CN109145276A - A kind of text correction method after speech-to-text based on phonetic - Google Patents
A kind of text correction method after speech-to-text based on phonetic Download PDFInfo
- Publication number
- CN109145276A CN109145276A CN201810922512.1A CN201810922512A CN109145276A CN 109145276 A CN109145276 A CN 109145276A CN 201810922512 A CN201810922512 A CN 201810922512A CN 109145276 A CN109145276 A CN 109145276A
- Authority
- CN
- China
- Prior art keywords
- text
- phonetic
- editable
- distance
- word
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F40/00—Handling natural language data
- G06F40/10—Text processing
- G06F40/166—Editing, e.g. inserting or deleting
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F40/00—Handling natural language data
- G06F40/20—Natural language analysis
- G06F40/279—Recognition of textual entities
- G06F40/289—Phrasal analysis, e.g. finite state techniques or chunking
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F40/00—Handling natural language data
- G06F40/30—Semantic analysis
-
- G—PHYSICS
- G10—MUSICAL INSTRUMENTS; ACOUSTICS
- G10L—SPEECH ANALYSIS OR SYNTHESIS; SPEECH RECOGNITION; SPEECH OR VOICE PROCESSING; SPEECH OR AUDIO CODING OR DECODING
- G10L15/00—Speech recognition
- G10L15/26—Speech to text systems
Abstract
The invention discloses the text correction methods after a kind of speech-to-text based on phonetic, this method will be by text information made of speech recognition conversion by preliminary judgement, when there is unidentified information relevant to content out, according to the Pinyin information identified, it is calculated by pinyin similarity and corresponding text is replaced, the correction of voice is realized, in the hope of obtaining accurate semanteme.Phonetic similarity judgement of the present invention is modified with faster speed and is exported with the speech text of high-accuracy, and its implementation is easy, and the accuracy rate and service quality of speech recognition can significantly be guaranteed.
Description
Technical field
The present invention relates to the texts after artificial intelligent voice identification field more particularly to a kind of speech-to-text based on phonetic
This bearing calibration.
Background technique
In the latest 20 years, speech recognition technology obtains marked improvement, starts to move towards market from laboratory.Speech recognition technology
Progress into the every field such as industry, household electrical appliances, communication, automotive electronics, medical treatment, home services, consumption electronic product.Speech recognition
Technology constantly promoted, the various robots based on speech recognition start to come into being, however because everyone birth
Transient causes, the actual uses of speech recognition such as the immanent causes such as ground, pronunciation habit and signal interference, network are bad are accurate
Rate is far below 97% that businessman is boasted.And actual speech identification accuracy rate greatly affect need using speech recognition into
The business and work of row subsequent operation, therefore during practice, it needs to expend a large amount of manpowers and the time goes processing to identify
Inaccuracy and bring trouble, undertake corresponding economic loss.
Existing technology main direction is the tuning and improvement to speech recognition, and the liter of technology is carried out on the algorithm of identification
Grade, reaches higher recognition capability, seldom directs attention to correct this aspect to progress secondary treatment after speech recognition, existing
That deposits is also corrected just for homonym.However it is that recognition capability is inadequate that many situations, which are not, in existing standard mandarin
Under the technical background that discrimination can nearly all accurately identify, causing the reason of identifying deviation is the pronunciation difference and ring due to people
Border bring interference etc., these problems are depended merely on the space that promotion recognition capability is difficult to capture or be promoted and are extremely limited.And unisonance
Although the correction of word can make up for it the mistake of a part, but more in the case of be non-homonym caused by various complicated reasons
Situation, thus market with greater need for be certain Fuzzy Processing ability bearing calibration.
Summary of the invention
In view of the deficiencies of the prior art, the present invention discloses the text correction side after a kind of speech-to-text based on phonetic
The accuracy rate of method, the speech recognition that this method obtains is high, and specific technical solution is as follows:
Text correction method after a kind of speech-to-text based on phonetic, which is characterized in that this method includes following step
It is rapid:
S1: the Chinese text information after speech recognition is subjected to cutting by Chinese Word Automatic Segmentation or tool, is obtained multiple
Word；
S2: it searches in database under the application scenarios of this section of voice, keyword relevant to the word obtained in S1, to S1
Multiple words of middle acquisition are matched with obtained keyword；The database includes the submodule of multiple application scenarios,
Multiple keywords relevant to the scene are stored in each submodule；According to different scenes, setting needs to match keyword
Number requires if reaching matching, does not need to be corrected, directly by text output；Otherwise, into S3；
S3: i-th of each keyword in each word that S1 is obtained in i-th of Chinese character and database under the scene is calculated
The editable discrepancy of distance D of the phonetic of a Chinese characteri, the editable discrepancy of distance of the phonetic is the single word to phonetic
Two phonetics are become duplicate minimal modifications number by way of increase, deletion or replacement by symbol, each word
Editable discrepancy of distance D=∑ Di, given threshold k, when the word that S1 is obtained is for the smallest editable of all keywords
Discrepancy of distance DminWhen≤k, then the word in the corresponding S1 of the editable diversity factor is replaced with into corresponding key in database
Word；
S4: by the text output after replacement, i.e. completion text correction.
Further, when searching keyword in the database of S2, only the noun after S1 cutting is matched.
Further, the database includes the submodule of multiple application scenarios, is stored and this in each submodule
The relevant multiple keywords of scape.
Further, the editable discrepancy of distance D of the phonetic of i-th of Chinese character in the S3iFor initial consonant, simple or compound vowel of a Chinese syllable and
The sum of three kinds of editable discrepancy of distance of tone, i.e. Di=d1+d2+d3, wherein the editable discrepancy of distance of initial consonant and simple or compound vowel of a Chinese syllable
D1, d2 are identical with the definition in S3, and the editable discrepancy of distance of the tone defines d3 and is, tone is mutually all 0, are not all
1。
Further, respectively w1, the w2 of the weight of the editable discrepancy of distance of the initial consonant, simple or compound vowel of a Chinese syllable and tone,
W3, then Di=w1d1+w2d2+w3d3, and w1 >=w2 >=w3.
Beneficial effects of the present invention are as follows:
The phonetic that the present invention uses is the basis of Chinese language, is most to carry model close to the language semantic of language voice,
The semantic loss converted in identification process is utmostly reduced, and efficiency more reasonable for the makeover process of phonetic is more
Height, its implementation is easy, and means are flexible, and the accuracy rate and service quality of speech recognition can significantly be guaranteed.
Detailed description of the invention
Fig. 1 is the text correction method flow chart after the speech-to-text based on phonetic.
Specific embodiment
Below according to attached drawing and preferred embodiment the present invention is described in detail, the objects and effects of the present invention will become brighter
White, below in conjunction with drawings and examples, the present invention will be described in further detail.It should be appreciated that described herein specific
Embodiment is only used to explain the present invention, is not intended to limit the present invention.
As shown in Figure 1, the text correction method after the speech-to-text of the invention based on phonetic, specifically includes following step
It is rapid:
Step 1: the Chinese text information after speech recognition is subjected to cutting by Chinese Word Automatic Segmentation or tool, is obtained
Multiple words；
Step 2: searching the keyword in database under the application scenarios of this section of voice, multiple by what is obtained in step 1
Word is matched with keyword；The database includes the submodule of multiple application scenarios, in each submodule storage with
The relevant multiple nominal keywords of the scene；According to different scenes, setting needs to match the number of keyword, if reached
It is required to matching, does not then need to be corrected, directly by text output；Otherwise, three are entered step；
Step 3: the correction of Chinese text information is carried out according to the editable of phonetic distance.
The step is core of the invention, is divided into following sub-step.
1) each keyword in each word that step 1 obtains in i-th of Chinese character and database under the scene is calculated
The editable discrepancy of distance D of the phonetic of i-th of Chinese characteri, the editable discrepancy of distance of the phonetic is the list to phonetic
Two phonetics are become duplicate minimal modifications number by way of increase, deletion or replacement by a character, right respectively
Initial consonant, the simple or compound vowel of a Chinese syllable harmony of corresponding Chinese character transfer in the calculating of row diversity factor, and the editable discrepancy of distance of tone is defined as, tone phase
It is all 0, is not all 1, calculated result d1, d2 and d3, while according to pronunciation law, for the difference of initial consonant, simple or compound vowel of a Chinese syllable and tone
Weight coefficient w1, w2 and w3, Di=d1*w1+d2*w2+d3*w3 is respectively set in degree；
2) the editable discrepancy of distance D=∑ D of single word is calculatedi, according to the threshold value k of setting, when the word that S1 is obtained
The smallest editable discrepancy of distance D of the language for all keywordsminIt, then will be in the corresponding S1 of the editable diversity factor when≤k
Word replace with corresponding keyword in database；
3) sub-step 1 and 2 is recycled, until all vocabulary obtained by step 1 are all calculated and are disposed；
Step 4: by the text output after replacement, i.e. completion text correction.
The present invention is described in detail using house property as 2 representational embodiments of Foreground selection below, herein data
Library includes " price ", " position ", " mating " keyword, and threshold k=5, initial consonant, simple or compound vowel of a Chinese syllable, tone weight are 2:1:0.5.Following reality
It applies example for convenience of explanation, is respectively provided with when matching a keyword i.e. it is believed that successful match.
In order to improve matching efficiency, the workload of calculating is reduced, is somebody's turn to do in the multiple words obtained in S1 and database
When keyword under the application scenarios of Duan Yuyan is matched, only noun is matched.
Embodiment 1:
Correct text: how is house price?
Does identification text: house price swell sample?
Step 1: the Chinese text information after speech recognition is subjected to cutting by Chinese Word Automatic Segmentation or tool, is obtained
Multiple words, this identify that text is split as " house ", " price ", " swollen sample ", and wherein noun has " house ", " price "；
Step 2: the keyword in database under house property application scenarios is searched, in the multiple words obtained in step 1
Noun matched with obtained keyword, identify " price " keyword, directly export result, because having been obtained
Required key message, " swollen sample " do not influence as secondary information to semantic judgement, can terminate without correcting.
Embodiment 2:
Correct text: here house mating (pei4 tao4) how?
Identify text: here house quilt cover (bei4 tao4) how?
Step 1: the Chinese text information after speech recognition is subjected to cutting by Chinese Word Automatic Segmentation or tool, is obtained
Multiple words, can will identify that in the example come character segmentation be " you ", " here ", " house ", " quilt cover ", " how "；
Step 2: relevant keyword under house property application scenarios in lookup database, it will be in the word that obtained in step 1
Noun " house ", the keyword " price " in " quilt cover " and database under the scene, " position ", " mating " carry out respectively
Match, two words are unrecognized to be come out, and enters step three；
Step 3: the correction of Chinese text information is carried out according to the editable of phonetic distance；
1) each keyword in each word that step 1 obtains in i-th of Chinese character and database under the scene is calculated
The editable discrepancy of distance D of the phonetic of i-th of Chinese characteri；
When word a is " house " in the example, keyword b is " price ", compares the initial consonant " f " of the 1st word " room " and " valence "
" j " need to only be replaced, therefore initial consonant editable discrepancy of distance d1=1, and simple or compound vowel of a Chinese syllable is respectively " ang " and " ia ", need by
" ng " removes along with " i " can just make two simple or compound vowel of a Chinese syllable identical, therefore the editable discrepancy of distance of simple or compound vowel of a Chinese syllable is d2=3, tone one
A is the rising tone, and the editable distance of another falling tone, tone is d3=1, calculate the phonetic editable of first Chinese character away from
From diversity factor D1=2*1+1*3+0.5*1=5.5；
Similarly second Chinese character is calculated, the D2=1*2+1*1+0.5=3.5 that can be calculated.
The finally editable discrepancy of distance D=D1+D2=9 of " house " for keyword " price "；
Similar calculating process,
The editable distance D=10 of " house " for keyword " position "；
The editable distance D=13 of " house " for keyword " mating "；
The Dmin=9 of word " house ", is greater than K, without any replacement；
3) sub-step 1 and 2 is recycled, until the word that all steps 1 provide has been calculated:
The editable distance D=2 of " quilt cover " for keyword " mating "；
The editable discrepancy of distance D=10 of " quilt cover " for keyword " price "；
" quilt cover " is D=8 for the editable distance of keyword " position "；
Therefore, for word " quilt cover ", the smallest editable is 2 apart from value, i.e. Dmin=2, Dmin ratio K value
Small, the corresponding vocabulary of Dmin and keyword are respectively " quilt cover " and " mating ", and " quilt cover " is replaced with " mating "；
4) text output that will be disposed, text " here house quilt cover how? " correction of a final proof be " you this
In house it is mating how? "；
It will appreciated by the skilled person that being not used to limit the foregoing is merely the preferred embodiment of invention
System invention, although invention is described in detail referring to previous examples, for those skilled in the art, still
It can modify to the technical solution of aforementioned each case history or equivalent replacement of some of the technical features.It is all
Within the spirit and principle of invention, modification, equivalent replacement for being made etc. be should be included within the protection scope of invention.
Claims (5)
1. the text correction method after a kind of speech-to-text based on phonetic, which is characterized in that this method comprises the following steps:
S1: the Chinese text information after speech recognition is subjected to cutting by Chinese Word Automatic Segmentation or tool, obtains multiple words；
S2: it searches in database under the application scenarios of this section of voice, keyword relevant to the word obtained in S1, to being obtained in S1
The multiple words obtained are matched with obtained keyword；The database includes the submodule of multiple application scenarios, each
Multiple keywords relevant to the scene are stored in submodule；According to different scenes, setting needs to match the number of keyword,
If reaching matching to require, do not need to be corrected, directly by text output；Otherwise, into S3；
S3: i-th of Chinese of each keyword in each word that S1 is obtained in i-th of Chinese character and database under the scene is calculated
The editable discrepancy of distance D of the phonetic of wordi, the editable discrepancy of distance of the phonetic is logical to the single character of phonetic
Two phonetics are become duplicate minimal modifications number by the mode for crossing increase, deletion or replacement, and each word is compiled
Collect discrepancy of distance D=∑ Di, given threshold k, when the word that S1 is obtained is for the smallest editable distance of all keywords
Diversity factor DminWhen≤k, then the word in the corresponding S1 of the editable diversity factor is replaced with into corresponding keyword in database.
S4: by the text output after replacement, i.e. completion text correction.
2. the text correction method after the speech-to-text according to claim 1 based on phonetic, which is characterized in that in S2
Database in search keyword when, only the noun after S1 cutting is matched.
3. the text correction method after the speech-to-text according to claim 1 based on phonetic, which is characterized in that described
Database include multiple application scenarios submodule, storage multiple keywords relevant to the scene in each submodule.
4. the text correction method after the speech-to-text according to claim 1 based on phonetic, which is characterized in that described
S3 in i-th of Chinese character phonetic editable discrepancy of distance DiFor three kinds of initial consonant, simple or compound vowel of a Chinese syllable and tone editable range differences
The sum of different degree, i.e. Di=d1+d2+d3, wherein the definition phase in editable discrepancy of distance d1, d2 and S3 of initial consonant and simple or compound vowel of a Chinese syllable
Together, it is that tone is mutually all 0 that the editable discrepancy of distance of the tone, which defines d3, is not all 1.
5. the text correction method after the speech-to-text according to claim 4 based on phonetic, which is characterized in that described
Initial consonant, simple or compound vowel of a Chinese syllable and tone the weight of editable discrepancy of distance be respectively w1, w2, w3, then Di=w1d1+w2d2+
W3d3, and w1 >=w2 >=w3.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201810922512.1A CN109145276A (en) | 2018-08-14 | 2018-08-14 | A kind of text correction method after speech-to-text based on phonetic |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201810922512.1A CN109145276A (en) | 2018-08-14 | 2018-08-14 | A kind of text correction method after speech-to-text based on phonetic |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN109145276A true CN109145276A (en) | 2019-01-04 |
Family
ID=64793340
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201810922512.1A Pending CN109145276A (en) | 2018-08-14 | 2018-08-14 | A kind of text correction method after speech-to-text based on phonetic |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN109145276A (en) |
Cited By (24)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109948124A (en) * | 2019-03-15 | 2019-06-28 | 腾讯科技（深圳）有限公司 | Voice document cutting method, device and computer equipment |
| CN109977412A (en) * | 2019-03-29 | 2019-07-05 | 北京林业大学 | A kind of field value error correction method, device, readable medium and storage control |
| CN110334348A (en) * | 2019-06-28 | 2019-10-15 | 珍岛信息技术（上海）股份有限公司 | A kind of text method of calibration based in plain text |
| CN110399608A (en) * | 2019-06-04 | 2019-11-01 | 深思考人工智能机器人科技（北京）有限公司 | A kind of conversational system text error correction system and method based on phonetic |
| CN110767217A (en) * | 2019-10-30 | 2020-02-07 | 爱驰汽车有限公司 | Audio segmentation method, system, electronic device and storage medium |
| CN110782892A (en) * | 2019-10-25 | 2020-02-11 | 四川长虹电器股份有限公司 | Voice text error correction method |
| CN110880316A (en) * | 2019-10-16 | 2020-03-13 | 苏宁云计算有限公司 | Audio output method and system |
| CN110970026A (en) * | 2019-12-17 | 2020-04-07 | 用友网络科技股份有限公司 | Voice interaction matching method, computer device and computer-readable storage medium |
| CN111028834A (en) * | 2019-10-30 | 2020-04-17 | 支付宝(杭州)信息技术有限公司 | Voice message reminding method and device, server and voice message reminding equipment |
| CN111611792A (en) * | 2020-05-21 | 2020-09-01 | 全球能源互联网研究院有限公司 | Entity error correction method and system for voice transcription text |
| CN111831201A (en) * | 2020-05-25 | 2020-10-27 | 中国人民解放军陆军军医大学第二附属医院 | Human-computer interaction system and method for automatically detecting bone marrow cell morphology |
| CN112114926A (en) * | 2020-09-25 | 2020-12-22 | 北京百度网讯科技有限公司 | Page operation method, device, equipment and medium based on voice recognition |
| CN112259182A (en) * | 2020-11-05 | 2021-01-22 | 中国联合网络通信集团有限公司 | Method and device for generating electronic medical record |
| CN112562668A (en) * | 2020-11-30 | 2021-03-26 | 广州橙行智动汽车科技有限公司 | Semantic information deviation rectifying method and device |
| CN112560493A (en) * | 2020-12-17 | 2021-03-26 | 金蝶软件（中国）有限公司 | Named entity error correction method, named entity error correction device, computer equipment and storage medium |
| CN112863531A (en) * | 2021-01-12 | 2021-05-28 | 蒋亦韬 | Method for speech audio enhancement by regeneration after computer recognition |
| CN113053359A (en) * | 2019-12-27 | 2021-06-29 | 深圳Tcl数字技术有限公司 | Voice recognition method, intelligent terminal and storage medium |
| CN113223509A (en) * | 2021-04-28 | 2021-08-06 | 华南理工大学 | Fuzzy statement identification method and system applied to multi-person mixed scene |
| CN113297348A (en) * | 2021-04-15 | 2021-08-24 | 国网江苏省电力有限公司南京供电分公司 | Correction method for speech recognition Chinese text |
| CN113743093A (en) * | 2020-06-17 | 2021-12-03 | 北京沃东天骏信息技术有限公司 | Text correction method and device |
| CN113744722A (en) * | 2021-09-13 | 2021-12-03 | 上海交通大学宁波人工智能研究院 | Off-line speech recognition matching device and method for limited sentence library |
| CN113763961A (en) * | 2020-06-02 | 2021-12-07 | 阿里巴巴集团控股有限公司 | Text processing method and device |
| CN116052657A (en) * | 2022-08-01 | 2023-05-02 | 荣耀终端有限公司 | Character error correction method and device for voice recognition |
| CN112560493B (en) * | 2020-12-17 | 2024-04-30 | 金蝶软件（中国）有限公司 | Named entity error correction method, named entity error correction device, named entity error correction computer equipment and named entity error correction storage medium |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103324621A (en) * | 2012-03-21 | 2013-09-25 | 北京百度网讯科技有限公司 | Method and device for correcting spelling of Thai texts |
| US20160179774A1 (en) * | 2014-12-18 | 2016-06-23 | International Business Machines Corporation | Orthographic Error Correction Using Phonetic Transcription |
| CN105869642A (en) * | 2016-03-25 | 2016-08-17 | 海信集团有限公司 | Voice text error correction method and device |
| CN107741928A (en) * | 2017-10-13 | 2018-02-27 | 四川长虹电器股份有限公司 | A kind of method to text error correction after speech recognition based on field identification |
-
2018
- 2018-08-14 CN CN201810922512.1A patent/CN109145276A/en active Pending
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103324621A (en) * | 2012-03-21 | 2013-09-25 | 北京百度网讯科技有限公司 | Method and device for correcting spelling of Thai texts |
| US20160179774A1 (en) * | 2014-12-18 | 2016-06-23 | International Business Machines Corporation | Orthographic Error Correction Using Phonetic Transcription |
| CN105869642A (en) * | 2016-03-25 | 2016-08-17 | 海信集团有限公司 | Voice text error correction method and device |
| CN107741928A (en) * | 2017-10-13 | 2018-02-27 | 四川长虹电器股份有限公司 | A kind of method to text error correction after speech recognition based on field identification |
Cited By (35)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109948124A (en) * | 2019-03-15 | 2019-06-28 | 腾讯科技（深圳）有限公司 | Voice document cutting method, device and computer equipment |
| CN109948124B (en) * | 2019-03-15 | 2022-12-23 | 腾讯科技（深圳）有限公司 | Voice file segmentation method and device and computer equipment |
| CN109977412A (en) * | 2019-03-29 | 2019-07-05 | 北京林业大学 | A kind of field value error correction method, device, readable medium and storage control |
| CN109977412B (en) * | 2019-03-29 | 2022-12-27 | 北京林业大学 | Method and device for correcting field value of voice recognition text and storage controller |
| CN110399608B (en) * | 2019-06-04 | 2023-04-25 | 深思考人工智能机器人科技（北京）有限公司 | Text error correction system and method for dialogue system based on pinyin |
| CN110399608A (en) * | 2019-06-04 | 2019-11-01 | 深思考人工智能机器人科技（北京）有限公司 | A kind of conversational system text error correction system and method based on phonetic |
| CN110334348A (en) * | 2019-06-28 | 2019-10-15 | 珍岛信息技术（上海）股份有限公司 | A kind of text method of calibration based in plain text |
| CN110334348B (en) * | 2019-06-28 | 2022-11-15 | 珍岛信息技术（上海）股份有限公司 | Character checking method based on plain text |
| CN110880316A (en) * | 2019-10-16 | 2020-03-13 | 苏宁云计算有限公司 | Audio output method and system |
| CN110782892A (en) * | 2019-10-25 | 2020-02-11 | 四川长虹电器股份有限公司 | Voice text error correction method |
| CN110782892B (en) * | 2019-10-25 | 2022-03-25 | 四川长虹电器股份有限公司 | Voice text error correction method |
| CN111028834A (en) * | 2019-10-30 | 2020-04-17 | 支付宝(杭州)信息技术有限公司 | Voice message reminding method and device, server and voice message reminding equipment |
| CN110767217A (en) * | 2019-10-30 | 2020-02-07 | 爱驰汽车有限公司 | Audio segmentation method, system, electronic device and storage medium |
| CN110767217B (en) * | 2019-10-30 | 2022-04-12 | 爱驰汽车有限公司 | Audio segmentation method, system, electronic device and storage medium |
| CN110970026A (en) * | 2019-12-17 | 2020-04-07 | 用友网络科技股份有限公司 | Voice interaction matching method, computer device and computer-readable storage medium |
| CN113053359A (en) * | 2019-12-27 | 2021-06-29 | 深圳Tcl数字技术有限公司 | Voice recognition method, intelligent terminal and storage medium |
| CN111611792B (en) * | 2020-05-21 | 2023-05-23 | 全球能源互联网研究院有限公司 | Entity error correction method and system for voice transcription text |
| CN111611792A (en) * | 2020-05-21 | 2020-09-01 | 全球能源互联网研究院有限公司 | Entity error correction method and system for voice transcription text |
| CN111831201A (en) * | 2020-05-25 | 2020-10-27 | 中国人民解放军陆军军医大学第二附属医院 | Human-computer interaction system and method for automatically detecting bone marrow cell morphology |
| CN113763961B (en) * | 2020-06-02 | 2024-04-09 | 阿里巴巴集团控股有限公司 | Text processing method and device |
| CN113763961A (en) * | 2020-06-02 | 2021-12-07 | 阿里巴巴集团控股有限公司 | Text processing method and device |
| CN113743093A (en) * | 2020-06-17 | 2021-12-03 | 北京沃东天骏信息技术有限公司 | Text correction method and device |
| CN112114926A (en) * | 2020-09-25 | 2020-12-22 | 北京百度网讯科技有限公司 | Page operation method, device, equipment and medium based on voice recognition |
| CN112259182A (en) * | 2020-11-05 | 2021-01-22 | 中国联合网络通信集团有限公司 | Method and device for generating electronic medical record |
| CN112259182B (en) * | 2020-11-05 | 2023-08-11 | 中国联合网络通信集团有限公司 | Method and device for generating electronic medical record |
| CN112562668A (en) * | 2020-11-30 | 2021-03-26 | 广州橙行智动汽车科技有限公司 | Semantic information deviation rectifying method and device |
| CN112560493A (en) * | 2020-12-17 | 2021-03-26 | 金蝶软件（中国）有限公司 | Named entity error correction method, named entity error correction device, computer equipment and storage medium |
| CN112560493B (en) * | 2020-12-17 | 2024-04-30 | 金蝶软件（中国）有限公司 | Named entity error correction method, named entity error correction device, named entity error correction computer equipment and named entity error correction storage medium |
| CN112863531A (en) * | 2021-01-12 | 2021-05-28 | 蒋亦韬 | Method for speech audio enhancement by regeneration after computer recognition |
| CN113297348A (en) * | 2021-04-15 | 2021-08-24 | 国网江苏省电力有限公司南京供电分公司 | Correction method for speech recognition Chinese text |
| CN113223509B (en) * | 2021-04-28 | 2022-06-10 | 华南理工大学 | Fuzzy statement identification method and system applied to multi-person mixed scene |
| CN113223509A (en) * | 2021-04-28 | 2021-08-06 | 华南理工大学 | Fuzzy statement identification method and system applied to multi-person mixed scene |
| CN113744722A (en) * | 2021-09-13 | 2021-12-03 | 上海交通大学宁波人工智能研究院 | Off-line speech recognition matching device and method for limited sentence library |
| CN116052657A (en) * | 2022-08-01 | 2023-05-02 | 荣耀终端有限公司 | Character error correction method and device for voice recognition |
| CN116052657B (en) * | 2022-08-01 | 2023-10-20 | 荣耀终端有限公司 | Character error correction method and device for voice recognition |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN109145276A (en) | A kind of text correction method after speech-to-text based on phonetic | |
| CN107315737B (en) | Semantic logic processing method and system | |
| CN109146610B (en) | Intelligent insurance recommendation method and device and intelligent insurance robot equipment | |
| CN105718586B (en) | The method and device of participle | |
| CN109388795B (en) | Named entity recognition method, language recognition method and system | |
| CN105869634B (en) | It is a kind of based on field band feedback speech recognition after text error correction method and system | |
| US6879951B1 (en) | Chinese word segmentation apparatus | |
| US10515292B2 (en) | Joint acoustic and visual processing | |
| CN108984529A (en) | Real-time court's trial speech recognition automatic error correction method, storage medium and computing device | |
| CN111931506B (en) | Entity relationship extraction method based on graph information enhancement | |
| WO2017127296A1 (en) | Analyzing textual data | |
| CN114580382A (en) | Text error correction method and device | |
| CN107564528B (en) | Method and equipment for matching voice recognition text with command word text | |
| Li et al. | Towards zero-shot learning for automatic phonemic transcription | |
| CN110516239B (en) | Segmentation pooling relation extraction method based on convolutional neural network | |
| CN113326702B (en) | Semantic recognition method, semantic recognition device, electronic equipment and storage medium | |
| CN112417823B (en) | Chinese text word order adjustment and word completion method and system | |
| CN110942767B (en) | Recognition labeling and optimization method and device for ASR language model | |
| CN111553157A (en) | Entity replacement-based dialog intention identification method | |
| CN111680524A (en) | Human-machine feedback translation method and system based on reverse matrix analysis | |
| CN114996463A (en) | Intelligent classification method and device for cases | |
| CN113157918B (en) | Commodity name short text classification method and system based on attention mechanism | |
| CN110705295B (en) | Entity name disambiguation method based on keyword extraction | |
| CN109241521B (en) | Scientific literature high-attention sentence extraction method based on citation relation | |
| CN112989839A (en) | Keyword feature-based intent recognition method and system embedded in language model |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| WD01 | Invention patent application deemed withdrawn after publication |
Application publication date: 20190104 |
|
| WD01 | Invention patent application deemed withdrawn after publication |