CN108920541A

CN108920541A - A kind of recognition methods and private tutor's machine based on phonetic search

Info

Publication number: CN108920541A
Application number: CN201810603485.1A
Authority: CN
Inventors: 徐杨
Original assignee: Guangdong Genius Technology Co Ltd
Current assignee: Guangdong Genius Technology Co Ltd
Priority date: 2018-06-12
Filing date: 2018-06-12
Publication date: 2018-11-30

Abstract

A kind of recognition methods and private tutor's machine based on phonetic search, including：Whether private tutor's machine testing receives the first voice messaging of user's input；If receiving the first voice messaging, private tutor's machine determines and target accent database corresponding to the accent of the user of the first voice messaging of input according to the first voice messaging from the total accent database got in advance；Total accent database is that the accent based on the user to match with default statistical dimension is formed, and default statistical dimension includes region；Private tutor's machine identifies the voice content of the first voice messaging, the content to be searched as user according to target accent database.Implement the embodiment of the present invention, can be improved private tutor's machine to the recognition accuracy of phonetic problem, and then improves private tutor's eedle to the accuracy of the search result of phonetic problem.

Description

A kind of recognition methods and private tutor's machine based on phonetic search

Technical field

The present invention relates to private tutor's machine technical fields, and in particular to a kind of recognition methods and private tutor's machine based on phonetic search.

Background technique

When student is learnt using private tutor's machine, the phonetic search function of private tutor's machine is often used, that is, student can be to private tutor Machine says problem, and in turn, private tutor's machine can identify problem and retrieve answer corresponding with problem.Practice discovery, student is not Sound with growth stage can change, this will lead to private tutor's machine and reduces to the recognition accuracy of phonetic problem, and then cause Private tutor's eedle reduces the accuracy of the search result of phonetic problem.

Summary of the invention

The embodiment of the invention discloses a kind of recognition methods based on phonetic search and private tutor's machines, can be improved private tutor's machine pair The recognition accuracy of phonetic problem, and then private tutor's eedle is improved to the accuracy of the search result of phonetic problem.

First aspect of the embodiment of the present invention discloses a kind of recognition methods based on phonetic search, the method includes：

Whether private tutor's machine testing receives the first voice messaging of user's input；

If receiving first voice messaging, private tutor's machine is obtained according to first voice messaging from preparatory To total accent database in determine and target accent data corresponding to the accent for the user for inputting first voice messaging Library；Total accent database is that the accent based on the user to match with default statistical dimension is formed, the default statistics Dimension includes region；

Private tutor's machine identifies the voice content of first voice messaging according to the target accent database, as with The content to be searched at family.

As an alternative embodiment, the default statistical dimension is also wrapped in first aspect of the embodiment of the present invention Include age and gender；

Private tutor's machine is determined and is inputted from the total accent database got in advance according to first voice messaging Target accent database corresponding to the accent of the user of first voice messaging, including：

Private tutor's machine determines age of user, user's gender and user region according to first voice messaging；

Private tutor's machine is obtained according to the age of user, user's gender and the user region from advance Determination is corresponding with the age of user, user's gender and the user region in the total accent database got Target accent database.

As an alternative embodiment, private tutor's machine is according to the mesh in first aspect of the embodiment of the present invention Mark accent database identifies the voice content of first voice messaging, after the content to be searched of user, the method Further include：

Private tutor's machine is from content corresponding with the user region, the age of user and user's gender The content to match with the content to be searched is searched in database as object content, and exports the object content；

And after private tutor's machine exports the object content, the method also includes：

The evaluation page of satisfaction grade of the private tutor's machine output for being directed to the object content for user's input, and Detect the objective satisfaction degree grade of user's input；

Private tutor's machine judges whether the objective satisfaction degree grade is greater than default satisfaction grade, if it is not, then by institute It states the first voice messaging, the content to be searched and the object content and is uploaded to user feedback platform.

As an alternative embodiment, in first aspect of the embodiment of the present invention, the method also includes：

Private tutor's machine determines the corresponding first sound variation curve of the target accent database；

Private tutor's machine optimizes first sound variation according to corresponding first voice data of first voice messaging Curve obtains the second sound change curve of user, and stores the second sound change curve.

As an alternative embodiment, in first aspect of the embodiment of the present invention, private tutor's machine storage described the After two sound variation curves, the method also includes：

Whether private tutor's machine testing receives the second voice messaging of user's input；

If receiving second voice messaging, private tutor's machine determines the corresponding rising tone of second voice messaging Sound data, and judge in the second sound change curve with the presence or absence of certain voice data and the second sound Data Matching；

If there is certain described voice data and the second sound Data Matching, private tutor's machine is according to certain described sound Data and the target accent database identify the voice content of second voice messaging, and by second voice messaging Voice content is as current content to be searched.

Second aspect of the embodiment of the present invention discloses a kind of private tutor's machine, and private tutor's machine includes：

Detection unit, for detecting whether receiving the first voice messaging of user's input；

First determination unit, for after the detection unit receives first voice messaging, according to described One voice messaging determines the accent with the user for inputting first voice messaging from the total accent database got in advance Corresponding target accent database；Total accent database is the accent based on the user to match with default statistical dimension It is formed, the default statistical dimension includes region；

Recognition unit is made for identifying the voice content of first voice messaging according to the target accent database For the content to be searched of user.

As an alternative embodiment, the default statistical dimension is also wrapped in second aspect of the embodiment of the present invention Include age and gender；

First determination unit is determined from the total accent database got in advance according to first voice messaging The mode of target accent database corresponding to accent with the user for inputting first voice messaging is specially：

First determination unit is determined according to first voice messaging where age of user, user's gender and user Region, further according to the age of user, user's gender and the user region, from the total accent got in advance Target accent data corresponding with the age of user, user's gender and the user region are determined in database Library.

As an alternative embodiment, in second aspect of the embodiment of the present invention, private tutor's machine further includes：

Search unit, for identifying first voice messaging according to the target accent database in the recognition unit Voice content, after the content to be searched of user, from the user region, the age of user and described The content to match with the content to be searched is searched in the corresponding content data base of user's gender as object content, and is exported The object content；

Content is exported, for after described search unit exports the object content, output to be for inputting needle for user To the evaluation page of the satisfaction grade of the object content, and detect the objective satisfaction degree grade of user's input；

Judging unit, for judging whether the objective satisfaction degree grade is greater than default satisfaction grade；

Uploading unit, for judging that the objective satisfaction degree grade is not more than default satisfaction etc. in the judging unit After grade, first voice messaging, the content to be searched and the object content are uploaded to user feedback platform.

Second determination unit, for determining the corresponding first sound variation curve of the target accent database；

Optimize storage unit, for optimizing first sound according to corresponding first voice data of first voice messaging Sound change curve obtains the second sound change curve of user, and stores the second sound change curve.

As an alternative embodiment, in second aspect of the embodiment of the present invention, the detection unit is also used to After the optimization storage unit stores the second sound change curve, the second voice for receiving user's input is detected whether Information；

Private tutor's machine further includes：

Third determination unit, for after the detection unit receives second voice messaging, determining described The corresponding second sound data of two voice messagings；

The judging unit, be also used to judge in the second sound change curve with the presence or absence of certain voice data with it is described Second sound Data Matching；

The recognition unit is also used to judge there is certain described voice data and the rising tone in the judging unit After sound Data Matching, the language of second voice messaging is identified according to certain described voice data and the target accent database Sound content, and using the voice content of second voice messaging as current content to be searched.

The third aspect of the embodiment of the present invention discloses a kind of private tutor's machine, including：

It is stored with the memory of executable program code；

The processor coupled with the memory；

The processor calls the executable program code stored in the memory, executes the embodiment of the present invention the On the one hand the disclosed recognition methods based on phonetic search.

Fourth aspect of the embodiment of the present invention discloses a kind of computer readable storage medium, stores computer program, In, the computer program makes computer execute the identification side based on phonetic search disclosed in first aspect of the embodiment of the present invention Method.

The 5th aspect of the embodiment of the present invention discloses a kind of computer program product, when the computer program product is being counted When being run on calculation machine, so that the computer executes the recognition methods based on phonetic search disclosed in first aspect.

Compared with prior art, the embodiment of the present invention has the advantages that：

In the embodiment of the present invention, private tutor's machine can detecte the first voice messaging of user's input, when detecting the first voice When information, private tutor's machine can according to the first voice messaging from the total accent database got in advance (it should be noted that total mouth Sound database is that the accent based on the user to match with default statistical dimension is formed, and default statistical dimension includes region) in The corresponding target accent database of accent (for example, Beijing accent database) for inputting the user of first voice messaging is obtained, Private tutor's machine further according to target accent database identify the corresponding voice content of the first voice messaging (for example, friend's word of friend how Write), and using above-mentioned voice content as content to be searched.In conclusion implementing the embodiment of the present invention, private tutor's machine pair can be improved The recognition accuracy of phonetic problem, and then private tutor's eedle is improved to the accuracy of the search result of phonetic problem.

Detailed description of the invention

It to describe the technical solutions in the embodiments of the present invention more clearly, below will be to needed in the embodiment Attached drawing is briefly described, it should be apparent that, drawings in the following description are only some embodiments of the invention, for ability For the those of ordinary skill of domain, without creative efforts, it can also be obtained according to these attached drawings other attached Figure.

Fig. 1 is a kind of flow diagram of recognition methods based on phonetic search disclosed by the embodiments of the present invention；

Fig. 2 is the flow diagram of another recognition methods based on phonetic search disclosed by the embodiments of the present invention；

Fig. 3 is the flow diagram of another recognition methods based on phonetic search disclosed by the embodiments of the present invention；

Fig. 4 is a kind of structural schematic diagram of private tutor's machine disclosed by the embodiments of the present invention；

Fig. 5 is the structural schematic diagram of another private tutor's machine disclosed by the embodiments of the present invention；

Fig. 6 is the structural schematic diagram of another private tutor's machine disclosed by the embodiments of the present invention；

Fig. 7 is the structural schematic diagram of another private tutor's machine disclosed by the embodiments of the present invention.

Specific embodiment

Following will be combined with the drawings in the embodiments of the present invention, and technical solution in the embodiment of the present invention carries out clear, complete Site preparation description, it is clear that the described embodiment is only a part of the embodiment of the present invention, instead of all the embodiments.Based on this Embodiment in invention, every other reality obtained by those of ordinary skill in the art without making creative efforts Example is applied, shall fall within the protection scope of the present invention.

It should be noted that term " includes " and " having " and their any changes in the embodiment of the present invention and attached drawing Shape, it is intended that cover and non-exclusive include.Such as contain the process, method of a series of steps or units, system, product or Equipment is not limited to listed step or unit, but optionally further comprising the step of not listing or unit or optional Ground further includes the other step or units intrinsic for these process, methods, product or equipment.

The embodiment of the invention discloses a kind of recognition methods based on phonetic search and private tutor's machines, can be improved private tutor's machine pair The recognition accuracy of phonetic problem, and then private tutor's eedle is improved to the accuracy of the search result of phonetic problem.Individually below into Row is described in detail.

Embodiment one

Referring to Fig. 1, Fig. 1 is a kind of process signal of recognition methods based on phonetic search disclosed by the embodiments of the present invention Figure.Being somebody's turn to do the recognition methods based on phonetic search as shown in Figure 1 may comprise steps of：

101, whether private tutor's machine testing receives the first voice messaging of user's input, if so, 102 are thened follow the steps, If not, thening follow the steps 101.

In the embodiment of the present invention, optionally, whether private tutor's machine testing receives the side of the first voice messaging of user's input Formula is specifically as follows：

Private tutor's machine opens user speech real-time detection function, and judges that the first voice is believed when detecting the first voice messaging Whether the content of breath includes target password (for example, the small cloth of small cloth), if comprising target password, the first voice for will receiving Information is determined as the first voice messaging of user's input, if not including target password, defaults and does not receive user's input First voice messaging.Determine that private tutor's machine receives user's input when can include in this way target password in detecting voice messaging The first voice messaging, reduce private tutor's machine to do not include target password voice messaging misrecognition probability, improve private tutor The response probability for the first voice messaging that machine inputs user improves user to the usage experience of private tutor's machine.

Another optional in the embodiment of the present invention, whether private tutor's machine testing receives the first voice letter of user's input The mode of breath is specifically as follows：

Whether private tutor's machine testing receives the first voice messaging, if receiving the first voice messaging, is known by vocal print Other function judge the corresponding target vocal print feature of the first voice messaging whether with prestore legal voice print matching, if it does, then will The first voice messaging received is determined as the first voice messaging of user's input, if it does not match, default does not receive use First voice messaging of family input.It in this way can be when the first voice messaging be legitimate user's input not to the first voice messaging Content recognition is carried out, to reduce the power consumption of private tutor's machine.

102, private tutor's machine determines and input first according to the first voice messaging from the total accent database got in advance Target accent database corresponding to the accent of the user of voice messaging；Total accent database is to be based on and default statistical dimension phase What the accent of matched user was formed, default statistical dimension includes region.

In the embodiment of the present invention, total accent database includes the rim of the mouth sound database of different zones (for example, Guangdong accent number According to library, Beijing accent database, Shanghai accent database).In addition, target accent database (for example, Guangdong accent database) It is that the pronunciation, word and the analysis of word order for corresponding to dialect to target accent database based on big data obtain.

In the embodiment of the present invention, optionally, private tutor's machine is according to the first voice messaging, from the total accent data got in advance It determines in library and is specifically as follows with the mode of target accent database corresponding to the accent of the user of the first voice messaging of input：

Whether total accent database that private tutor's machine testing is got in advance is latest edition, if not latest edition, then Version updating is carried out to total accent database, then is determined and input the first voice letter from total accent database after version updating Target accent database (Guangdong accent database) corresponding to the accent (for example, Guangdong accent) of the user of breath.In this way can It is determined by the update to total accent database with the matched accent database of user's accent, the language that user is inputted with raising The recognition accuracy of sound；In addition, if the voice content of the first voice messaging of user's input is problem content, the present invention is implemented Example then can be improved private tutor's eedle to the accuracy of the search result of phonetic problem.

103, private tutor's machine identifies the voice content of the first voice messaging according to target accent database, as user wait search Rope content.

As an alternative embodiment, private tutor's machine identifies the voice of the first voice messaging according to target accent database After content, following operation is can also be performed in private tutor's machine：

Judge whether voice content includes problem content, if not including problem content (for example, voice content is " to accompany me Chat "), then preset object content is exported (for example, " I am to you with the corresponding target accent of target accent database Tell a story ")；If executed above-mentioned using voice content as the content to be searched of user comprising problem content.

As it can be seen that implementing the optional embodiment, it can be realized and not carried out in the case where by search engine with user Interactive voice, and then improve user experience.

As another optional embodiment, after step 103, following steps are can also be performed in private tutor's machine：

Judge in historical search record with the presence or absence of target search content and content matching (in simple terms, private tutor to be searched Machine judges whether the content to be searched is content that user once searched for)；

If it does not exist, then the search engine by private tutor's machine searches for content to be searched, search result is obtained；

Search result is exported, and is searched using search result and content to be searched as target histories search record storage in history In Suo Jilu；

If it is present the corresponding search result of output target search content.

As it can be seen that implementing the optional embodiment of the another kind, it can store and work as by the content for searching for user When user searches again for the content, the corresponding search result of the content is directly exported, the search efficiency of user is improved.

For example, machine examination of teaching in a family measure student say " the small small cloth of cloth, recommend it is some describe springs good sentences " when, Private tutor's machine can say region locating for the student of the words according to the total accent database judgement got in advance, and determine The target accent database of the words that the student says for identification, for example, Guangdong accent database；In turn, private tutor's machine can be with The words that student says is identified according to Guangdong accent database, obtains recognition result, and using the recognition result as Content to be searched, so that private tutor's machine scans for the content to be searched by search engine.

As it can be seen that implementing method described in Fig. 1, private tutor's machine can be improved to the recognition accuracy of phonetic problem, Jin Erti Accuracy of the high private tutor's eedle to the search result of phonetic problem；It can also be when the first voice messaging be legitimate user's input Content recognition is not carried out to the first voice messaging, to reduce the power consumption of private tutor's machine；It can also realize not by search engine In the case of with user carry out interactive voice, and then improve user experience.

Embodiment two

Referring to Fig. 2, the process that Fig. 2 is another recognition methods based on phonetic search disclosed by the embodiments of the present invention is shown It is intended to.Wherein, default statistical dimension can also include age and gender, as shown in Fig. 2, should the identification side based on phonetic search Method may comprise steps of：

Step 201~step 203 please refers in embodiment one for the description of step 201~step 203 for step The detailed description of 101~step 103, the embodiment of the present invention repeat no more.

204, private tutor's machine is searched for from content data base corresponding with user region, age of user and user's gender The content to match with content to be searched exports object content as object content.

In the embodiment of the present invention, content data base at least may include different subjects and different geographical electronic teaching material, often Know encyclopaedic knowledge and different types of story.

In the embodiment of the present invention, optionally, private tutor's machine is from corresponding with user region, age of user and user's gender Content data base in the content that matches with content to be searched of search as object content, and the mode for exporting object content has Body can be：

Whether deposited in private tutor's machine testing content data base corresponding with user region, age of user and user's gender In the content to match with content to be searched, if it is present using the content to match with content to be searched as object content And export object content.

Further, if there is no the content that matches with content to be searched in content data base, private tutor's machine can be with Execute following operation：

Private tutor's machine searches for content on the line to match with content to be searched, and content on output line by search engine.This Sample can be improved the speed of response to the phonetic problem of user, and then improve the search efficiency to phonetic problem.

205, the output of private tutor's machine is for the evaluation page for user's input for the satisfaction grade of object content, and detects The objective satisfaction degree grade of user's input.

In the embodiment of the present invention, the output form in the evaluation page for the satisfaction grade of object content can be The five-pointed star of predetermined number may be electronics marking plate, and the embodiment of the present invention is not construed as limiting.Wherein, if being directed to object content Satisfaction grade output form be predetermined number five-pointed star, then, machine examination of teaching in a family measure user's point touch one it is hollow Five-pointed star when private tutor's machine the hollow five-pointed star can be converted into solid five-pointed star.Machine examination of teaching in a family measures user's point and has touched entirely When the hollow five-pointed star in portion, then private tutor's machine default user is very satisfied to object content；Machine examination of teaching in a family measures user's point and touches When the hollow five-pointed star of more than half, then private tutor's machine default user is satisfied to object content；Machine examination of teaching in a family measures the touching of user's point When half and hollow five-pointed star below, then private tutor's machine default user is dissatisfied to object content.In addition, if being directed to target The output form of the satisfaction grade of content is electronics marking plate, is exported in the class then private tutor's machine can detecte user in marking Target fractional, if target fractional be greater than gross score 70%, default user is very satisfied to object content；If target Score is greater than the 40% of gross score and is less than or equal to the 70% of gross score, then default user is satisfied to object content；If target Score is less than or equal to the 40% of gross score, then default user is dissatisfied to object content.In this way for User, increasing Except the enjoyment for having added human-computer interaction, the developer also for private tutor's machine provides user feedback, so as to developer according to Family feedback optimizes private tutor's machine further progress, and then improves user experience.

206, private tutor's machine judges whether objective satisfaction degree grade is greater than default satisfaction grade, if it is, terminating this Process, if not, thening follow the steps 107.

207, the first voice messaging, content to be searched and object content are uploaded to user feedback platform by private tutor's machine.

As an alternative embodiment, private tutor's machine is according to the first voice messaging, from the total accent number got in advance Target accent database corresponding to accent according to the user that the first voice messaging is determined and inputted in library, may include following step Suddenly：

Private tutor's machine determines age of user, user's gender and user region according to the first voice messaging；

Private tutor's machine is according to age of user, user's gender and user region, from the total accent data got in advance Target accent database corresponding with age of user, user's gender and user region is determined in library.

It, can be by age (for example, five years old) of user, gender (for example, female as it can be seen that implement the optional embodiment Property) and region (for example, Guangdong) determine corresponding target accent database (for example, five years old accent number of Guangdong & women & According to library), the target accent database determined so is more matched with the information of user, when identifying the voice messaging of user's input Recognition accuracy it is also higher, in turn, can be improved the search accuracy rate of search result relevant to voice messaging.

As another optional embodiment, after step 207, following steps are can also be performed in private tutor's machine：

Remote connection request is sent to the mobile terminal (for example, mobile phone of teacher) connecting with private tutor's machine, and detection is The no response for receiving mobile terminal to remote connection request；

If receiving the response of remote connection request, it is connect with mobile terminal remote, and to mobile terminal transmission the One voice messaging, so that mobile terminal user feeds back for the first voice messaging.

As it can be seen that implementing another optional embodiment, user can ask when the search result to private tutor's machine is dissatisfied Help teacher, increases the enquirement approach of user, and then improve the learning effect of user.In addition, the satisfaction that private tutor's machine has Grade Function of Evaluation can supervise user to think deeply and judge whether the search result of private tutor's machine is oneself desired knot Fruit, to help user to improve learning ability.

As it can be seen that implementing method described in Fig. 2, private tutor's machine can be improved to the recognition accuracy of phonetic problem, Jin Erti Accuracy of the high private tutor's eedle to the search result of phonetic problem；It can also be when the first voice messaging be legitimate user's input Content recognition is not carried out to the first voice messaging, to reduce the power consumption of private tutor's machine；It can also realize not by search engine In the case of with user carry out interactive voice, and then improve user experience；The speed of the response to the phonetic problem of user can also be improved Rate, and then improve the search efficiency to phonetic problem；User can also be supervised to think deeply, user is helped to improve study energy Power.

Embodiment three

Referring to Fig. 3, the process that Fig. 3 is another recognition methods based on phonetic search disclosed by the embodiments of the present invention is shown It is intended to.Being somebody's turn to do the recognition methods based on phonetic search as shown in Figure 3 may comprise steps of：

Step 301~step 307, for the description of step 301~step 307, please refer in embodiment two for 201~ The detailed description of step 207, the embodiment of the present invention repeat no more.

308, private tutor's machine determines the corresponding first sound variation curve of target accent database.

In the embodiment of the present invention, optionally, private tutor's machine determines the corresponding first sound variation curve of target accent database Mode be specifically as follows：

Private tutor's machine determines the corresponding user's gender of target accent database, user region；

Private tutor's machine obtains according to user's gender and the user region from sound variation diagram database One sound variation curve is (for example, the first sound variation curve is the sound variation curve of Guangdong women, then the sound variation Curve covers the Guangdong women of all age brackets).It can determine that the first sound becomes by the gender and region of user in this way Change curve, to improve the recognition accuracy for increasing the voice messaging to user's input with age of user.

Another optional in the embodiment of the present invention, private tutor's machine determines that corresponding first sound of target accent database becomes The mode for changing curve is specifically as follows：

Private tutor's machine determines the vocal print feature of user according to target accent database and the first voice messaging, and is based on accent Big data analysis vocal print feature, and then generate the first sound variation curve.It in this way can be by the voice messaging that is inputted to user Analyze and then generates unique the first sound variation curve for the user (it should be noted that first sound Sound change curve can show the corresponding vocal print feature of each age of the user), improve the growth pair with age of user The recognition accuracy of the voice messaging of user's input.

309, private tutor's machine optimizes the first sound variation curve according to corresponding first voice data of the first voice messaging, obtains The second sound change curve of user, and store second sound change curve.

In the embodiment of the present invention, second sound change curve is the first sound variation curve after optimization.

310, whether private tutor's machine testing receives the second voice messaging of user's input, if so, 311 are thened follow the steps, If not, thening follow the steps 310.

311, private tutor's machine determines the corresponding second sound data of the second voice messaging, and judges in second sound change curve With the presence or absence of certain voice data and second sound Data Matching, if so, 312 are thened follow the steps, if it is not, then according to the second language Message breath is determined and is inputted from the total accent database got in advance corresponding to the accent of the user of the second voice messaging Target accent database, and according to the voice content of target accent database the first voice messaging of identification, as user wait search Rope content.

It include the second vocal print feature in second sound data, and the second vocal print feature is input language in the embodiment of the present invention The vocal print feature of the user of message breath.

312, private tutor's machine identifies the voice content of the second voice messaging according to certain voice data and target accent database, and Using the voice content of the second voice messaging as current content to be searched.

For example, if private tutor's machine determines that the first sound variation curve of five years old Guangdong female user, private tutor's machine are every It is secondary the first sound variation curve to be optimized when detecting user's input voice information, therefore, the first sound Change curve is more and more with optimization number, which also increasingly matches with the user, in turn, private tutor The recognition accuracy for the voice messaging that machine inputs the user also can be higher and higher.

As it can be seen that method described in implementing Fig. 3, can be improved private tutor's machine to the recognition accuracy of phonetic problem, Jin Erti Accuracy of the high private tutor's eedle to the search result of phonetic problem；It can also be when the first voice messaging be legitimate user's input Content recognition is not carried out to the first voice messaging, to reduce the power consumption of private tutor's machine；It can also realize not by search engine In the case of with user carry out interactive voice, and then improve user experience；The speed of the response to the phonetic problem of user can also be improved Rate, and then improve the search efficiency to phonetic problem；User can also be supervised to think deeply, user is helped to improve study energy Power；The recognition accuracy of the voice messaging inputted with the growth of age of user to user can also be improved.

Example IV

Referring to Fig. 4, Fig. 4 is a kind of structural schematic diagram of private tutor's machine disclosed by the embodiments of the present invention.As shown in figure 4, should Private tutor's machine may include：Detection unit 401, the first determination unit 402 and recognition unit 403, wherein

Detection unit 401, for detecting whether receiving the first voice messaging of user's input.

In the embodiment of the present invention, optionally, detection unit 401 detects whether to receive the first voice messaging of user's input Mode be specifically as follows：

Detection unit 401 opens user speech real-time detection function, and first is judged when detecting the first voice messaging Whether the content of voice messaging includes target password (for example, the small cloth of small cloth), if comprising target password, will receive One voice messaging is determined as the first voice messaging of user's input.It in this way can include target password in detecting voice messaging When determine private tutor's machine receive user input the first voice messaging, reduce private tutor's machine to do not include target password voice believe The misrecognition probability of breath improves the response probability for the first voice messaging that private tutor's machine inputs user, improves user to family The usage experience of religion machine.

Another optional in the embodiment of the present invention, detection unit 401 detects whether to receive the first language of user's input The mode of message breath is specifically as follows：

Detection unit 401 detects whether to receive the first voice messaging, if receiving the first voice messaging, passes through sound Line identification function judge the corresponding target vocal print feature of the first voice messaging whether with prestore legal voice print matching, if it does, The first voice messaging received is then determined as to the first voice messaging of user's input.The first voice messaging can be worked as not in this way Content recognition is not carried out to the first voice messaging when being legitimate user's input, to reduce the power consumption of private tutor's machine.

First determination unit 402, for after detection unit 401 receives the first voice messaging, according to the first voice Information determines and mesh corresponding to the accent of the user of the first voice messaging of input from the total accent database got in advance Mark accent database；Total accent database is that the accent based on the user to match with default statistical dimension is formed, and presets system Counting dimension includes region.

In the embodiment of the present invention, optionally, the first determination unit 402 is according to the first voice messaging, from what is got in advance The mode with target accent database corresponding to the accent of the user of the first voice messaging of input is determined in total accent database It is specifically as follows：

First determination unit 402 detects whether the total accent database got in advance is latest edition, if not most New version then carries out version updating to total accent database, then determines and input from total accent database after version updating Target accent database (Guangdong accent data corresponding to the accent (for example, Guangdong accent) of the user of first voice messaging Library).Can be determined in this way by the update to total accent database with the matched accent database of user's accent, with improve pair The recognition accuracy of the voice of user's input；In addition, if the voice content of the first voice messaging of user's input is in problem Hold, the embodiment of the present invention then can be improved private tutor's eedle to the accuracy of the search result of phonetic problem.

In the embodiment of the present invention, in the first determination unit 402 according to the first voice messaging, from the total accent got in advance Determine that triggering identifies with after target accent database corresponding to the accent of the user of the first voice messaging of input in database Unit 403 starts.

Recognition unit 403, for identifying the voice content of the first voice messaging according to target accent database, as user Content to be searched.

As an alternative embodiment, recognition unit 403, can be also used for identifying according to target accent database After the voice content of first voice messaging,

As another optional embodiment, recognition unit 403 can be also used for knowing according to target accent database The voice content of other first voice messaging, after the content to be searched of user：

As it can be seen that implementing private tutor's machine described in Fig. 4, private tutor's machine can be improved to the recognition accuracy of phonetic problem, in turn Private tutor's eedle is improved to the accuracy of the search result of phonetic problem；It can also not be legitimate user's input when the first voice messaging When not to the first voice messaging carry out content recognition, to reduce the power consumption of private tutor's machine；It can also realize not by search engine In the case where with user carry out interactive voice, and then improve user experience.

Embodiment five

Referring to Fig. 5, Fig. 5 is the structural schematic diagram of another private tutor's machine disclosed by the embodiments of the present invention.Wherein, Fig. 5 institute The private tutor's machine shown is that private tutor's machine as shown in Figure 4 optimizes；In addition, default statistical dimension further includes age and property Not.Compared with private tutor's machine shown in Fig. 4, private tutor's machine shown in fig. 5 can also include：Search unit 404, output content 405, Judging unit 406 and uploading unit 407, wherein

Search unit 404, for identifying the voice of the first voice messaging according to target accent database in recognition unit 403 Content, after the content to be searched of user, from content corresponding with user region, age of user and user's gender The content to match with content to be searched is searched in database as object content, and exports object content.

In the embodiment of the present invention, optionally, search unit 404 from user region, age of user and user's property The content to match with content to be searched is searched in not corresponding content data base as object content, and exports object content Mode is specifically as follows：

Search unit 404 detects in content data base corresponding with user region, age of user and user's gender With the presence or absence of the content to match with content to be searched, if it is present using the content to match with content to be searched as mesh Mark content simultaneously exports object content.

Further, if search unit 404 is also used to be not present and content phase to be searched in content data base After the content matched,

Content on the line to be matched by search engine search with content to be searched, and content on output line.In this way can The speed of response to the phonetic problem of user is improved, and then improves the search efficiency to phonetic problem.

Content 405 is exported, for after search unit 404 exports object content, output for user's input for be directed to The evaluation page of the satisfaction grade of object content, and detect the objective satisfaction degree grade of user's input.

In the embodiment of the present invention, the output form in the evaluation page for the satisfaction grade of object content can be The five-pointed star of predetermined number may be electronics marking plate, and the embodiment of the present invention is not construed as limiting.Wherein, if being directed to object content Satisfaction grade output form be predetermined number five-pointed star, then, when output content 405 detect user's point touch one The hollow five-pointed star can be converted to solid five-pointed star by private tutor's machine when hollow five-pointed star.When output content 405 detects use When family point has touched whole hollow five-pointed stars, then it is very satisfied to object content to export 405 default user of content；When output content 405 when detecting that user's point has touched the hollow five-pointed star of more than half, then it is full to object content to export 405 default user of content Meaning；When output content 405 detects that user's point has touched half and hollow five-pointed star below, then exports the default of content 405 and use Family is dissatisfied to object content.In addition, if the output form for the satisfaction grade of object content is electronics marking plate, that Output content 405 can detecte the target fractional that user exports in the class in marking, if target fractional is greater than gross score 70%, then default user is very satisfied to object content；If target fractional is greater than the 40% of gross score and is less than or equal to total score Several 70%, then default user is satisfied to object content；If target fractional is less than or equal to the 40% of gross score, default user It is dissatisfied to object content.It in this way for User, is increasing except the enjoyment of human-computer interaction, is being also private tutor's machine Developer provides user feedback, so that developer optimizes private tutor's machine further progress according to user feedback, Jin Ergai Kind user experience.

In the embodiment of the present invention, in the output of output content 405 for being directed to the satisfaction etc. of object content for user's input Grade the evaluation page, and detect user input objective satisfaction degree grade after, triggering judging unit 406 start.

Judging unit 406, for judging whether objective satisfaction degree grade is greater than default satisfaction grade.

In the embodiment of the present invention, judge whether objective satisfaction degree grade is greater than default satisfaction grade in judging unit 406 Later, triggering uploading unit 407 starts.

Uploading unit 407, for judging that objective satisfaction degree grade is not more than default satisfaction grade in judging unit 406 Later, the first voice messaging, content to be searched and object content are uploaded to user feedback platform.

First determination unit 402 according to the first voice messaging, determined from the total accent database got in advance with it is defeated The mode for entering target accent database corresponding to the accent of the user of the first voice messaging is specially：

First determination unit 402 determines age of user, user's gender and user region according to the first voice messaging, Further according to age of user, user's gender and user region, determines and use from the total accent database got in advance Family age, the corresponding target accent database of user's gender and user region.

As an alternative embodiment, uploading unit 407, can be also used for by the first voice messaging, to be searched Content and object content are uploaded to after user feedback platform,

As it can be seen that implementing a kind of optional embodiment, user can seek help when the search result to private tutor's machine is dissatisfied Teacher Yu increases the enquirement approach of user, and then improves the learning effect of user.In addition, the satisfaction etc. that private tutor's machine has Grade Function of Evaluation, can supervise user think deeply and judge private tutor's machine search result whether be oneself want as a result, To help user to improve learning ability.

As it can be seen that implementing private tutor's machine described in Fig. 5 can be improved private tutor's machine to the recognition accuracy of phonetic problem, Jin Erti Accuracy of the high private tutor's eedle to the search result of phonetic problem；It can also be when the first voice messaging be legitimate user's input Content recognition is not carried out to the first voice messaging, to reduce the power consumption of private tutor's machine；It can also realize not by search engine In the case of with user carry out interactive voice, and then improve user experience；The speed of the response to the phonetic problem of user can also be improved Rate, and then improve the search efficiency to phonetic problem；User can also be supervised to think deeply, user is helped to improve study energy Power.

Embodiment six

Referring to Fig. 6, Fig. 6 is the structural schematic diagram of another private tutor's machine disclosed by the embodiments of the present invention.Wherein, Fig. 6 institute The private tutor's machine shown is that private tutor's machine as shown in Figure 5 optimizes.It is shown in fig. 6 compared with private tutor's machine shown in fig. 5 Private tutor's machine can also include：Second determination unit 408, optimization storage unit 409 and third determination unit 410, wherein

Second determination unit 408, for determining the corresponding first sound variation curve of target accent database.

In the embodiment of the present invention, optionally, the second determination unit 408 determines corresponding first sound of target accent database The mode of change curve is specifically as follows：

Second determination unit 408 determines the corresponding user's gender of target accent database, user region；

Second determination unit 408 is according to user's gender and the user region from sound variation diagram database The middle first sound variation curve that obtains is (for example, the first sound variation curve is the sound variation curve of Guangdong women, then should Sound variation curve covers the Guangdong women of all age brackets).Can be determined by the gender and region of user in this way One sound variation curve, to improve the recognition accuracy for increasing the voice messaging to user's input with age of user.

Another optional in the embodiment of the present invention, the second determination unit 408 determines target accent database corresponding the The mode of one sound variation curve is specifically as follows：

Second determination unit 408 determines the vocal print feature of user according to target accent database and the first voice messaging, And it is based on accent big data analysis vocal print feature, and then generate the first sound variation curve.It in this way can be by being inputted to user Voice messaging analyze and then generate unique the first sound variation curve for the user and (need to illustrate It is that the first sound variation curve can show the corresponding vocal print feature of each age of the user), it improves with user year The recognition accuracy for the voice messaging that the growth in age inputs user.

In the embodiment of the present invention, corresponding first sound variation of target accent database is determined in the second determination unit 408 After curve, triggering optimization storage unit 409 starts.

Optimize storage unit 409, becomes for optimizing the first sound according to corresponding first voice data of the first voice messaging Change curve, obtains the second sound change curve of user, and store second sound change curve.

Detection unit 401 is also used to after optimization storage unit 409 stores second sound change curve, detects whether Receive the second voice messaging of user's input.

Third determination unit 410, for determining the second voice after detection unit 401 receives the second voice messaging The corresponding second sound data of information.

In the embodiment of the present invention, third determination unit 410 determine the corresponding second sound data of the second voice messaging it Afterwards, the execution of triggering judging unit 406 judges in second sound change curve with the presence or absence of certain voice data and second sound data Matched operation.

Judging unit 406 is also used to judge in second sound change curve with the presence or absence of certain voice data and second sound Data Matching.

Recognition unit 403 is also used to judge that there are certain voice datas and second sound Data Matching in judging unit 406 Later, the voice content of the second voice messaging is identified according to certain voice data and target accent database, and the second voice is believed The voice content of breath is as current content to be searched.

As it can be seen that implementing private tutor's machine described in Fig. 6 can be improved private tutor's machine to the recognition accuracy of phonetic problem, Jin Erti Accuracy of the high private tutor's eedle to the search result of phonetic problem；It can also be when the first voice messaging be legitimate user's input Content recognition is not carried out to the first voice messaging, to reduce the power consumption of private tutor's machine；It can also realize not by search engine In the case of with user carry out interactive voice, and then improve user experience；The speed of the response to the phonetic problem of user can also be improved Rate, and then improve the search efficiency to phonetic problem；User can also be supervised to think deeply, user is helped to improve study energy Power；The recognition accuracy of the voice messaging inputted with the growth of age of user to user can also be improved.

Embodiment seven

Referring to Fig. 7, Fig. 7 is the structural schematic diagram of another private tutor's machine disclosed by the embodiments of the present invention.As shown in fig. 7, For ease of description, only parts related to embodiments of the present invention are shown, disclosed by specific technical details, please refers to this hair Bright embodiment method part.

With reference to Fig. 7, private tutor's machine includes：Radio frequency (Radio Frequency, RF) circuit 710, memory 720, input unit 730, display unit 740, sensor 750, voicefrequency circuit 760, Wireless Fidelity (wireless fidelity, WiFi) module 770, the components such as processor 780, power supply 790 and camera (not shown).It will be understood by those skilled in the art that showing in Fig. 7 Private tutor's machine structure out does not constitute the restriction to private tutor's machine, may include than illustrating more or fewer components, or combination Certain components or different component layouts.

It is specifically introduced below with reference to each component parts of the Fig. 7 to private tutor's machine：

RF circuit 710 can be used for receiving and sending messages or communication process in, signal sends and receivees, particularly, by base station After downlink information receives, handled to processor 780；In addition, the data for designing uplink are sent to base station.In general, RF circuit 710 Including but not limited to antenna, at least one amplifier, transceiver, coupler, low-noise amplifier (Low Noise Amplifier, LNA), duplexer etc..In addition, RF circuit 710 can also be communicated with network and other equipment by wireless communication. Any communication standard or agreement, including but not limited to global system for mobile communications (Global can be used in above-mentioned wireless communication System of Mobile communication, GSM), general packet radio service (General Packet Radio Service, GPRS), CDMA (Code Division Multiple Access, CDMA), wideband code division multiple access (Wideband Code Division Multiple Access, WCDMA), long term evolution (Long Term Evolution, LTE), Email, short message service (Short Messaging Service, SMS) etc..

Memory 720 can be used for storing software program and module, and processor 780 is stored in memory 720 by operation Software program and module, thereby executing the various function application and data processing of private tutor's machine.Memory 720 can be wrapped mainly Include storing program area and storage data area, wherein storing program area can application needed for storage program area, at least one function Program (such as sound-playing function, image player function etc.) etc.；Storage data area can be stored is created according to using for private tutor's machine Data (such as audio data, phone directory etc.) built etc..In addition, memory 720 may include high-speed random access memory, also It may include nonvolatile memory, a for example, at least disk memory, flush memory device or the storage of other volatile solid-states Device.

Input unit 730 can be used for receiving the number or character information of input, and generate the user setting with private tutor's machine And the related key signals input of function control.Specifically, input unit 730 may include touch panel 731 and other inputs Equipment 732.Touch panel 731, also referred to as touch screen, collecting the touch operation of user on it or nearby, (for example user makes With the operation of any suitable object or attachment such as finger, stylus on touch panel 731 or near touch panel 731), and Corresponding attachment device is driven according to preset formula.Optionally, touch panel 731 may include touch detecting apparatus and touching Touch two parts of controller.Wherein, the touch orientation of touch detecting apparatus detection user, and detect touch operation bring letter Number, transmit a signal to touch controller；Touch controller receives touch information from touch detecting apparatus, and is converted into Contact coordinate, then give processor 780, and order that processor 780 is sent can be received and executed.Furthermore, it is possible to using The multiple types such as resistance-type, condenser type, infrared ray and surface acoustic wave realize touch panel 731.It is defeated in addition to touch panel 731 Entering unit 730 can also include other input equipments 732.Specifically, other input equipments 732 can include but is not limited to physics One of keyboard, function key (such as volume control button, switch key etc.), trace ball, mouse, operating stick etc. are a variety of.

Display unit 740 can be used for showing information input by user or be supplied to user information and private tutor's machine it is each Kind menu.Display unit 740 may include display panel 741, optionally, can use liquid crystal display (Liquid Crystal Display, LCD), the forms such as Organic Light Emitting Diode (Organic Light-Emitting Diode, OLED) it is aobvious to configure Show panel 741.Further, touch panel 731 can cover display panel 741, when touch panel 731 detect it is on it or attached After close touch operation, processor 780 is sent to determine the type of touch event, is followed by subsequent processing device 780 according to touch event Type corresponding visual output is provided on display panel 741.Although in Fig. 7, touch panel 731 and display panel 741 It is the input and input function of private tutor's machine to be realized as two independent components, but in some embodiments it is possible to will touch Control panel 731 and display panel 741 are integrated and that realizes private tutor's machine output and input function.

Private tutor's machine may also include at least one sensor 750, such as optical sensor, motion sensor and other sensings Device.Specifically, optical sensor may include ambient light sensor and proximity sensor, wherein ambient light sensor can be according to environment The light and shade of light adjusts the brightness of display panel 741, and proximity sensor can close display surface when private tutor's machine is moved in one's ear Plate 741 and/or backlight.As a kind of motion sensor, accelerometer sensor can detect in all directions (generally three axis) The size of acceleration can detect that size and the direction of gravity when static, the application that can be used to identify private tutor's machine posture is (such as horizontal Vertical screen switching, dependent game, magnetometer pose calibrating), Vibration identification correlation function (such as pedometer, tap) etc.；As for family The other sensors such as gyroscope, barometer, hygrometer, thermometer, infrared sensor that religion machine can also configure, it is no longer superfluous herein It states.

Voicefrequency circuit 760, loudspeaker 761, microphone 762 can provide the audio interface between user and private tutor's machine.Audio Electric signal after the audio data received conversion can be transferred to loudspeaker 761, be converted to sound by loudspeaker 761 by circuit 760 Sound signal output；On the other hand, the voice signal of collection is converted to electric signal by microphone 762, after being received by voicefrequency circuit 760 Audio data is converted to, then by after the processing of audio data output processor 780, such as another family is sent to through RF circuit 710 Religion machine, or audio data is exported to memory 720 to be further processed.

WiFi belongs to short range wireless transmission technology, and private tutor's machine can help user's transceiver electronics by WiFi module 770 Mail, browsing webpage and access streaming video etc., it provides wireless broadband internet access for user.Although Fig. 7 is shown WiFi module 770, but it is understood that, and it is not belonging to must be configured into for private tutor's machine, it can according to need completely Do not change in the range of the essence of invention and omits.

Processor 780 is the control centre of private tutor's machine, utilizes each portion of various interfaces and the entire private tutor's machine of connection Point, by running or execute the software program and/or module that are stored in memory 720, and calls and be stored in memory 720 Interior data execute the various functions and processing data of private tutor's machine, to carry out integral monitoring to private tutor's machine.Optionally, it handles Device 780 may include one or more processing units；Preferably, processor 780 can integrate application processor and modulation /demodulation processing Device, wherein the main processing operation system of application processor, user interface and application program etc., modem processor is mainly located Reason wireless communication.It is understood that above-mentioned modem processor can not also be integrated into processor 780.

Private tutor's machine further includes the power supply 790 (such as battery) powered to all parts, it is preferred that power supply can pass through power supply Management system and processor 780 are logically contiguous, to realize management charging, electric discharge and power consumption pipe by power-supply management system The functions such as reason.

Although being not shown, private tutor's machine can also be including bluetooth module etc., and details are not described herein.

In embodiments of the present invention, processor 780 included by private tutor's machine calls the software being stored in memory 720 Program specifically executes following operation：

It detects whether to receive the first voice messaging that user inputs；

After receiving the first voice messaging, according to the first voice messaging, from the total accent database got in advance Target accent database corresponding to the accent of the user of the first voice messaging of middle determination and input；Total accent database is to be based on What the accent of the user to match with default statistical dimension was formed, default statistical dimension includes region；

The voice content that the first voice messaging is identified according to target accent database, the content to be searched as user.

As it can be seen that implementing private tutor's machine described in Fig. 7 can be improved private tutor's machine to the recognition accuracy of phonetic problem, Jin Erti Accuracy of the high private tutor's eedle to the search result of phonetic problem.

The above, above embodiments are only to illustrate the technical solution of the application rather than its limitations；Although referring to before Embodiment is stated the application is described in detail, however those skilled in the art should understand that；It still can be with It modifies the technical solutions described in the foregoing embodiments or equivalent replacement of some of the technical features；And These are modified or replaceed, the spirit and model of each embodiment technical solution of the application that it does not separate the essence of the corresponding technical solution It encloses.

Claims

1. a kind of recognition methods based on phonetic search, which is characterized in that the method includes：

If receiving first voice messaging, private tutor's machine is according to first voice messaging, from what is got in advance It is determined in total accent database and target accent database corresponding to the accent for the user for inputting first voice messaging；Institute Stating total accent database is that the accent based on the user to match with default statistical dimension is formed, the default statistical dimension packet Include region；

Private tutor's machine identifies the voice content of first voice messaging according to the target accent database, as user's Content to be searched.

2. the method according to claim 1, wherein the default statistical dimension further includes age and gender；

Private tutor's machine according to first voice messaging, determined from the total accent database got in advance with described in input Target accent database corresponding to the accent of the user of first voice messaging, including：

Private tutor's machine is got according to the age of user, user's gender and the user region from advance Total accent database in determine corresponding with the age of user, user's gender and user region target Accent database.

3. according to the method described in claim 2, it is characterized in that, private tutor's machine is identified according to the target accent database The voice content of first voice messaging, after the content to be searched of user, the method also includes：

Private tutor's machine is from content-data corresponding with the user region, the age of user and user's gender The content to match with the content to be searched is searched in library as object content, and exports the object content；

Private tutor's machine output is detected for the evaluation page for user's input for the satisfaction grade of the object content The objective satisfaction degree grade of user's input；

Private tutor's machine judges whether the objective satisfaction degree grade is greater than default satisfaction grade, if it is not, then by described One voice messaging, the content to be searched and the object content are uploaded to user feedback platform.

4. method according to claim 1-3, which is characterized in that the method also includes：

Private tutor's machine optimizes the first sound variation curve according to corresponding first voice data of first voice messaging, The second sound change curve of user is obtained, and stores the second sound change curve.

5. according to the method described in claim 4, it is characterized in that, private tutor's machine store the second sound change curve it Afterwards, the method also includes：

If receiving second voice messaging, private tutor's machine determines the corresponding second sound number of second voice messaging According to, and judge in the second sound change curve with the presence or absence of certain voice data and the second sound Data Matching；

If there is certain described voice data and the second sound Data Matching, private tutor's machine is according to certain described voice data Identify the voice content of second voice messaging with the target accent database, and by the voice of second voice messaging Content is as current content to be searched.

6. a kind of private tutor's machine, which is characterized in that private tutor's machine includes：

First determination unit, for after the detection unit receives first voice messaging, according to first language Message breath, it is right with the accent for the user for inputting first voice messaging institute to determine from the total accent database got in advance The target accent database answered；Total accent database is that the accent based on the user to match with default statistical dimension is formed , the default statistical dimension includes region；

Recognition unit, for identifying the voice content of first voice messaging according to the target accent database, as with The content to be searched at family.

7. private tutor's machine according to claim 6, which is characterized in that the default statistical dimension further includes age and property Not；

First determination unit according to first voice messaging, determined from the total accent database got in advance with it is defeated The mode for entering target accent database corresponding to the accent of the user of first voice messaging is specially：

First determination unit determines age of user, user's gender and user location according to first voice messaging Domain, further according to the age of user, user's gender and the user region, from the total accent number got in advance According to target accent data corresponding with the age of user, user's gender and the user region determining in library Library.

8. private tutor's machine according to claim 7, which is characterized in that private tutor's machine further includes：

Search unit, for identifying the language of first voice messaging according to the target accent database in the recognition unit Sound content, after the content to be searched of user, from the user region, the age of user and the user The content to match with the content to be searched is searched in the corresponding content data base of gender as object content, and described in output Object content；

Content is exported, for after described search unit exports the object content, output to be for being directed to institute for user's input The evaluation page of the satisfaction grade of object content is stated, and detects the objective satisfaction degree grade of user's input；

Uploading unit, for the judging unit judge the objective satisfaction degree grade be not more than default satisfaction grade it Afterwards, first voice messaging, the content to be searched and the object content are uploaded to user feedback platform.

9. according to the described in any item private tutor's machines of claim 6~8, which is characterized in that private tutor's machine further includes：

Optimize storage unit, becomes for optimizing first sound according to corresponding first voice data of first voice messaging Change curve, obtains the second sound change curve of user, and store the second sound change curve.

10. private tutor's machine according to claim 9, which is characterized in that

The detection unit is also used to after the optimization storage unit stores the second sound change curve, and detection is No the second voice messaging for receiving user's input；

Private tutor's machine further includes：

Third determination unit, for determining second language after the detection unit receives second voice messaging Message ceases corresponding second sound data；

The judging unit is also used to judge in the second sound change curve with the presence or absence of certain voice data and described second Voice data matching；

The recognition unit is also used to judge there is certain described voice data and the second sound number in the judging unit After matching, identified in the voice of second voice messaging according to certain described voice data and the target accent database Hold, and using the voice content of second voice messaging as current content to be searched.