcast.

[0079] In one embodiment of the present disclosure, the intelligent question answering method further includes: In response to the user's speaking speed being greater than or equal to a first speaking speed threshold, speeding up the broadcasting speaking speed at a first speaking speed step length; In response to the user's speaking speed being less than the first speaking speed threshold, the announcement speaking speed is reduced by a second speaking speed step.

[0080] In this embodiment, the intelligent question answering method further includes: In response to the number of questions asked by the user within the first time period being greater than or equal to a first question number threshold and the user's speech speed being greater than or equal to a first speech speed threshold, increasing the first speech speed threshold by a first speech speed adjustment step; In response to the number of questions asked by the user within the first time period being greater than the first number of questions and the user's speaking speed being less than the first speaking speed threshold, the first speaking speed threshold is lowered with a second speaking speed adjustment step.

[0081] In this embodiment, the first speech rate threshold is a pre-set speed standard used to measure the speed of the user's speech. By comparing the size relationship between the user's speech rate and the threshold, it is determined whether to adjust the broadcast speech rate and how to adjust it.

[0082] The first speech speed step is an adjustment step, which is used to speed up the adjustment range of the broadcast speech speed when the user speech speed is greater than or equal to the first speech speed threshold. The second speech speed step is an adjustment step, which is used to reduce the adjustment range of the broadcast speech speed when the user speech speed is less than the first speech speed threshold.

[0083] The above-mentioned first speaking speed threshold, first speaking speed step, second speaking speed step, first duration, first question number threshold, first speaking speed adjustment step and second speaking speed adjustment step can all be considered to be determined.

[0084] For example, the first speech rate threshold is set to 160 words per minute, the first speech rate step is increased by 8 words per minute, and the second speech rate step is decreased by 6 words per minute.

[0085] If the user's speaking speed is 180 words per minute (greater than the first speaking speed threshold), the broadcasting speed is increased by the first speaking speed step. For example, the original broadcasting speed is 120 words per minute, and after adjustment, the broadcasting speed becomes 128 words per minute.

[0086] If the user's speaking speed is 140 words per minute (less than the first speaking speed threshold), the broadcasting speed is reduced by the second speaking speed step. For example, the original broadcasting speed is 150 words per minute, and after adjustment, the broadcasting speed becomes 144 words per minute.

[0087] For another example, the first time period is set to 7 days, the first question number threshold is set to 20 times, the first speech speed adjustment step is set to 10 words per minute, and the second speech speed adjustment step is set to 8 words per minute.

[0088] If the user asks 25 questions within 7 days (greater than the first question number threshold), and the speaking speed of each question is greater than the first speaking speed threshold (assuming the first speaking speed threshold is 150 words per minute), the first speaking speed threshold is increased by the first speaking speed adjustment step. For example, the original first speaking speed threshold is 150 words per minute, and after adjustment it becomes 160 words per minute.

[0089] If the user asks questions 26 times within 7 days (less than the first question number threshold), and the speaking speed of each question is less than the first speaking speed threshold, the first speaking speed threshold is lowered by the second speaking speed adjustment step. For example, the original first speaking speed threshold is 150 words per minute, and after adjustment it becomes 142 words per minute.

[0090] From the above, it can be concluded that the present disclosure dynamically adjusts the broadcast speed according to the user's actual speaking speed so that the broadcast speed matches the user's auditory habits, which can improve the user's auditory experience and reduce the difficulty in understanding or discomfort caused by speech speed mismatch. The present disclosure records the number of questions and speech speed asked by the user within a certain period of time. The method can automatically adjust the speech speed threshold to make it more in line with the user's long-term speech speed habits. This adaptive mechanism helps to maintain the consistency between the broadcast speed and the user's speech speed, further improving the user experience. The method realizes the dynamic adjustment of the broadcast speed by introducing parameters such as the speech speed threshold and the adjustment step size. This dynamic adjustment capability enables the present disclosure to perform adaptive optimization according to the actual needs and habits of the user, thereby improving the intelligence level of the present disclosure.

[0091] In one embodiment of the present disclosure, the first broadcast feature includes: broadcast timbre, first broadcast speech speed and first broadcast emotion; The second broadcast characteristics include: the second broadcast speech speed and the second broadcast emotion; The first broadcast feature is determined based on the user's age, gender and speech speed, including: Determine the announcement tone based on the user's age and gender; Determine the first broadcast speech rate based on the user's speech rate; Determining a first broadcast emotion based on the user's emotion; The second broadcast feature is determined based on the target answer text, including: Determining a second broadcast speech rate based on the target answer text; A second broadcast emotion is determined based on the target answer text.

[0092] In this embodiment, the first broadcast feature can be a feature generated based on the user's own attributes for determining the answer broadcast method. It takes into account factors such as the user's age, gender, and speech speed, including the broadcast timbre, the first broadcast speech speed, and the first broadcast emotion, so that the broadcast is more in line with the user's individual characteristics and provides a personalized interactive experience. The second broadcast feature is a feature generated based on the content characteristics of the target answer text for determining the broadcast method, including the second broadcast speech speed and the second broadcast emotion. The purpose is to adjust the broadcast method according to the nature of the answer content so that the broadcast matches the expression effect of the content.

[0093] The broadcast timbre refers to the characteristics or type of the sound during voice broadcast. The first broadcast speed is the initial broadcast speed determined according to the user's own speaking speed. If the user speaks fast, the present disclosure will set a faster first broadcast speed accordingly; conversely, if the user speaks slowly, the first broadcast speed will also be slower. The first broadcast emotion is the broadcast emotion tendency determined based on the user's current emotional state. If the user is in a positive mood, the first broadcast emotion will also be a positive style, such as enthusiasm and cheerfulness; if the user is in a low mood, the first broadcast emotion will be a comforting and gentle style, so that the broadcast is more in line with the user's psychological state.

[0094] The second broadcast speed is the speed determined by the target answer text. For example, for urgent content (such as fire escape guide), the second broadcast speed will be faster; for detailed explanations of complex concepts, the speed will be slowed down to ensure accurate communication and understanding of the information. The second broadcast emotion is the broadcast emotion determined by the emotional tendency and content nature of the target answer text. For example, if the answer is a congratulatory content (such as "Congratulations on winning the prize!"), the second broadcast emotion is excitement and joy; if it is content that informs unfortunate news, the emotion will be serious and sympathetic.

[0095] For example, user F is a 35-year-old male who speaks quickly and in a positive mood. He asks, "I am going to attend an important business dinner. Are there any etiquettes I need to pay attention to?" Based on the user's identity as a 35-year-old male, a steady and magnetic voice can be selected, similar to the voice of a professional business person, giving people a sense of reliability.

[0096] Since the user speaks fast, the first broadcast speed can be set to about 180 words per minute to match the user's rhythm.

[0097] Because the user emotions were positive, the first broadcast emotion was determined to be a positive and professional style that conveyed enthusiasm and confidence.

[0098] The target answer text is about business dinner etiquette. The content is relatively important but does not need to be conveyed particularly quickly. Considering the seriousness of the business scene, the second broadcast speed is set at about 160 words per minute, slightly slower than the first broadcast speed, to reflect solemnity.

[0099] The answer content is to provide advice and guidance, which belongs to the neutral to positive emotional tendency. The second broadcast emotion is professional, gentle and positive. Taking into account the first and second broadcast characteristics, the final broadcast tone remains calm and attractive, the broadcast speed is about 170 words per minute (a combination of the first and second broadcast speeds), and the broadcast emotion is positive, professional, and gentle, and the above answer content is broadcast to the user. This not only takes into account the individual characteristics of the user, but also makes appropriate broadcasts according to the nature of the answer content.

[0100] For example, "When attending a business dinner, you need to pay attention to the following etiquette: First, the dress should be formal and appropriate. Men generally choose dark suits with ties and leather shoes. Arriving at the dinner on time is basic courtesy. If you are late, you must sincerely apologize to the host. During the dinner, pay attention to table manners and use tableware in a standardized manner... These etiquette can help you show a good image at the dinner. I wish you a successful dinner!"

[0101] In one embodiment of the present disclosure, the broadcast features include: broadcast timbre, broadcast speech speed, and broadcast emotion; The target answer text is adjusted based on the broadcast characteristics to determine the target answer corresponding to the target question, including: The target answer text is adjusted based on the broadcast tone to obtain a first answer text; The target answer text is adjusted based on the broadcast speech speed to obtain a second answer text; The target answer text is adjusted based on the broadcast emotion to obtain a third answer text; A target answer to the target question is determined based on the first answer text, the second answer text, and the third answer text.

[0102] In this embodiment, the first answer text is a result obtained by adjusting the target answer text based on the announcement tone.

[0103] For example, the target answer text is "The food in this restaurant tastes great and the prices are reasonable." If the selected broadcast voice is a friendly and gentle female voice, the adjusted first answer text can be "Dear, the food in this restaurant tastes great and the prices are reasonable."

[0104] The second answer text is the result of adjusting the target answer text based on the broadcast speech speed.

[0105] For example, if the target answer text is "Exercise has many benefits for the body, it can enhance cardiopulmonary function, improve immunity, and improve sleep quality", if the determined broadcast speed is fast, the adjusted second answer text can be "Exercise has many benefits, it can enhance cardiopulmonary function, improve immunity, and improve sleep". By simplifying the sentence structure and simplifying the parallel content into verb-object phrases, the number of words is reduced, which is more suitable for fast broadcasting.

[0106] The third answer text is the result of adjusting the target answer text based on the broadcast emotion.

[0107] For example, if the target answer text is "You didn't get the reward this time", if you need to convey sympathy, the adjusted third answer text can be "It's a pity that you didn't get the reward this time, but don't be discouraged, there will be opportunities in the future." If you want to convey a relaxed emotion, you can adjust it to "Hey, you missed the reward this time, but it doesn't matter, maybe you will get better luck next time!".

[0108] The broadcast tone, broadcast speed and broadcast emotion have been determined.

[0109] In this embodiment, the three texts obtained after the adjustment of timbre, speaking speed and emotion can be comprehensively considered and integrated to form the final target answer.

[0110] For example, user G asks: "Are there any fun places you can recommend for travel?" User G is a 23-year-old female.

[0111] The voice of the broadcast can be that of an enthusiastic and energetic young woman, the speech speed can be about 180 words per minute, and the emotion can be positive and cheerful. The target answer text can be "You can consider going to Sanya, where there are beautiful beaches and warm sunshine, as well as many water activities; you can also go to Chengdu, where there are cute giant pandas and a variety of delicious food."

[0112] Based on the adjustment of the broadcast tone, the first answer text is obtained, and the first answer text can be "Dear, you can go to Sanya! There are super beautiful beaches and super warm sunshine, and there are a lot of water activities! And Chengdu is also great, with super cute giant pandas, and there are so many delicious foods!".

[0113] Based on the adjustment of the broadcast speed, the second answer text is obtained, and the second answer text can be "Dear, you must go to Sanya! There are super beautiful beaches, super warm sunshine and so many water activities. Chengdu is also great, with cute giant pandas and so many delicious foods."

[0114] Based on the adjustment of the broadcast emotion, the third answer text is obtained, and the third answer text can be "Wow! Honey, you must go to Sanya! There are super beautiful beaches, super warm sunshine, and super fun water activities! Chengdu is also awesome! There are super cute giant pandas, and there are so many delicious foods that you can't stop! It's amazing!".

[0115] Based on the first answer text, the second answer text and the third answer text, the target answer is determined as "Wow! Honey, you must go to Sanya! There are super beautiful beaches, super warm sunshine, and super fun water activities! Chengdu is also awesome! There are super cute giant pandas, and there are so many delicious foods that you can't stop! It's amazing!".

[0116] In this embodiment, element-by-element analysis and fusion can be performed. At the vocabulary level, the merging principle can be used to compare the vocabulary in the three texts, and those vocabulary that are important in different adjustment dimensions can be retained. If a certain vocabulary appears repeatedly in multiple texts and is indispensable for expressing the core content or emotion, then this vocabulary should be retained in the target answer first.

[0117] In terms of sentence structure, we can analyze the sentence structures of the three texts based on the principle of coordination and select the most clear, concise and effective structure as the basic structure of the target answer. If the first answer text uses complex long sentences to create a friendly atmosphere, the second answer text is simplified to short sentences to adapt to the speaking speed, and the third answer text enhances the expressiveness of the sentences through rich modifiers, then we need to find a balance between them.

[0118] In one embodiment of the present disclosure, the relative importance of the three dimensions of timbre, speaking speed and emotion can also be determined based on the core objectives of the broadcast, user needs and question types, and weights can be assigned to the first answer text, the second answer text and the third answer text.

[0119] For example, in an emergency notification broadcast scenario, speaking speed is the most important factor, so the weight of the second answer text will be higher; in an emotional comfort scenario, the emotion-related third answer text will have a larger weight.

[0120] From the above, it can be concluded that the present disclosure can provide users with an auditory experience that is more in line with their personal characteristics by selecting the announcement tone according to the user's age and gender, which helps to improve the user's overall experience. The present disclosure can match the announcement method with the content expression effect by adjusting the announcement speed and emotion according to the content characteristics of the target answer text. The present disclosure significantly improves the user experience and content adaptability by introducing personalized announcement features and dynamic adjustment mechanisms, improves the intelligence level of intelligent question and answer, and improves the user experience.

[0121] Corresponding to the intelligent question-answering method in the above embodiment, Figure 2 This is a block diagram of the structure of an intelligent question-answering device provided by an embodiment of the present disclosure. For ease of explanation, only the parts related to the embodiment of the present disclosure are shown. Figure 2 The intelligent question-answering device 20 includes: a mode selection module 21, a text output module 22 and an answering module 23.

[0122] The mode selection module 21 is used to determine a target question-answering mode from multiple question-answering modes based on the amount of information of the target question, wherein each question-answering mode needs to combine a different number of feature types; A text output module 22, used to determine a target answer text corresponding to a target question based on a target question-answering mode; The answer module 23 is used to adjust the target answer text based on the broadcast characteristics, determine the target answer corresponding to the target question, and output the target answer.

[0123] In one embodiment of the present disclosure, the mode selection module 21 is specifically configured to determine that the target question-and-answer mode is the first question-and-answer mode in response to the amount of information being greater than or equal to the first amount of information; In response to the amount of information being less than the first amount of information, determining the target question-answering mode to be the second question-answering mode; Among them, the number of feature types that need to be combined in the first question-and-answer mode is greater than the number of types that need to be combined in the second question-and-answer mode.

[0124] In one embodiment of the present disclosure, the target question-answering mode is a first question-answering mode; The text output module 22 is specifically used to search for the target question based on user characteristics and question characteristics to determine the target answer text. The user characteristics are the characteristics corresponding to the user who asks the target question, and the question characteristics are the characteristics of the target question.

[0125] In one embodiment of the present disclosure, the target question-answering mode is the second question-answering mode; The text output module 22 is further specifically used to search for a target question based on question features and determine a target answer text, wherein the question features are features of the target question.

[0126] In one embodiment of the present disclosure, the intelligent question-answering device 20 further includes: a feature determination module; A feature determination module, used to determine the first broadcast feature based on the user's age, gender, speech speed and emotion; Determining a second broadcast feature based on the target answer text; Obtaining a broadcast feature based on the first broadcast feature and the second broadcast feature; Among them, the feature type of the first broadcast feature is greater than the feature type of the second broadcast feature.

[0127] In one embodiment of the present disclosure, the first broadcast feature includes: broadcast timbre, first broadcast speech speed and first broadcast emotion; The second broadcast characteristics include: the second broadcast speech speed and the second broadcast emotion; A feature determination module, specifically used to determine the announcement timbre based on the user's age and gender; Determine the first broadcast speech rate based on the user's speech rate; Determining a first broadcast emotion based on the user's emotion; Determining a second broadcast speech rate based on the target answer text; A second broadcast emotion is determined based on the target answer text.

[0128] In one embodiment of the present disclosure, the broadcast features include: broadcast timbre, broadcast speech speed, and broadcast emotion; The answer module 23 is specifically used to adjust the target answer text based on the broadcast tone to obtain a first answer text; The target answer text is adjusted based on the broadcast speech speed to obtain a second answer text; The target answer text is adjusted based on the broadcast emotion to obtain a third answer text; A target answer to the target question is determined based on the first answer text, the second answer text, and the third answer text.

[0129] See also Figure 3 , Figure 3 A schematic block diagram of an electronic device provided by an embodiment of the present disclosure. Figure 3 The electronic device 300 in the embodiment shown may include: one or more processors 301, one or more input devices 302, one or more output devices 303 and one or more memories 304. The processors 301, input devices 302, output devices 303 and memories 304 communicate with each other via a communication bus 305. The memory 304 is used to store computer programs, which include program instructions. The processor 301 is used to execute the program instructions stored in the memory 304. The processor 301 is configured to call the program instructions to execute the functions of each module / unit in the above-mentioned device embodiments, such as Figure 2 The functions of modules 21 to 23 are shown.

[0130] It should be understood that in the embodiment of the present disclosure, the processor 301 may be a central processing unit (CPU), and the processor may also be other general-purpose processors, digital signal processors (DSP), application-specific integrated circuits (ASIC), field-programmable gate arrays (FPGA) or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. A general-purpose processor may be a microprocessor or the processor may also be any conventional processor, etc.

[0131] The input device 302 may include a touch panel, a fingerprint collection sensor (for collecting the user's fingerprint information and fingerprint direction information), a microphone, etc., and the output device 303 may include a display (LCD, etc.), a speaker, etc.

[0132] The memory 304 may include a read-only memory and a random access memory, and provide instructions and data to the processor 301. A portion of the memory 304 may also include a non-volatile random access memory. For example, the memory 304 may also store information about the device type.

[0133] In a specific implementation, the processor 301, input device 302, and output device 303 described in the embodiments of the present disclosure can execute the implementation methods described in the first and second embodiments of the intelligent question and answer method provided in the embodiments of the present disclosure, and can also execute the implementation methods of the electronic device described in the embodiments of the present disclosure, which will not be repeated here.

[0134] In another embodiment of the present disclosure, a computer-readable storage medium is provided, wherein the computer-readable storage medium stores a computer program, wherein the computer program includes program instructions, and when the program instructions are executed by the processor, all or part of the processes in the above-mentioned embodiment method are implemented, and the computer program can also be completed by instructing the relevant hardware through the computer program. The computer program can be stored in a computer-readable storage medium, and when the computer program is executed by the processor, the steps of each of the above-mentioned method embodiments can be implemented. Among them, the computer program includes computer program code, and the computer program code can be in source code form, object code form, executable file or some intermediate form, etc. The computer-readable medium can include: any entity or device capable of carrying computer program code, recording medium, U disk, mobile hard disk, magnetic disk, optical disk, computer memory, read-only memory (ROM, Read-Only Memory), random access memory (RAM, Random Access Memory), electric carrier signal, telecommunication signal and software distribution medium, etc.

[0135] The computer-readable storage medium may be an internal storage unit of the electronic device of any of the aforementioned embodiments, such as a hard disk or memory of the electronic device. The computer-readable storage medium may also be an external storage device of the electronic device, such as a plug-in hard disk, a smart memory card (Smart Media Card, SMC), a secure digital (SecureDigital, SD) card, a flash card (Flash Card), etc. equipped on the electronic device. Furthermore, the computer-readable storage medium may also include both an internal storage unit of the electronic device and an external storage device. The computer-readable storage medium is used to store computer programs and other programs and data required by the electronic device. The computer-readable storage medium may also be used to temporarily store data that has been output or is to be output.

[0136] Those of ordinary skill in the art will appreciate that the units and algorithm steps of each example described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, computer software, or a combination of the two. In order to clearly illustrate the interchangeability of hardware and software, the composition and steps of each example have been generally described in terms of function in the above description. Whether these functions are performed in hardware or software depends on the specific application and design constraints of the technical solution. Professional and technical personnel can use different methods to implement the described functions for each specific application, but such implementation should not be considered to be beyond the scope of this disclosure.

[0137] Those skilled in the art can clearly understand that, for the convenience and brevity of description, the specific working processes of the electronic devices and units described above can refer to the corresponding processes in the aforementioned method embodiments and will not be repeated here.

[0138] In the several embodiments provided in the present application, it should be understood that the disclosed electronic devices and methods can be implemented in other ways. For example, the device embodiments described above are only schematic, for example, the division of units is only a logical function division, and there may be other division methods in actual implementation, such as multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. In addition, the mutual coupling or direct coupling or communication connection shown or discussed can be an indirect coupling or communication connection through some interfaces or units, or it can be an electrical, mechanical or other form of connection.

[0139] The units described as separate components may or may not be physically separated, and the components shown as units may or may not be physical units, that is, they may be located in one place or distributed on multiple network units. Some or all of the units may be selected according to actual needs to achieve the purpose of the embodiments of the present disclosure.

[0140] In addition, each functional unit in each embodiment of the present disclosure may be integrated into one processing unit, or each unit may exist physically separately, or two or more units may be integrated into one unit. The above-mentioned integrated unit may be implemented in the form of hardware or in the form of software functional units.

[0141] The above are only specific embodiments of the present disclosure, but the protection scope of the present disclosure is not limited thereto. Any technician familiar with the technical field can easily think of various equivalent modifications or replacements within the technical scope disclosed in the present disclosure, and these modifications or replacements should be included in the protection scope of the present disclosure. Therefore, the protection scope of the present disclosure should be based on the protection scope of the claims.

Claims

1. An intelligent question-answering method, characterized in that: include: Determining a target question-answering model from a plurality of question-answering models based on the amount of information of the target question, wherein each of the question-answering models needs to combine a different number of feature types; Determining a target answer text corresponding to the target question based on the target question-answering mode; The target answer text is adjusted based on the broadcast characteristics, the target answer corresponding to the target question is determined, and the target answer is output.

2. The intelligent question-answering method according to claim 1, characterized in that: The step of determining a target question-answering mode from a plurality of question-answering modes based on the amount of information of the target question includes: In response to the amount of information being greater than or equal to the first amount of information, determining the target question-and-answer mode to be the first question-and-answer mode; In response to the amount of information being less than the first amount of information, determining the target question-answering mode to be a second question-answering mode; Among them, the number of feature types that the first question and answer mode needs to combine is greater than the number of types that the second question and answer mode needs to combine.

3. The intelligent question-answering method according to claim 1, characterized in that: The target question-answering mode is a first question-answering mode; The determining a target answer text corresponding to the target question based on the target question-answering mode includes: The target question is retrieved based on user features and question features to determine a target answer text, wherein the user features are features corresponding to the user who asks the target question, and the question features are features of the target question.

4. The intelligent question-answering method according to claim 1, wherein: The target question-answering mode is the second question-answering mode; The determining a target answer text corresponding to the target question based on the target question-answering mode includes: The target question is retrieved based on question features to determine a target answer text, wherein the question features are features of the target question.

5. The intelligent question-answering method according to claim 1, wherein: Before adjusting the target answer text based on the broadcast feature to determine the target answer corresponding to the target question, the method further includes: Determine the first broadcast feature based on the user's age, gender, speech speed and emotion; Determining a second broadcast feature based on the target answer text; Obtaining the broadcast feature based on the first broadcast feature and the second broadcast feature; The feature type of the first broadcast feature is greater than the feature type of the second broadcast feature.

6. The intelligent question-answering method according to claim 5, characterized in that: The first broadcast feature includes: broadcast tone, first broadcast speech speed and first broadcast emotion; The second broadcast feature includes: a second broadcast speech speed and a second broadcast emotion; The determining of the first broadcast feature based on the user's age, gender and speech speed includes: Determining the announcement tone based on the age and gender of the user; Determining the first broadcast speech rate based on the speech rate of the user; Determining the first broadcast emotion based on the emotion of the user; The determining of the second broadcast feature based on the target answer text includes: Determining the second broadcast speech rate based on the target answer text; The second broadcast emotion is determined based on the target answer text.

7. The intelligent question-answering method according to claim 1, characterized in that: The broadcast features include: broadcast tone, broadcast speed and broadcast emotion; The step of adjusting the target answer text based on the broadcast feature to determine the target answer corresponding to the target question includes: Adjusting the target answer text based on the broadcast tone to obtain a first answer text; Adjusting the target answer text based on the broadcast speech speed to obtain a second answer text; Adjusting the target answer text based on the broadcast emotion to obtain a third answer text; A target answer to the target question is determined based on the first answer text, the second answer text, and the third answer text.

8. An intelligent question-answering device, characterized in that: include: A mode selection module, used for determining a target question-answering mode from multiple question-answering modes based on the amount of information of the target question, wherein each question-answering mode needs to combine a different number of feature types; A text output module, used to determine a target answer text corresponding to the target question based on the target question-answering mode; The answer module is used to adjust the target answer text based on the broadcast characteristics, determine the target answer corresponding to the target question, and output the target answer.

9. An electronic device comprising a memory, a processor, and a computer program stored in the memory and running on the processor, characterized in that: When the processor executes the computer program, the steps of the method according to any one of claims 1 to 7 are implemented.

10. A computer-readable storage medium storing a computer program, characterized in that: When the computer program is executed by a processor, the steps of the method according to any one of claims 1 to 7 are implemented.

Images