CN107480118A - Method for editing text and device - Google Patents
Method for editing text and device Download PDFInfo
- Publication number
- CN107480118A CN107480118A CN201710702071.XA CN201710702071A CN107480118A CN 107480118 A CN107480118 A CN 107480118A CN 201710702071 A CN201710702071 A CN 201710702071A CN 107480118 A CN107480118 A CN 107480118A
- Authority
- CN
- China
- Prior art keywords
- groove
- text
- semantic
- edited
- edit commands
- 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
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/21—Design or setup of recognition systems or techniques; Extraction of features in feature space; Blind source separation
- G06F18/214—Generating training patterns; Bootstrap methods, e.g. bagging or boosting
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/24—Classification techniques
- G06F18/241—Classification techniques relating to the classification model, e.g. parametric or non-parametric approaches
-
- 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
- G06F40/284—Lexical analysis, e.g. tokenisation or collocates
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F40/00—Handling natural language data
- G06F40/30—Semantic analysis
Abstract
The invention discloses a kind of method for editing text and device, this method includes：Obtain urtext to be edited；Receive user and edit speech data；Speech data is edited to the user and carries out speech recognition, obtains edit commands；Semantic understanding is carried out to the edit commands, to determine the value of current text to be edited and editing mode；The editing mode is used to describe whether to can determine that the current position of the text to be edited in urtext；Determine whether to perform the edit commands according to the value of the current text to be edited and editing mode；If it is, edlin is entered to the urtext according to the edit commands；Otherwise, with user mutual, with obtain more correspond to it is described currently text to be edited editing mode value.The present invention can improve text editing efficiency, and ensure the accuracy of edit operation.
Description
Technical field
The present invention relates to field of voice signal, and in particular to a kind of method for editing text and device.
Background technology
Text editing is primarily referred to as carrying out word some font formats, typesetting processing, bug patch etc., in prior art
In, usually by manually being completed using keyboard and mouse, first by position of the cursor to text to be edited in urtext
Positioned, edlin then is entered to corresponding text in urtext.Such as text data " today blue whale weather very
Good, I goes that and played several days.", it is necessary to " blue whale " therein is made into " Nanjing ", in addition also by " Nanjing " overstriking, will " my god
Gas " underlines " etc..
This edit methods need manual operation, and when the urtext for needing to edit is more, editorial efficiency is relatively low, and
Easily error, cost of labor are higher.
The content of the invention
The present invention provides a kind of method for editing text and device, with solve efficiency existing for existing text editing it is low, into
The problem of this is high.
Therefore, the present invention provides following technical scheme：
A kind of method for editing text, including：
Obtain urtext to be edited；
Receive user and edit speech data；
Speech data is edited to the user and carries out speech recognition, obtains edit commands；
Semantic understanding is carried out to the edit commands, to determine the value of current text to be edited and editing mode；It is described
Editing mode is used to describe whether to can determine that the current position of the text to be edited in urtext；
Determine whether to perform the edit commands according to the value of the current text to be edited and editing mode；
If it is, edlin is entered to the urtext according to the edit commands；
Otherwise, with user mutual, with obtain more correspond to it is described currently text to be edited editing mode value.
Preferably, methods described also includes：Preset the type of semantic groove；
It is described that semantic understanding is carried out to the edit commands, to determine that current text to be edited and editing mode include：
Semantic understanding is carried out to the edit commands, determines semantic groove and its type that the edit commands includes；
The semantic groove and its type included according to the edit commands determines current text to be edited and editing mode.
Preferably, the type of the semantic groove includes being used to describe the semantic groove of user's editing mode, for describing user
The semantic groove of edit operation and the semantic groove for describing editor's text；
It is described be used for describe user's editing mode semantic groove include it is following any one or more：
Sequential semantics groove, the sequential semantics groove are used to describe order of the text to be edited in urtext；
Type semanteme groove, the type semanteme groove are used for the type for describing text to be edited；
Hanzi structure semanteme groove, the Hanzi structure semanteme groove are used for the Hanzi structure for describing text to be edited；
Radical semanteme groove, the radical semanteme groove are used for the radical for describing text to be edited；
Chinese character description semanteme groove, the Chinese character description semanteme groove are used to be defined text to be edited；
It is described to be used to describe referenced text of the text to be edited in urtext with reference to semantic groove with reference to semantic groove；
Global position semanteme groove, the semantic groove of the overall situation refer to global position of the text to be edited in urtext；
Relative position semanteme groove, the relative position semanteme groove refer to relative position of the text to be edited in urtext；
It is described be used for describe user's edit operation semantic groove include it is following any one or more：Add semantic groove, delete
Except semantic groove, modification semanteme groove, modification font size semanteme groove, modification font color semanteme groove, modification font weight semanteme groove,
Enter a new line semantic groove, retraction semanteme groove；
The semantic groove for being used to describe editor's text includes：Semantic groove after semantic groove to be edited, editor.
Preferably, methods described also includes：
Semantic groove extraction model is built in advance；
It is described that semantic understanding is carried out to the edit commands, determine semantic groove and its type bag that the edit commands includes
Include：
After each word in the edit commands is carried out into vectorization feature is extracted as semantic groove；
The semantic groove extraction feature is inputted into the semantic groove extraction model, according to the defeated of the semantic groove extraction model
Go out to obtain semantic groove and its type that the edit commands includes.
Preferably, semantic groove extraction model is built in the following manner：
Edit commands template is set, the edit commands template includes one or more semantic grooves, and each semantic groove has really
Fixed type；
Collect the filling text for being largely used to fill the semantic groove；
The filling text is filled into the semantic groove, obtains a large amount of edit commands data and as training number
According to, mark each word in the training data whether be semantic groove and semantic groove type；
The training data is segmented, and obtains the term vector of each word；
Trained to obtain semantic groove extraction model according to the term vector and markup information.
Preferably, methods described also includes：
Structure text editing syntax storehouse in advance, the text editing syntax storehouse includes multiple grammatical templates；
It is described that semantic understanding is carried out to the edit commands, determine semantic groove and its type bag that the edit commands includes
Include：
The edit commands is matched with each grammatical template in the text editing syntax storehouse, it is true according to matching result
The semantic groove and its type that the fixed edit commands includes.
Preferably, text editing syntax storehouse is built in the following manner：
The grammatical template in structure basis, the grammatical template in basis include one or more semantic grooves, and each semantic groove has really
Fixed type；
The grammatical template in basis is extended, the grammatical template being expanded；
By the grammatical template in the basis and the grammatical template generation text editing syntax storehouse of the extension.
Preferably, the semantic groove included according to the edit commands and its type determine current text to be edited and volume
The value for the state of collecting includes：
Using the text in the semantic groove to be edited that the edit commands includes as current text to be edited, according to the volume
Collect the text being used to describe in the semantic groove of user's editing mode that order includes and obtain the value of current editing mode.
Preferably, the value according to the current text to be edited and editing mode determines whether to perform the editor
Order includes：
The current text to be edited is subjected to fuzzy matching with the urtext, obtains the text to be edited in institute
State a matched position in urtext；
If a match bit is equipped with one, it is determined that can perform the edit commands；
Otherwise, then the value of current editing mode is added to the current text to be edited and the mould of the urtext
In paste matching, Secondary Match position of the text to be edited in the urtext is obtained；
If the Secondary Match position has one, it is determined that can perform the edit commands；
Otherwise, it determines the edit commands cannot be performed.
A kind of text editing apparatus, including：
Text acquisition module, for obtaining urtext to be edited；
Speech reception module, speech data is edited for receiving user；
Sound identification module, speech recognition is carried out for editing speech data to the user, obtains edit commands；
Semantic understanding module, for carrying out semantic understanding to the edit commands, to determine current text to be edited and volume
The value for the state of collecting；The editing mode is used to describe whether to can determine that the current position of the text to be edited in urtext
Put；
Judge module, for determining whether to perform institute according to the value of the current text to be edited and editing mode
State edit commands；
Editor module, for after the judge module determines that the edit commands can be performed, being ordered according to the editor
Order enters edlin to the urtext；
Interactive module, for after the judge module determines that the edit commands cannot be performed, and user mutual, with
Obtain the value for the editing mode for more corresponding to the current text to be edited.
Preferably, described device also includes：
Semantic groove setup module, for presetting the type of semantic groove；
The semantic understanding module includes：
Semantic groove information determination module, for carrying out semantic understanding to the edit commands, determine the edit commands bag
The semantic groove and its type contained；
Information determination module is edited, the semantic groove and its type for being included according to the edit commands determine currently to wait to compile
Collect text and editing mode.
Preferably, the type of the semantic groove includes being used to describe the semantic groove of user's editing mode, for describing user
The semantic groove of edit operation and the semantic groove for describing editor's text；
It is described be used for describe user's editing mode semantic groove include it is following any one or more：
Sequential semantics groove, the sequential semantics groove are used to describe order of the text to be edited in urtext；
Type semanteme groove, the type semanteme groove are used for the type for describing text to be edited；
Hanzi structure semanteme groove, the Hanzi structure semanteme groove are used for the Hanzi structure for describing text to be edited；
Radical semanteme groove, the radical semanteme groove are used for the radical for describing text to be edited；
Chinese character description semanteme groove, the Chinese character description semanteme groove are used to be defined text to be edited；
It is described to be used to describe referenced text of the text to be edited in urtext with reference to semantic groove with reference to semantic groove；
Global position semanteme groove, the semantic groove of the overall situation refer to global position of the text to be edited in urtext；
Relative position semanteme groove, the relative position semanteme groove refer to relative position of the text to be edited in urtext；
It is described be used for describe user's edit operation semantic groove include it is following any one or more：Add semantic groove, delete
Except semantic groove, modification semanteme groove, modification font size semanteme groove, modification font color semanteme groove, modification font weight semanteme groove,
Enter a new line semantic groove, retraction semanteme groove；
The semantic groove for being used to describe editor's text includes：Semantic groove after semantic groove to be edited, editor.
Preferably, described device also includes：
Semantic groove extraction model builds module, for building semantic groove extraction model in advance；
The semantic groove information determination module includes：
Semantic groove extracts feature generation unit, for after each word in the edit commands is carried out into vectorization as language
Adopted groove extracts feature；
Semantic groove information extraction unit, for the semantic groove extraction feature to be inputted into the semantic groove extraction model, root
Semantic groove and its type that the edit commands includes are obtained according to the output of the semantic groove extraction model.
Preferably, the semantic groove extraction model structure module includes：
Template setup unit, for setting edit commands template, the edit commands template includes one or more semantic
Groove, each semantic groove have the type of determination；
Text collector unit, it is largely used to fill the filling text of the semantic groove for collection；
Semantic groove fills unit, for the filling text to be filled into the semantic groove, obtain a large amount of edit commands
Data and as training data, mark each word in the training data whether be semantic groove and semantic groove type；
Term vector acquiring unit, for being segmented to the training data, and obtain the term vector of each word；
Training unit, for being trained to obtain semantic groove extraction model according to the term vector and markup information.
Preferably, described device also includes：
Text editing syntax storehouse builds module, for building text editing syntax storehouse, the text editing syntax storehouse in advance
Including multiple grammatical templates；
The semantic groove information determination module, specifically for will in the edit commands and the text editing syntax storehouse it is each
Grammatical template is matched, and the semantic groove and its type that the edit commands includes are determined according to matching result.
Preferably, the text editing syntax storehouse structure module includes：
The grammatical template construction unit in basis, for building the grammatical template in basis, the grammatical template in basis comprising one or
Multiple semantic grooves, each semantic groove have the type of determination；
Expanding element, for being extended to the grammatical template in basis, the grammatical template being expanded；
Grammatical storehouse generation unit, for the grammatical template generation text editing by the grammatical template in the basis and the extension
Grammatical storehouse.
Preferably, editor's information determination module, specifically for the semantic groove to be edited for including the edit commands
In text as current text to be edited, the semantic groove for being used to describe user's editing mode included according to the edit commands
In text obtain the value of current editing mode.
Preferably, the judge module includes：
First matching unit, for the current text to be edited to be carried out into fuzzy matching with the urtext, obtain
A matched position of the text to be edited in the urtext；
Judging unit, for when a match bit is equipped with one, it is determined that the edit commands can be performed；It is no
Then, trigger second matching unit and carry out fuzzy matching operation；
Second matching unit, for the value of current editing mode to be added into the current text to be edited and institute
In the fuzzy matching for stating urtext, Secondary Match position of the text to be edited in the urtext is obtained；
The judging unit, when being additionally operable to have one in the Secondary Match position, it is determined that editor's life can be performed
Order；Otherwise, it determines the edit commands cannot be performed.
Method for editing text and device provided in an embodiment of the present invention, speech data is edited to user and carries out speech recognition,
Obtain edit commands；Semantic understanding is carried out to edit commands again and determines current text to be edited and editing mode, according to the volume
The state of collecting determines whether that executive editor orders, if performed, directly enters edlin to urtext；Otherwise, putd question to user,
User's editing mode value is further determined that, and then obtains more accurately positional information of the text to be edited in urtext.
Due to performing relative users edit commands, whole process performs automatically, substantially increases text editing efficiency, and ensure that volume
The accuracy of operation is collected, improves Consumer's Experience.
Brief description of the drawings
, below will be to institute in embodiment in order to illustrate more clearly of the embodiment of the present application or technical scheme of the prior art
The accompanying drawing needed to use is briefly described, it should be apparent that, drawings in the following description are only one described in the present invention
A little embodiments, for those of ordinary skill in the art, other accompanying drawings can also be obtained according to these accompanying drawings.
Fig. 1 is the flow chart of method for editing text of the embodiment of the present invention；
Fig. 2 is the flow chart that semantic groove extraction model is built in the embodiment of the present invention；
Fig. 3 is the flow chart that text editing syntax storehouse is built in the embodiment of the present invention；
Fig. 4 is the structural representation of text editing apparatus of the embodiment of the present invention；
Fig. 5 is a kind of structural representation of the semantic groove information determination module based on model in the embodiment of the present invention；
Fig. 6 is a kind of structural representation of the semantic groove information determination module based on the syntax in the embodiment of the present invention.
Embodiment
In order that those skilled in the art more fully understand the scheme of the embodiment of the present invention, below in conjunction with the accompanying drawings and implement
Mode is described in further detail to the embodiment of the present invention.
In view of the convenience and high efficiency of interactive voice, enter edlin to text data using voice and be undoubtedly quick height
The mode of effect, therefore, for prior art by manually being asked using efficiency is low existing for keyboard and mouse progress text editing
Topic, the embodiment of the present invention propose a kind of method for editing text and device, realize the automatic editor to text by voice, improve
Text editing efficiency；Moreover, in the case where position of the text to be edited in urtext accurately can be determined, perform to original
Text enters edlin, otherwise by further interacting, obtains more accurately position letter of the text to be edited in urtext
Breath, ensure the accuracy of edit operation.
As shown in figure 1, being the flow chart of method for editing text of the embodiment of the present invention, comprise the following steps：
Step 101, urtext to be edited is obtained.
Wherein, the urtext to be edited can input information according to user speech or input information of taking pictures obtains；
Information is inputted for user speech, text can be identified as using speech recognition technology, using the text as to be edited
Urtext；For user take pictures input information, corresponding text can be obtained using image recognition technology, using the text as
Urtext to be edited.Certainly, the text data that the urtext to be edited can also directly be provided by user.It is described
Speech recognition and image recognition can use some existing correlation techniques, and this embodiment of the present invention is not limited.
Step 102, receive user and edit speech data.
The user edits speech data and refers to data corresponding to the voice said when user changes text data, such as waits to compile
The urtext collected is " helping me to order opening and closing fertilizer to the train ticket in Nanjing ", and user wants for " Nanjing " therein to be revised as " north
Capital ", it is " Nanjing is revised as into Beijing " that corresponding user, which edits speech data,.
Step 103, speech data is edited to the user and carries out speech recognition, obtains edit commands.
The edit commands is the identification text that speech recognition obtains.The speech recognition can use related existing skill
Art, it will not be described in detail herein.
Step 104, semantic understanding is carried out to the edit commands, to determine taking for current text to be edited and editing mode
Value；The editing mode is used to describe whether to can determine that the current position of the text to be edited in urtext.
Specifically, semantic understanding is carried out to the edit commands first, determine semantic groove that the edit commands includes and
Its type；Then the semantic groove and its type included according to the edit commands determines current text to be edited and editing mode.
The type of the semantic groove includes being used to describe the semantic groove of user's editing mode, for describing user's edit operation
Semantic groove and for describe editor text semantic groove.The semantic groove of this several types is carried out specifically separately below
It is bright.
1) be used for describe user's editing mode semantic groove include it is following any one or more：
Sequential semantics groove, the sequential semantics groove are used to describe order of the text to be edited in urtext, and such as first
It is individual, second, last, penultimate etc.；
Type semanteme groove, the type semanteme groove are used for the type for describing text to be edited, such as Chinese character, Chinese, English, number
Word, Chinese capitalization, Arabic numerals, numeric string etc.；
Hanzi structure semanteme groove, the Hanzi structure semanteme groove are used for the Hanzi structure for describing text to be edited, such as length of bow
, Wen Daoliu, ear east is old, three horizontal strokes, one perpendicular king etc.；
Radical semanteme groove, the radical semanteme groove is used for the radical for describing text to be edited, such as one
Side, by vertical cutter, by king's word, precious headkerchief, by word etc.；
Chinese character description semanteme groove, the Chinese character description semanteme groove are used to be defined text to be edited, such as we I,
The section, Chinese state etc. of University of Science and Technology；
It is described to be used to describe referenced text of the text to be edited in urtext with reference to semantic groove with reference to semantic groove, such as
I wants to go to Nanjing below, wherein, " I wants to go to " is referenced text；
Global position semanteme groove, the semantic groove of the overall situation refer to global position of the text to be edited in urtext, such as treated
Edit chapter of the text in urtext, paragraph, sentence etc.；Such as the Nanjing in the 3rd section second；
Relative position semanteme groove, the relative position semanteme groove refer to relative position of the text to be edited in urtext,
As I will go to Nanjing below, behind be relative position semanteme groove；
2) it is used for the semantic groove for describing user's edit operation, such semantic groove is used for the specific edit operation for describing user,
Such as can include it is following any one or more：Add semantic groove, deletion semanteme groove, the semantic groove of modification, modification font size
Semantic groove, modification font color semanteme groove, modification font weight semanteme groove, the semantic groove of line feed, retraction semanteme groove；
3) being used to describe the semantic groove of editor's text includes：Semantic groove after semantic groove to be edited, editor.Wherein, it is described to treat
It is text to be edited to edit semantic groove Chinese version, and semantic groove Chinese version is text after editor after editor.
It is determined that when semantic groove and its type that the edit commands includes, method or use based on model can be used
Based on the method for the syntax, detailed process will be described in detail later.
After semantic groove that the edit commands includes and its type is obtained, directly it can be included according to the edit commands
Semantic groove and its type determine current text to be edited and editing mode.Specifically, what the edit commands included is to be edited
Text in semantic groove is current text to be edited；According to the edit commands include be used for user's editing mode is described
Text in semantic groove, obtain the value of current editing mode.
Such as user's edit commands " to go to Nanjing to be below revised as Beijing me ", the semantic groove type extracted for "/
O I/RefText will go/RefText behind/RelativePos Nanjing/OriText modifications/O is /O Beijing/Text ", its
Middle O represents non-semantic groove, and OriText is semantic groove to be edited, and RefText is to refer to semantic groove, and Text is semantic groove after editor；
Current text to be edited is " Nanjing ", and the value of editing mode is：
Sequential semantics groove：Null
Type semanteme groove：Null
Hanzi structure semanteme groove：Null
Radical semanteme groove：Null
Chinese character description semanteme groove：Null
With reference to semantic groove：I will go
Global position semanteme groove：Null
Relative position semanteme groove：Below
Step 105, determine whether to perform the volume according to the value of the current text to be edited and editing mode
Collect order；If it is, perform step 106；Otherwise, step 107 is performed.
Specifically, pre-editing text will be worked as and carries out fuzzy matching with the urtext, and obtained the text to be edited and exist
A matched position in the urtext, a match bit are equipped with one or more.
If one time matched position only has one, it is determined that current position of the text to be edited in urtext is described
Matched position；Because the position of current text to be edited can accurately determine, therefore in such a case, it is possible to perform use
Family edit commands, i.e., edlin is entered to the urtext according to the edit commands.
If a match bit is equipped with multiple, text to be edited further can be determined according to the value of current editing mode
Particular location in urtext.Specifically, the value of current editing mode directly can be added to current text to be edited
This with the fuzzy matching of urtext, if the Secondary Match position matched only has one, it is determined that can perform described
Edit commands；Otherwise need to interact with user, it is current to be edited until obtaining to obtain the value of more editing modes
Exact position of the text in urtext.
Step 106, edlin is entered to the urtext according to the edit commands, is then back to step 102.
If user's edit commands is " to go to Nanjing to be below revised as Beijing me ", it is determined that text to be edited is " south
After the position in " Nanjing " edited is needed in capital ", and urtext, " north directly is revised as in " Nanjing " on the position
Capital "；Concrete modification method can use existing correlation technique, will not be described in detail herein.
Step 107, with user mutual, with obtain more correspond to it is described currently text to be edited editing mode value.
, can be according to the value condition of each semantic groove and addition in active user's editing mode when being interacted with user
Fuzzy matching result in user's editing mode after semantic groove value is putd question to user, such as " I will be gone to Nanjing below
Be revised as Beijing " the Secondary Match result that adds after the value of current editing mode still have it is multiple, as first in urtext
Occur " Nanjing " in section, occur " Nanjing ", " blue whale ", " hardly possible warp " in the 3rd section；" the south of which section such as can be putd question to user
The problems such as capital ", " being the provincial capital in Jiangsu ", continue to obtain the value of semantic groove in more editing modes；Be tod after puing question to every time
To user's editing mode semanteme groove value be added in the matching of text to be edited and urtext, until obtaining text to be edited
Originally the exact position in urtext.
It should be noted that when user enters edlin to text to be edited, general interaction times can not be too many, otherwise, uses
Family experience is poor, so in order to lift Consumer's Experience, can control interaction times within twice；It is of course also possible to carry out
Repeatedly interaction, to obtain more accurately information, determine user view.
, can it is determined that when semantic groove and its type that the edit commands includes it is previously noted that in embodiments of the present invention
Using the method based on model or using the method based on the syntax, both approaches are described in detail separately below.
(1) method based on model determines the semantic groove and its type that edit commands includes
In the method, it is necessary to build semantic groove extraction model in advance, as shown in Fig. 2 being built in the embodiment of the present invention
The flow chart of semantic groove extraction model, comprises the following steps：
Step 21, edit commands template is set, the edit commands template includes one or more semantic grooves, each semantic
Groove has the type of determination.
As " it is edit commands template that A, which is made into B ", and wherein A and B are semantic groove, and specifically, semantic groove A type is to treat
Edit semantic groove, semantic groove B type is semantic groove after editor.
Step 22, the filling text for being largely used to fill the semantic groove is collected.
Step 23, the filling text is filled into the semantic groove, obtains a large amount of edit commands data and made
For training data, mark each word in the training data whether be semantic groove and semantic groove type.
It should be noted that when carrying out information labeling, if the word in edit commands data is not semantic groove, can make
Represented with unified symbol, such as represent non-semantic groove using character O.
Step 24, the training data is segmented, and obtains the term vector of each word.
Step 25, trained to obtain semantic groove extraction model according to the term vector and markup information.
, can be using the term vector of each word in user's edit commands as mode input, to edit commands during specific training
Whether middle word is that the type of semantic groove and semantic groove is predicted, and utilizes the prediction result and mark of every edit commands semanteme groove
Information is updated to model parameter, after training terminates, obtains semantic groove extraction model.
Based on above-mentioned semantic groove extraction model, it is determined that when semantic groove and its type that the edit commands includes, it is necessary to
After each word progress vectorization in the edit commands that above-mentioned steps 103 are obtained feature is extracted as semantic groove；Then by institute
Predicate justice groove extraction feature inputs the semantic groove extraction model, and the volume is obtained according to the output of the semantic groove extraction model
Collect semantic groove and its type that order includes.If active user's edit commands is " Beijing is revised as in Nanjing ", the semanteme is utilized
Groove extraction model, obtained semantic groove extraction result are：" it is /O Beijing/Text ", wherein O /O Nanjing/OriText modifications/O
Non-semantic groove is represented, oriText is semantic groove, and type is semantic groove to be edited, and Text is semantic groove, and type is semantic after editor
Groove.
(2) method based on the syntax determines the semantic groove and its type that edit commands includes
In the method, it is necessary to text editing syntax storehouse be built in advance, as shown in figure 3, being built in the embodiment of the present invention
The flow chart in text editing syntax storehouse, comprises the following steps：
Step 31, the grammatical template in structure basis, the grammatical template in basis includes one or more semantic grooves, each semantic
Groove has the type of determination.
Step 32, the grammatical template in basis is extended, the grammatical template being expanded.
As the grammatical template in basis is：A is revised as B, can be expanded to " A be modified as B, exchange A for B, I thinks handle
A makes B, A into and makes B " etc. into, so as to the grammatical template being expanded, semantic groove in grammatical template and its type after extension with
Semantic groove and its type in the corresponding grammatical template in basis is identical.
Step 33, by the grammatical template in the basis and the grammatical template generation text editing syntax storehouse of the extension.
, can be with it is determined that when semantic groove and its type that the edit commands includes based on above-mentioned text editing syntax storehouse
The edit commands that above-mentioned steps 103 obtain is matched with each grammatical template in the text editing syntax storehouse, according to matching
As a result the semantic groove and its type that the edit commands includes be can determine that.Specific matching process can use existing related skill
Art, the method that can such as use accurate matching or fuzzy matching, will not be described in detail herein.
Method for editing text provided in an embodiment of the present invention, speech data is edited to user and carries out speech recognition, is compiled
Collect order；Semantic understanding is carried out to edit commands again and determines current text to be edited and editing mode, according to the editing mode
Determine whether that executive editor orders, if performed, edlin directly is entered to urtext；Otherwise, putd question to user, further
User's editing mode value is determined, and then obtains more accurately positional information of the text to be edited in urtext.Due to holding
Row relative users edit commands, whole process perform automatically, substantially increase text editing efficiency, and ensure that edit operation
Accuracy, improve Consumer's Experience.
Correspondingly, the embodiment of the present invention also provides a kind of text editing apparatus, as shown in figure 4, being that the structure of the device is shown
It is intended to.
In this embodiment, described device includes：
Text acquisition module 401, for obtaining urtext to be edited；
Speech reception module 402, speech data is edited for receiving user；
Sound identification module 403, speech recognition is carried out for editing speech data to the user, obtains edit commands；
Semantic understanding module 404, for carrying out semantic understanding to the edit commands, with determine current text to be edited and
The value of editing mode；Whether the editing mode is used to describe to can determine that the current text to be edited in urtext
Position；
Judge module 405, for determining whether to hold according to the value of the current text to be edited and editing mode
The row edit commands；
Editor module 406, for the judge module determine can perform the edit commands after, according to the editor
Order enters edlin to the urtext；
Interactive module 407, for after the judge module determines that the edit commands cannot be performed, being handed over user
Mutually, the value of the editing mode of the current text to be edited is more corresponded to acquisition.
Above-mentioned speech reception module 402 can specifically be edited speech data to the user using prior art and carry out voice
Identification, is identified text, the identification text is the edit commands of user.
Above-mentioned semantic understanding module 404 to the semantic understanding of the edit commands, be mainly based upon pre-set it is various
Semantic groove and its type, to determine current text to be edited and editing mode.The type of semantic groove can be preset at described device
In, certainly, for the ease of operation, corresponding semantic groove setup module (not shown) can also be set, with pre- in said device
First set the type of semantic groove.
In embodiments of the present invention, the type of the semantic groove includes being used for semantic groove, the use for describing user's editing mode
In description user's edit operation semantic groove and for describe editor text semantic groove, all types of semantic grooves it is specific
Implication will not be repeated here described in detail above.
Correspondingly, as shown in figure 4, the semantic understanding module 404 can include semantic groove information determination module 441 and compile
Collect information determination module 442.Wherein, semantic groove information determination module 441 is used to carry out semantic understanding to the edit commands, really
The semantic groove and its type that the fixed edit commands includes；Information determination module 442 is edited to be used for according to the edit commands bag
The semantic groove and its type contained determines current text to be edited and editing mode.
Above-mentioned semantic groove information determination module 441, can it is determined that when semantic groove and its type that the edit commands includes
Using the method based on model or using the method based on the syntax, this is further explained below.
As shown in figure 5, show semantic when the semantic groove and its type that the method based on model determines that edit commands includes
A kind of structural representation of groove information determination module.
In this embodiment, semantic groove extraction model 50 is built in advance by semantic groove extraction model structure module 51, the language
Adopted groove extraction model structure module 51 can be as a part for the device of the embodiment of the present invention or independently of the device
Outside, this embodiment of the present invention is not limited.
Correspondingly, in this embodiment, semantic groove information determination module 441 includes：
Semantic groove extraction feature generation unit 52, for after each word in the edit commands is carried out into vectorization as
Semantic groove extracts feature；
Semantic groove information extraction unit 53, for the semantic groove extraction feature to be inputted into the semantic groove extraction model,
Semantic groove and its type that the edit commands includes are obtained according to the output of the semantic groove extraction model.
A kind of concrete structure of above-mentioned semantic groove extraction model structure module 51 includes following each unit：
Template setup unit, for setting edit commands template, the edit commands template includes one or more semantic
Groove, each semantic groove have the type of determination；
Text collector unit, it is largely used to fill the filling text of the semantic groove for collection；
Semantic groove fills unit, for the filling text to be filled into the semantic groove, obtain a large amount of edit commands
Data and as training data, mark each word in the training data whether be semantic groove and semantic groove type；
Term vector acquiring unit, for being segmented to the training data, and obtain the term vector of each word；
Training unit, for being trained to obtain semantic groove extraction model according to the term vector and markup information.
As shown in fig. 6, show semantic when the semantic groove and its type that the method based on the syntax determines that edit commands includes
A kind of structural representation of groove information determination module.
In this embodiment, text editing syntax storehouse 60 is built in advance by text editing syntax storehouse structure module 61, this article
The grammatical storehouse structure module 61 of this editor can be as a part for the device of the embodiment of the present invention or independently of the device
Outside, this embodiment of the present invention is not limited.Text editing syntax storehouse 60 includes multiple grammatical templates.
Correspondingly, in this embodiment, semantic groove information determination module 441 is by the edit commands and the text editing
Each grammatical template is matched in grammatical storehouse, and the semantic groove and its type that the edit commands includes are determined according to matching result.
A kind of concrete structure of above-mentioned text editing syntax storehouse structure module 61 includes following each unit：
The grammatical template construction unit in basis, for building the grammatical template in basis, the grammatical template in basis comprising one or
Multiple semantic grooves, each semantic groove have the type of determination；
Expanding element, for being extended to the grammatical template in basis, the grammatical template being expanded；
Grammatical storehouse generation unit, for the grammatical template generation text editing by the grammatical template in the basis and the extension
Grammatical storehouse.
Semantic groove information determination module 441 no matter based on which kind of above-mentioned mode obtain semantic groove that edit commands includes and its
After type,
Editor's information determination module 442 can be determined according to the semantic groove and its type current text to be edited and
The value of editing mode, specifically, using the text in the semantic groove to be edited that the edit commands includes as current to be edited
Text, the text being used to describe in the semantic groove of user's editing mode included according to the edit commands obtain working as pre-editing shape
The value of state.
Judge module in Fig. 4 can specifically include following each unit：
First matching unit, for the current text to be edited to be carried out into fuzzy matching with the urtext, obtain
A matched position of the text to be edited in the urtext；
Judging unit, for when a match bit is equipped with one, it is determined that the edit commands can be performed, now
The instruction that triggering editor module 406 starts editor can be exported；Otherwise, trigger second matching unit and carry out fuzzy matching behaviour
Make；
Second matching unit, for the value of current editing mode to be added into the current text to be edited and institute
In the fuzzy matching for stating urtext, Secondary Match position of the text to be edited in the urtext is obtained；
The judging unit, when being additionally operable to have one in the Secondary Match position, it is determined that editor's life can be performed
Order, it can now export the instruction that triggering editor module 406 starts editor；Otherwise, it determines the edit commands cannot be performed,
The instruction that triggering interactive module 407 interacts can now be exported.
The concrete operations that the editor module 406 enters edlin according to the edit commands to the urtext can adopt
Use prior art.
, can be according to each semantic groove in active user's editing mode when the interactive module 407 interacts with user
Value condition and the fuzzy matching result that adds in user's editing mode after semantic groove value putd question to user.Carry every time
Obtained user's editing mode semanteme groove value is added in the matching of text to be edited and urtext after asking, until obtaining
Exact position of the text to be edited in urtext.
Text editing apparatus provided in an embodiment of the present invention, speech data is edited to user and carries out speech recognition, is compiled
Collect order；Semantic understanding is carried out to edit commands again and determines current text to be edited and editing mode, according to the editing mode
Determine whether that executive editor orders, if performed, edlin directly is entered to urtext；Otherwise, putd question to user, further
User's editing mode value is determined, and then obtains more accurately positional information of the text to be edited in urtext.Due to holding
Row relative users edit commands, whole process perform automatically, substantially increase text editing efficiency, and ensure that edit operation
Accuracy, improve Consumer's Experience.
Each embodiment in this specification is described by the way of progressive, identical similar portion between each embodiment
Divide mutually referring to what each embodiment stressed is the difference with other embodiment.It is real especially for device
For applying example, because it is substantially similar to embodiment of the method, so describing fairly simple, related part is referring to embodiment of the method
Part explanation.Device embodiment described above is only schematical, wherein described be used as separating component explanation
Unit can be or may not be physically separate, can be as the part that unit is shown or may not be
Physical location, you can with positioned at a place, or can also be distributed on multiple NEs.Can be according to the actual needs
Some or all of module therein is selected to realize the purpose of this embodiment scheme.Those of ordinary skill in the art are not paying
In the case of creative work, you can to understand and implement.
The embodiment of the present invention is described in detail above, embodiment used herein is carried out to the present invention
Illustrate, the explanation of above example is only intended to help to understand method and system of the invention；Meanwhile for the one of this area
As technical staff, according to the thought of the present invention, there will be changes in specific embodiments and applications, to sum up institute
State, this specification content should not be construed as limiting the invention.
Claims (18)
- A kind of 1. method for editing text, it is characterised in that including：Obtain urtext to be edited；Receive user and edit speech data；Speech data is edited to the user and carries out speech recognition, obtains edit commands；Semantic understanding is carried out to the edit commands, to determine the value of current text to be edited and editing mode；The editor State is used to describe whether to can determine that the current position of the text to be edited in urtext；Determine whether to perform the edit commands according to the value of the current text to be edited and editing mode；If it is, edlin is entered to the urtext according to the edit commands；Otherwise, with user mutual, with obtain more correspond to it is described currently text to be edited editing mode value.
- 2. according to the method for claim 1, it is characterised in that methods described also includes：Preset the type of semantic groove；It is described that semantic understanding is carried out to the edit commands, to determine that current text to be edited and editing mode include：Semantic understanding is carried out to the edit commands, determines semantic groove and its type that the edit commands includes；The semantic groove and its type included according to the edit commands determines current text to be edited and editing mode.
- 3. according to the method for claim 2, it is characterised in that the type of the semantic groove includes editing for describing user The semantic groove of state, the semantic groove for describing user's edit operation and the semantic groove for describing editor's text；It is described be used for describe user's editing mode semantic groove include it is following any one or more：Sequential semantics groove, the sequential semantics groove are used to describe order of the text to be edited in urtext；Type semanteme groove, the type semanteme groove are used for the type for describing text to be edited；Hanzi structure semanteme groove, the Hanzi structure semanteme groove are used for the Hanzi structure for describing text to be edited；Radical semanteme groove, the radical semanteme groove are used for the radical for describing text to be edited；Chinese character description semanteme groove, the Chinese character description semanteme groove are used to be defined text to be edited；It is described to be used to describe referenced text of the text to be edited in urtext with reference to semantic groove with reference to semantic groove；Global position semanteme groove, the semantic groove of the overall situation refer to global position of the text to be edited in urtext；Relative position semanteme groove, the relative position semanteme groove refer to relative position of the text to be edited in urtext；It is described be used for describe user's edit operation semantic groove include it is following any one or more：Add semantic groove, delete language Adopted groove, modification semanteme groove, modification font size semanteme groove, modification font color semanteme groove, modification font weight semanteme groove, line feed Semantic groove, retraction semanteme groove；The semantic groove for being used to describe editor's text includes：Semantic groove after semantic groove to be edited, editor.
- 4. according to the method for claim 2, it is characterised in that methods described also includes：Semantic groove extraction model is built in advance；It is described that semantic understanding is carried out to the edit commands, determine that semantic groove that the edit commands includes and its type include：After each word in the edit commands is carried out into vectorization feature is extracted as semantic groove；The semantic groove extraction feature is inputted into the semantic groove extraction model, according to exporting for the semantic groove extraction model The semantic groove and its type included to the edit commands.
- 5. according to the method for claim 4, it is characterised in that build semantic groove extraction model in the following manner：Edit commands template is set, the edit commands template includes one or more semantic grooves, and each semantic groove has determination Type；Collect the filling text for being largely used to fill the semantic groove；The filling text is filled into the semantic groove, obtains a large amount of edit commands data and as training data, Mark each word in the training data whether be semantic groove and semantic groove type；The training data is segmented, and obtains the term vector of each word；Trained to obtain semantic groove extraction model according to the term vector and markup information.
- 6. according to the method for claim 2, it is characterised in that methods described also includes：Structure text editing syntax storehouse in advance, the text editing syntax storehouse includes multiple grammatical templates；It is described that semantic understanding is carried out to the edit commands, determine that semantic groove that the edit commands includes and its type include：The edit commands is matched with each grammatical template in the text editing syntax storehouse, institute is determined according to matching result State semantic groove and its type that edit commands includes.
- 7. according to the method for claim 6, it is characterised in that build text editing syntax storehouse in the following manner：The grammatical template in structure basis, the grammatical template in basis include one or more semantic grooves, and each semantic groove has determination Type；The grammatical template in basis is extended, the grammatical template being expanded；By the grammatical template in the basis and the grammatical template generation text editing syntax storehouse of the extension.
- 8. according to the method for claim 3, it is characterised in that the semantic groove included according to the edit commands and its Type determines that the value of current text to be edited and editing mode includes：Using the text in the semantic groove to be edited that the edit commands includes as current text to be edited, ordered according to the editor The text being used to describe in the semantic groove of user's editing mode included is made to obtain the value of current editing mode.
- 9. according to the method described in any one of claim 1 to 8, it is characterised in that described according to the current text to be edited And the value of editing mode determines whether that performing the edit commands includes：The current text to be edited is subjected to fuzzy matching with the urtext, obtains the text to be edited in the original A matched position in beginning text；If a match bit is equipped with one, it is determined that can perform the edit commands；Otherwise, then the value of current editing mode is added to fuzzy of the current text to be edited and the urtext In matching somebody with somebody, Secondary Match position of the text to be edited in the urtext is obtained；If the Secondary Match position has one, it is determined that can perform the edit commands；Otherwise, it determines the edit commands cannot be performed.
- A kind of 10. text editing apparatus, it is characterised in that including：Text acquisition module, for obtaining urtext to be edited；Speech reception module, speech data is edited for receiving user；Sound identification module, speech recognition is carried out for editing speech data to the user, obtains edit commands；Semantic understanding module, for carrying out semantic understanding to the edit commands, to determine current text to be edited and editor's shape The value of state；The editing mode is used to describe whether to can determine that the current position of the text to be edited in urtext；Judge module, for determining whether to perform the volume according to the value of the current text to be edited and editing mode Collect order；Editor module, for the judge module determine can perform the edit commands after, according to the edit commands pair The urtext enters edlin；Interactive module, for after the judge module determines that the edit commands cannot be performed, and user mutual, to obtain More correspond to the value of the editing mode of the current text to be edited.
- 11. device according to claim 10, it is characterised in that described device also includes：Semantic groove setup module, for presetting the type of semantic groove；The semantic understanding module includes：Semantic groove information determination module, for carrying out semantic understanding to the edit commands, determine what the edit commands included Semantic groove and its type；Information determination module is edited, the semantic groove and its type for being included according to the edit commands determine current text to be edited Sheet and editing mode.
- 12. device according to claim 11, it is characterised in that the type of the semantic groove includes being used to describe user's volume Semantic groove, the semantic groove for describing user's edit operation and the semantic groove for describing editor's text for the state of collecting；It is described be used for describe user's editing mode semantic groove include it is following any one or more：Sequential semantics groove, the sequential semantics groove are used to describe order of the text to be edited in urtext；Type semanteme groove, the type semanteme groove are used for the type for describing text to be edited；Hanzi structure semanteme groove, the Hanzi structure semanteme groove are used for the Hanzi structure for describing text to be edited；Radical semanteme groove, the radical semanteme groove are used for the radical for describing text to be edited；Chinese character description semanteme groove, the Chinese character description semanteme groove are used to be defined text to be edited；It is described to be used to describe referenced text of the text to be edited in urtext with reference to semantic groove with reference to semantic groove；Global position semanteme groove, the semantic groove of the overall situation refer to global position of the text to be edited in urtext；Relative position semanteme groove, the relative position semanteme groove refer to relative position of the text to be edited in urtext；It is described be used for describe user's edit operation semantic groove include it is following any one or more：Add semantic groove, delete language Adopted groove, modification semanteme groove, modification font size semanteme groove, modification font color semanteme groove, modification font weight semanteme groove, line feed Semantic groove, retraction semanteme groove；The semantic groove for being used to describe editor's text includes：Semantic groove after semantic groove to be edited, editor.
- 13. device according to claim 11, it is characterised in that described device also includes：Semantic groove extraction model builds module, for building semantic groove extraction model in advance；The semantic groove information determination module includes：Semantic groove extracts feature generation unit, for after each word in the edit commands is carried out into vectorization as semantic groove Extract feature；Semantic groove information extraction unit, for the semantic groove extraction feature to be inputted into the semantic groove extraction model, according to institute The output of predicate justice groove extraction model obtains semantic groove and its type that the edit commands includes.
- 14. device according to claim 13, it is characterised in that the semantic groove extraction model structure module includes：Template setup unit, for setting edit commands template, the edit commands template includes one or more semantic grooves, often Individual semantic groove has the type of determination；Text collector unit, it is largely used to fill the filling text of the semantic groove for collection；Semantic groove fills unit, for the filling text to be filled into the semantic groove, obtain a large amount of edit commands data And as training data, mark each word in the training data whether be semantic groove and semantic groove type；Term vector acquiring unit, for being segmented to the training data, and obtain the term vector of each word；Training unit, for being trained to obtain semantic groove extraction model according to the term vector and markup information.
- 15. device according to claim 11, it is characterised in that described device also includes：Text editing syntax storehouse builds module, and for building text editing syntax storehouse in advance, the text editing syntax storehouse includes Multiple grammatical templates；The semantic groove information determination module, specifically for by each syntax in the edit commands and the text editing syntax storehouse Template is matched, and the semantic groove and its type that the edit commands includes are determined according to matching result.
- 16. device according to claim 15, it is characterised in that the text editing syntax storehouse structure module includes：The grammatical template construction unit in basis, for building the grammatical template in basis, the grammatical template in basis includes one or more Semantic groove, each semantic groove have the type of determination；Expanding element, for being extended to the grammatical template in basis, the grammatical template being expanded；Grammatical storehouse generation unit, for the grammatical template generation text editing syntax by the grammatical template in the basis and the extension Storehouse.
- 17. device according to claim 12, it is characterised in thatEditor's information determination module, specifically for using the text in the semantic groove to be edited that the edit commands includes as Current text to be edited, the text being used to describe in the semantic groove of user's editing mode included according to the edit commands obtain The value of current editing mode.
- 18. according to the device described in any one of claim 10 to 17, it is characterised in that the judge module includes：First matching unit, for the current text to be edited to be carried out into fuzzy matching with the urtext, obtain described A matched position of the text to be edited in the urtext；Judging unit, for when a match bit is equipped with one, it is determined that the edit commands can be performed；Otherwise, touch Send out the second matching unit described and carry out fuzzy matching operation；Second matching unit, for the value of current editing mode to be added into the current text to be edited and the original In the fuzzy matching of beginning text, Secondary Match position of the text to be edited in the urtext is obtained；The judging unit, when being additionally operable to have one in the Secondary Match position, it is determined that the edit commands can be performed；It is no Then, it is determined that the edit commands cannot be performed.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201710702071.XA CN107480118A (en) | 2017-08-16 | 2017-08-16 | Method for editing text and device |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201710702071.XA CN107480118A (en) | 2017-08-16 | 2017-08-16 | Method for editing text and device |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN107480118A true CN107480118A (en) | 2017-12-15 |
Family
ID=60600480
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201710702071.XA Pending CN107480118A (en) | 2017-08-16 | 2017-08-16 | Method for editing text and device |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN107480118A (en) |
Cited By (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109087645A (en) * | 2018-10-24 | 2018-12-25 | 科大讯飞股份有限公司 | A kind of decoding network generation method, device, equipment and readable storage medium storing program for executing |
| CN109388792A (en) * | 2018-09-30 | 2019-02-26 | 珠海格力电器股份有限公司 | Text handling method, device, equipment, computer equipment and storage medium |
| CN110321534A (en) * | 2018-03-28 | 2019-10-11 | 科大讯飞股份有限公司 | A kind of method for editing text, device, equipment and readable storage medium storing program for executing |
| CN111046674A (en) * | 2019-12-20 | 2020-04-21 | 科大讯飞股份有限公司 | Semantic understanding method and device, electronic equipment and storage medium |
| CN113378520A (en) * | 2021-04-20 | 2021-09-10 | 北京灵伴即时智能科技有限公司 | Text editing method and system |
| CN113723050A (en) * | 2021-07-23 | 2021-11-30 | 内蒙古工业大学 | Word document processing system and method based on voice interaction |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103106061A (en) * | 2013-03-05 | 2013-05-15 | 北京车音网科技有限公司 | Voice input method and device |
| US20130311179A1 (en) * | 2010-09-27 | 2013-11-21 | Apple Inc. | Electronic Device with Text Error Correction Based on Voice Recognition Data |
| CN105869632A (en) * | 2015-01-22 | 2016-08-17 | 北京三星通信技术研究有限公司 | Speech recognition-based text revision method and device |
| CN106534548A (en) * | 2016-11-17 | 2017-03-22 | 科大讯飞股份有限公司 | Voice error correction method and device |
-
2017
- 2017-08-16 CN CN201710702071.XA patent/CN107480118A/en active Pending
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20130311179A1 (en) * | 2010-09-27 | 2013-11-21 | Apple Inc. | Electronic Device with Text Error Correction Based on Voice Recognition Data |
| CN103106061A (en) * | 2013-03-05 | 2013-05-15 | 北京车音网科技有限公司 | Voice input method and device |
| CN105869632A (en) * | 2015-01-22 | 2016-08-17 | 北京三星通信技术研究有限公司 | Speech recognition-based text revision method and device |
| CN106534548A (en) * | 2016-11-17 | 2017-03-22 | 科大讯飞股份有限公司 | Voice error correction method and device |
Cited By (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110321534A (en) * | 2018-03-28 | 2019-10-11 | 科大讯飞股份有限公司 | A kind of method for editing text, device, equipment and readable storage medium storing program for executing |
| CN110321534B (en) * | 2018-03-28 | 2023-11-24 | 科大讯飞股份有限公司 | Text editing method, device, equipment and readable storage medium |
| CN109388792A (en) * | 2018-09-30 | 2019-02-26 | 珠海格力电器股份有限公司 | Text handling method, device, equipment, computer equipment and storage medium |
| CN109087645A (en) * | 2018-10-24 | 2018-12-25 | 科大讯飞股份有限公司 | A kind of decoding network generation method, device, equipment and readable storage medium storing program for executing |
| CN111046674A (en) * | 2019-12-20 | 2020-04-21 | 科大讯飞股份有限公司 | Semantic understanding method and device, electronic equipment and storage medium |
| CN113378520A (en) * | 2021-04-20 | 2021-09-10 | 北京灵伴即时智能科技有限公司 | Text editing method and system |
| CN113378520B (en) * | 2021-04-20 | 2024-04-09 | 北京灵伴即时智能科技有限公司 | Text editing method and system |
| CN113723050A (en) * | 2021-07-23 | 2021-11-30 | 内蒙古工业大学 | Word document processing system and method based on voice interaction |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN107480118A (en) | Method for editing text and device | |
| CN103810998B (en) | Based on the off-line audio recognition method of mobile terminal device and realize method | |
| CN109753636A (en) | Machine processing and text error correction method and device calculate equipment and storage medium | |
| CN104346319B (en) | Method and system for inspecting document style | |
| CN105975625A (en) | Chinglish inquiring correcting method and system oriented to English search engine | |
| CN108664474A (en) | A kind of resume analytic method based on deep learning | |
| CN105632484A (en) | Voice synthesis database pause information automatic marking method and system | |
| CN107885844A (en) | Automatic question-answering method and system based on systematic searching | |
| CN105701089A (en) | Post-editing processing method for correction of wrong words in machine translation | |
| CN105261358A (en) | N-gram grammar model constructing method for voice identification and voice identification system | |
| CN108090400A (en) | A kind of method and apparatus of image text identification | |
| CN108052499A (en) | Text error correction method, device and computer-readable medium based on artificial intelligence | |
| CN110287482A (en) | Semi-automation participle corpus labeling training device | |
| CN103235775B (en) | A kind of statistical machine translation method merging translation memory and phrase translation model | |
| CN105068990B (en) | A kind of English long sentence dividing method of more strategies of Machine oriented translation | |
| CN102915122B (en) | Based on the intelligent family moving platform spelling input method of language model | |
| CN109634935A (en) | Method of speech processing, storage medium and device | |
| CN107992476B (en) | Corpus generation method and system for sentence-level biological relation network extraction | |
| CN105045583B (en) | A kind of IETM failure classes data module composition apparatus and its creative method based on visible process | |
| CN110335594A (en) | Automatic speech recognition difficulty sample method for digging based on multi-instance learning | |
| CN106326307A (en) | Language interaction method | |
| CN106339455A (en) | Webpage text extracting method based on text tag feature mining | |
| CN106648819A (en) | Internationalized code conversion method based on editor | |
| CN109992770A (en) | A kind of Laotian name entity recognition method based on combination neural net | |
| CN106372053A (en) | Syntactic analysis method and apparatus |
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 |