CN106610930B - Foreign language writing methods automatic error correction method and system - Google Patents
Foreign language writing methods automatic error correction method and system Download PDFInfo
- Publication number
- CN106610930B CN106610930B CN201510691638.9A CN201510691638A CN106610930B CN 106610930 B CN106610930 B CN 106610930B CN 201510691638 A CN201510691638 A CN 201510691638A CN 106610930 B CN106610930 B CN 106610930B
- Authority
- CN
- China
- Prior art keywords
- sentence
- foreign language
- error
- error correction
- text
- 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.)
- Active
Links
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F40/00—Handling natural language data
- G06F40/20—Natural language analysis
- G06F40/253—Grammatical analysis; Style critique
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F40/00—Handling natural language data
- G06F40/20—Natural language analysis
- G06F40/232—Orthographic correction, e.g. spell checking or vowelisation
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Health & Medical Sciences (AREA)
- Artificial Intelligence (AREA)
- Audiology, Speech & Language Pathology (AREA)
- Computational Linguistics (AREA)
- General Health & Medical Sciences (AREA)
- Physics & Mathematics (AREA)
- General Engineering & Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Electrically Operated Instructional Devices (AREA)
- Machine Translation (AREA)
Abstract
The invention discloses a kind of foreign language writing methods automatic error correction method and systems, this method comprises: building is used for the error correcting model of foreign language sentence automatic error-correcting in advance；Obtain the foreign language sentence that user writes；Extract the sentence vector of the term vector of each word and the sentence in the sentence；The sentence vector of the term vector of word each in the sentence and the sentence is successively inputted into the error correcting model, obtains the correction term vector of each word of the error correcting model output, and according to text after the generation error correction of the correction term vector of each word；Show text after the error correction.Using the present invention, the automatic error-correcting of statement error in Foreign Language writing may be implemented, improve student written and learning effect.
Description
Technical field
The present invention relates to text-processing technical fields, and in particular to a kind of foreign language writing methods automatic error correction method and system.
Background technique
In recent years, in traditional education sector, the IT application in education sector upgrading for also gradually carrying out a new generation is explored, such as online religion
The advantages such as study is convenient and efficient with its, threshold is low are educated, are gradually received by numerous teachers or student, and then gradually popularize.Foreign language book
Evaluation is write as an important application function in online education, is always the hot spot of Faculty and Students' concern, in recent years, with
Related fields researcher's is continually striving to, and validity, the accuracy of evaluation result are all stepping up, especially as Sinology
Raw english composition is corrected.Existing english composition is corrected, and is merely given as an overall score mostly, or further give
The overall assessment write a composition out and the miscue being likely to occur, and guided bone cannot be provided for the mistake occurred in composition
Correct option is not obvious the promotion effect of Foreign Languages writing ability so that students ' learning performance is had a greatly reduced quality.
Summary of the invention
The present invention provides a kind of foreign language writing methods automatic error correction method and system, to realize statement error in Foreign Language writing
Automatic error-correcting improves user's writing and learning effect.
For this purpose, the invention provides the following technical scheme:
A kind of foreign language writing methods automatic error correction method, comprising:
Building is used for the error correcting model of foreign language sentence automatic error-correcting in advance；
Obtain the foreign language sentence that user writes；
Extract the sentence vector of the term vector of each word and the sentence in the sentence；
The sentence vector of the term vector of word each in the sentence and the sentence is successively inputted into the error correcting model, is obtained
The correction term vector of each word of the error correcting model output, and according to text after the generation error correction of the correction term vector of each word；
Show text after the error correction.
Preferably, error correcting model of the building for foreign language sentence automatic error-correcting includes:
Training corpus is collected, the training corpus includes the foreign language standard sentence corpus for setting quantity and corresponding mistake
Accidentally corpus；
Determine the topological structure of the error correcting model for foreign language sentence automatic error-correcting；
Error correcting model is obtained based on training corpus training.
Preferably, the otherness that writing mistake and mother tongue grammer and foreign language grammer are easily made according to user constructs and described outer
The corresponding wrong corpus of language standard sentence corpus.
Preferably, the topological structure of the error correcting model be multi-story and multi-span, including an input layer, one it is hidden
Layer and an output layer；The input layer include three nodes, be respectively as follows: the term vector of current word in foreign language sentence, it is previous when
The sentence vector of the hidden layer vector at quarter, foreign language sentence；The output layer includes a node, is entangled for current word in foreign language sentence
Positive term vector.
Preferably, the method also includes:
After showing the error correction before text, whether text is identical as the foreign language sentence after judging the error correction；
If it is, executing the step of showing text after the error correction；
Otherwise, show text after the error correction, and carry out miscue.
Preferably, the progress miscue includes:
In text after showing the error correction, the error correction is shown using highlighting or being different from other colors for showing content
Text afterwards；Or
In text after showing the error correction, while showing the foreign language sentence, and mark after the error correction in text with
Different content in the foreign language sentence；Or
In text after showing the error correction, show prompt information to user；Or
In text after showing the error correction, voice broadcast prompt information.
A kind of foreign language writing methods automatic error correction system, comprising:
Error correcting model constructs module, for constructing the error correcting model for foreign language sentence automatic error-correcting in advance；
Sentence obtains module, for obtaining the foreign language sentence of user's writing；
Vector extraction module, for extracting the sentence vector of the term vector of each word and the sentence in the sentence；
Correction module, for successively will the sentence vector of the term vector of word each in the sentence and the sentence input described in
Error correcting model obtains the correction term vector of each word of the error correcting model output, and raw according to the correction term vector of each word
At text after error correction；
Display module, for showing text after the error correction.
Preferably, the error correcting model building module includes:
Corpus collector unit, for collecting training corpus, the training corpus includes the foreign language standard sentence for setting quantity
Corpus and corresponding wrong corpus；
Topological structure determination unit, for determining the topological structure of the error correcting model for foreign language sentence automatic error-correcting；
Training unit, for obtaining error correcting model based on training corpus training.
Preferably, the corpus collector unit includes:
Standard sentence corpus collects subelement, for collecting the foreign language standard sentence corpus of setting quantity；
Mistake corpus constructs subelement, for easily making the difference of writing mistake and mother tongue grammer and foreign language grammer according to user
Property, construct wrong corpus corresponding with the foreign language standard sentence corpus.
Preferably, the topological structure of the error correcting model be multi-story and multi-span, including an input layer, one it is hidden
Layer and an output layer；The input layer include three nodes, be respectively as follows: the term vector of current word in foreign language sentence, it is previous when
The sentence vector of the hidden layer vector at quarter, foreign language sentence；The output layer includes a node, is entangled for current word in foreign language sentence
Positive term vector.
Preferably, the system also includes:
Miscue module, for showing after the error correction before text, after judging the error correction in the display module
Whether text is identical as the foreign language sentence；If it is not, then carrying out miscue.
Preferably, the miscue module specifically carries out miscue in the following ways:
In text after showing the error correction, the error correction is shown using highlighting or being different from other colors for showing content
Text afterwards；Or
In text after showing the error correction, while showing the foreign language sentence, and mark after the error correction in text with
Different content in the foreign language sentence；Or
In text after showing the error correction, show prompt information to user；Or
In text after showing the error correction, voice broadcast prompt information.
Foreign language writing methods automatic error correction method provided in an embodiment of the present invention and system, by being constructed in advance for automatic error-correcting
Error correcting model, be then based on trained error correcting model Foreign Language sentence and carry out automatic error-correcting, realize in Foreign Language writing
The automatic error-correcting of statement error improves user's writing and learning effect.
Detailed description of the invention
In order to illustrate the technical solutions in the embodiments of the present application or in the prior art more clearly, below will be to institute in embodiment
Attached drawing to be used is needed to be briefly described, it should be apparent that, the accompanying drawings in the following description is only one recorded in the present invention
A little embodiments are also possible to obtain other drawings based on these drawings for those of ordinary skill in the art.
Fig. 1 is a kind of flow chart of foreign language writing methods automatic error correction method of the embodiment of the present invention；
Fig. 2 is the building flow chart of error correcting model in the embodiment of the present invention；
Fig. 3 is the topological structure schematic diagram of error correcting model in the embodiment of the present invention；
Fig. 4 is another flow chart of foreign language writing methods automatic error correction method of the embodiment of the present invention；
Fig. 5 is a kind of structural schematic diagram of foreign language writing methods automatic error correction system of the embodiment of the present invention；
Fig. 6 is the structural schematic diagram of error correcting model building model in the embodiment of the present invention；
Fig. 7 is another structural schematic diagram of foreign language writing methods automatic error correction system of the embodiment of the present invention.
Specific embodiment
The scheme of embodiment in order to enable those skilled in the art to better understand the present invention with reference to the accompanying drawing and is implemented
Mode is described in further detail the embodiment of the present invention.
The prior art aiming at the problem that mistake occurred in student written cannot provide the correct option of guided bone, this
Inventive embodiments provide a kind of foreign language writing methods automatic error correction method and system, by constructing the error correction mould for automatic error-correcting in advance
Type is then based on trained error correcting model Foreign Language sentence and carries out automatic error-correcting, realizes statement error in Foreign Language writing
Automatic error-correcting, improve user writing and learning effect.The foreign language refers to the languages other than mother tongue, such as using Chinese as mother tongue
Chinese, English is exactly one of its foreign language.
As shown in Figure 1, being a kind of flow chart of foreign language writing methods automatic error correction method of the embodiment of the present invention, including following step
It is rapid:
Step 101, building in advance is used for the error correcting model of foreign language sentence automatic error-correcting.
Specifically, can collect a certain number of includes foreign language standard sentence corpus and corresponding wrong corpus conduct
Training corpus, and determine the topological structure of the error correcting model for foreign language sentence automatic error-correcting, it is then based on these training corpus
Training obtains error correcting model.
Certainly, wrong corpus corresponding with the foreign language standard sentence corpus can also be according to the foreign language standard sentence language
Material automatically generates.
For example, first collect certain amount (such as about 50,000,000 sentences) foreign language standard sentence corpus, and according to
The otherness of writing mistake and mother tongue grammer and foreign language grammer is easily made at family, constructs mistake corresponding with the foreign language standard sentence corpus
Accidentally corpus.
When constructing wrong corpus corresponding with the foreign language standard sentence corpus, common wrong word replacement, word can be passed through
The modes Foreign Language standard sentence corpus such as language dependence sequencing is handled, and the wrong corpus is obtained.
Wherein, common wrong word replacement can be replaced according to user's list of common error of collection, or by correct language
All predicate verbs in material translate into mother tongue by dictionary for translation, are then again translated into translation result outside by dictionary for translation
Language, to obtain a large amount of mother tongue formula synonyms of the predicate verb.Then these mother tongue formula synonyms are replaced to pair in former sentence
The sentence of big often mistake can be obtained in the word answered.
Such as: standard sentence " Would you like to come to our party on Friday night? ",
Its translation is " you are ready to participate in our party? "
Wherein, the Chinese meaning of predicate verb " come to " is " participation ", and " participation " corresponding English can have
" join, attend ", therefore the wrong sentence of available following corresponding above-mentioned standard sentence:
Would you like to join our party on Friday night?
Would you like to attend our party on Friday night?
Wherein, word dependence relationship sequencing such as may is that adjective (either Adjective Phrases) with by adjective (or
By Adjective Phrases) location swap, qualifier and modificand verb location swap etc..
The specific training process of error correcting model will be described in detail later.
Step 102, the foreign language sentence that user writes is obtained.
The foreign language sentence that user writes can be and be inputted online by terminal, be also possible to through the acquisitions side such as scan, take pictures
Formula is collected hand-written or text sentence of block letter, without limitation to this embodiment of the present invention.
It should be noted that it is subsequent to the foreign language sentence carry out correction process when, be as unit of single statement into
Row processing.Therefore, when collecting multiple foreign language sentences or user continuously inputs multiple foreign language sentences, it is also necessary to first to these languages
Sentence is distinguished according to separators such as punctuate etc., obtains each single foreign language sentence.
Step 103, the sentence vector of the term vector of each word and the sentence in the sentence is extracted.
The acquisition of the term vector can use existing some extracting methods, for example, by preparatory trained word to
Amount is extracted model and is obtained, and the term vector extracts model and can be based on word2vec in google_1_billion corpus
Upper training obtains, naturally it is also possible to obtain term vector by the training of other approach and extract model, not do to this embodiment of the present invention
It limits.
The sentence vector can obtain in the sentence by way of being averaged in the vector summation of each word.
Step 104, the sentence vector of the term vector of word each in the sentence and the sentence is successively inputted into the error correction
Model obtains the correction term vector of each word of the error correcting model output, and is entangled according to the generation of the correction term vector of each word
Text after mistake.
It should be noted that if being based only on current term vector, previous moment hidden layer vector sentence as input entangles
Mismatch type because its error correction result is entirely an open content, thus will appear the sentence after error correction and former sentence in many cases
The poor situation of correlation.In this regard, in embodiments of the present invention, sentence vector is introduced in the input of error correcting model, by
In the powerful memory of network and smoothing capability, the reliability of error correction result is made to be effectively guaranteed.
Step 105, show text after the error correction.
As shown in Fig. 2, being the building flow chart of error correcting model in the embodiment of the present invention, comprising the following steps:
Step 201, collect training corpus, the training corpus include set quantity foreign language standard sentence corpus and and its
Corresponding mistake corpus.
It should be noted that it is corresponding with foreign language standard sentence corpus mistake corpus can by artificially collecting to obtain,
It can be constructed and be obtained automatically according to the foreign language standard sentence corpus, without limitation to this embodiment of the present invention.
Step 202, the topological structure of the error correcting model for foreign language sentence automatic error-correcting is determined.
In embodiments of the present invention, the topological structure of error correcting model can be a neural network, such as Multi-Layer Feedback mind
Through network (RNN), which includes an input layer, a hidden layer and an output layer, as shown in Figure 3.Input layer is 3 sections
Point, the term vector (for example be 200 dimension) of current word respectively in foreign language sentence, previous moment hidden layer vector (for example be 1024
Dimension) and foreign language sentence sentence term vector (for example be 200 dimension).Hidden layer is a node, is a hidden layer at current time
Vector (for example being 1024 dimensions).Output layer is a node, and node output is a term vector (for example tieing up for 200).
Step 203, error correcting model is obtained based on training corpus training.
As shown in figure 3, the hidden layer vector of current time t are as follows:
ht=f (WS+UXt+V·ht-1) (1)
Wherein, S is sentence vector, XtFor current term vector, ht-1For the hidden layer vector at t-1 moment, W be sentence vector with
The connection weight matrix of current hidden layer vector, U is the connection weight matrix of current term vector and current hidden layer vector, when V is t-1
The connection weight matrix of hidden layer vector and current hidden layer vector is carved, f is tanh function.
Output are as follows:
Outt=Oht (2)
Wherein, O is the connection weight matrix of hidden layer and output layer.
The objective function of error correcting model training are as follows:
T=Outt-Et=Of (WS+UXt+V·ht-1)-Et (3)
Wherein, EtPreferably to export term vector, the as term vector at t+1 moment.
It can be based on the sentence automatic error correction method of RNN, using error back propagation (Error Back
Propagation, BP) algorithm progress model training, finally make target function value T minimum.
Foreign language writing methods automatic error correction method provided in an embodiment of the present invention, by constructing the error correction for automatic error-correcting in advance
Model is then based on trained error correcting model Foreign Language sentence and carries out automatic error-correcting, and it is wrong to realize sentence in Foreign Language writing
Automatic error-correcting accidentally improves user's writing and learning effect.
As shown in figure 4, being another flow chart of foreign language writing methods automatic error correction method of the embodiment of the present invention, including following step
It is rapid:
Step 401, building in advance is used for the error correcting model of foreign language sentence automatic error-correcting.
Step 402, the foreign language sentence that user writes is obtained.
Step 403, the sentence vector of the term vector of each word and the sentence in the sentence is extracted.
Step 404, the sentence vector of the term vector of word each in the sentence and the sentence is successively inputted into the error correction
Model obtains the correction term vector of each word of the error correcting model output, and is entangled according to the generation of the correction term vector of each word
Text after mistake.
Step 405, whether text is identical as the foreign language sentence after judging the error correction.If so, thening follow the steps
406；Otherwise, step 407 is executed.
Step 406, show text after the error correction.
Step 407, show text after the error correction, and carry out miscue.
Carry out miscue mode can there are many, such as:
In text after showing the error correction, the error correction is shown using highlighting or being different from other colors for showing content
Text afterwards；Or
In text after showing the error correction, while showing the foreign language sentence, and mark after the error correction in text with
Different content in the foreign language sentence；Or
In text after showing the error correction, show prompt information to user；Or
In text after showing the error correction, voice broadcast prompt information.
It is, of course, also possible to have other miscue modes, without limitation to this embodiment of the present invention.
Compared with embodiment illustrated in fig. 1, the foreign language writing methods automatic error correction method of the embodiment not only realizes Foreign Language and writes
The automatic error-correcting of statement error in work improves user's writing and learning effect, and can preferably prompt user's sentence wrong
Whereabouts accidentally facilitates user and learns to use, and the user experience is improved.
Correspondingly, the embodiment of the present invention also provides a kind of foreign language writing methods automatic error correction system, as shown in figure 5, being the system
A kind of structural schematic diagram.
In this embodiment, the system comprises:
Error correcting model constructs module 501, for constructing the error correcting model for foreign language sentence automatic error-correcting in advance；
Sentence obtains module 502, for obtaining the foreign language sentence of user's writing；
Vector extraction module 503, for extracting the sentence vector of the term vector of each word and the sentence in the sentence；
Correction module 504, for successively inputting the sentence vector of the term vector of word each in the sentence and the sentence
The error correcting model, obtains the correction term vector of each word of error correcting model output, and according to the correction word of each word to
Amount generates text after error correction；
Display module 506, for showing text after the error correction.
As shown in fig. 6, being the structural schematic diagram of error correcting model building model in the embodiment of the present invention, including following list
Member:
Corpus collector unit 61, for collecting training corpus, the training corpus includes the foreign language standard speech for setting quantity
Sentence corpus and corresponding wrong corpus；
Topological structure determination unit 62, for determining the topological structure of the error correcting model for foreign language sentence automatic error-correcting,
A kind of topological structure of error correcting model is as shown in Figure 3, naturally it is also possible to use other structures, not limit this embodiment of the present invention
It is fixed；
Training unit 63, for obtaining error correcting model based on training corpus training.
Wherein, corpus collector unit 61 can collect a certain number of foreign language standard sentence corpus and corresponding respectively
Mistake corpus can also only collect a certain number of foreign language standard sentence corpus, then according to these foreign languages as training corpus
Standard sentence corpus construction generates corresponding wrong corpus.For example, a kind of specific implementation of corpus collector unit 61
It may include: that standard sentence corpus collects subelement and wrong corpus construction subelement, wherein the standard sentence corpus is collected
Subelement is used to collect the foreign language standard sentence corpus of setting quantity；The mistake corpus construction subelement is used for easy according to user
The otherness for making writing mistake and mother tongue grammer and foreign language grammer constructs wrong language corresponding with the foreign language standard sentence corpus
Material.Specific configuration process is referred to the description in the embodiment of the present invention method of front, and details are not described herein.
Foreign language writing methods automatic error correction system provided in an embodiment of the present invention, by constructing the error correction for automatic error-correcting in advance
Model is then based on trained error correcting model Foreign Language sentence and carries out automatic error-correcting, and it is wrong to realize sentence in Foreign Language writing
Automatic error-correcting accidentally improves user's writing and learning effect.
The convenience used in order to further increase allows users to the whereabouts for intuitively knowing statement error,
In another embodiment of present system, the system may also include that
Miscue module 506 judges the error correction for showing after the error correction before text in display module 505
Whether text is identical as the foreign language sentence afterwards；If it is not, then carrying out miscue.
The miscue module 506 specifically can carry out miscue using following various ways:
In text after showing the error correction, the error correction is shown using highlighting or being different from other colors for showing content
Text afterwards；Or
In text after showing the error correction, while showing the foreign language sentence, and mark after the error correction in text with
Different content in the foreign language sentence；Or
In text after showing the error correction, show prompt information to user；Or
In text after showing the error correction, voice broadcast prompt information.
The foreign language writing methods automatic error correction system of the embodiment not only realizes the automatic of statement error in Foreign Language writing and entangles
Mistake improves user's writing and learning effect, and can preferably prompt the whereabouts of user's statement error, facilitates use
Family study uses, and the user experience is improved.
All the embodiments in this specification are described in a progressive manner, same and similar portion between each embodiment
Dividing may refer to each other, and each embodiment focuses on the differences from other embodiments.Especially for system reality
For applying example, since it is substantially similar to the method embodiment, so describing fairly simple, related place is referring to embodiment of the method
Part explanation.System embodiment described above is only schematical, wherein described be used as separate part description
Unit may or may not be physically separated, component shown as a unit may or may not be
Physical unit, it can it is in one place, or may be distributed over multiple network units.It can be according to the actual needs
Some or all of the modules therein is selected to achieve the purpose of the solution of this embodiment.Those of ordinary skill in the art are not paying
In the case where creative work, it can understand and implement.
The embodiment of the present invention has been described in detail above, and specific embodiment used herein carries out the present invention
It illustrates, method and system of the invention that the above embodiments are only used to help understand；Meanwhile for the one of this field
As technical staff, according to the thought of the present invention, there will be changes in the specific implementation manner and application range, to sum up institute
It states, the contents of this specification are not to be construed as limiting the invention.
Claims (12)
1. a kind of foreign language writing methods automatic error correction method characterized by comprising
Building in advance is used for the error correcting model of foreign language sentence automatic error-correcting, and the training corpus of the error correcting model includes based on single
The mother tongue formula synonym of standard foreign language word；
Obtain the foreign language sentence that user writes；
Extract the sentence vector of the term vector of each word and the sentence in the sentence；
The sentence vector of the term vector of word each in the sentence and the sentence is successively inputted into the error correcting model, is obtained described
The correction term vector of each word of error correcting model output, and according to text after the generation error correction of the correction term vector of each word；
Show text after the error correction.
2. the method according to claim 1, wherein the building is used for the error correction mould of foreign language sentence automatic error-correcting
Type includes:
Training corpus is collected, the training corpus includes the foreign language standard sentence corpus and corresponding wrong language for setting quantity
Material；
Determine the topological structure of the error correcting model for foreign language sentence automatic error-correcting；
Error correcting model is obtained based on training corpus training.
3. according to the method described in claim 2, it is characterized in that, easily making writing mistake and mother tongue grammer and foreign language according to user
The otherness of grammer constructs wrong corpus corresponding with the foreign language standard sentence corpus.
4. according to the method described in claim 2, it is characterized in that, the topological structure of the error correcting model is Multi-Layer Feedback nerve
Network, including an input layer, a hidden layer and an output layer；The input layer includes three nodes, is respectively as follows: foreign language language
The sentence vector of the term vector of current word, the hidden layer vector of previous moment, foreign language sentence in sentence；The output layer includes a section
Point is the correction term vector of current word in foreign language sentence.
5. method according to any one of claims 1 to 4, which is characterized in that the method also includes:
After showing the error correction before text, whether text is identical as the foreign language sentence after judging the error correction；
If it is, executing the step of showing text after the error correction；
Otherwise, show text after the error correction, and carry out miscue.
6. according to the method described in claim 5, it is characterized in that, the progress miscue includes:
In text after showing the error correction, the error correction is shown hereinafter using highlighting or being different from other colors for showing content
This；Or
In text after showing the error correction, while showing the foreign language sentence, and mark after the error correction in text with it is described
Different content in foreign language sentence；Or
In text after showing the error correction, show prompt information to user；Or
In text after showing the error correction, voice broadcast prompt information.
7. a kind of foreign language writing methods automatic error correction system characterized by comprising
Error correcting model constructs module, for constructing the error correcting model for foreign language sentence automatic error-correcting, the error correcting model in advance
Training corpus include the mother tongue formula synonym based on single standard foreign language word；
Sentence obtains module, for obtaining the foreign language sentence of user's writing；
Vector extraction module, for extracting the sentence vector of the term vector of each word and the sentence in the sentence；
Correction module, for the sentence vector of the term vector of word each in the sentence and the sentence successively to be inputted the error correction
Model obtains the correction term vector of each word of the error correcting model output, and is entangled according to the generation of the correction term vector of each word
Text after mistake；
Display module, for showing text after the error correction.
8. system according to claim 7, which is characterized in that the error correcting model constructs module and includes:
Corpus collector unit, for collecting training corpus, the training corpus includes the foreign language standard sentence corpus for setting quantity
And corresponding wrong corpus；
Topological structure determination unit, for determining the topological structure of the error correcting model for foreign language sentence automatic error-correcting；
Training unit, for obtaining error correcting model based on training corpus training.
9. system according to claim 8, which is characterized in that the corpus collector unit includes:
Standard sentence corpus collects subelement, for collecting the foreign language standard sentence corpus of setting quantity；
Mistake corpus constructs subelement, for easily making the otherness of writing mistake and mother tongue grammer and foreign language grammer according to user,
Construct wrong corpus corresponding with the foreign language standard sentence corpus.
10. system according to claim 8, which is characterized in that the topological structure of the error correcting model is Multi-Layer Feedback mind
Through network, including an input layer, a hidden layer and an output layer；The input layer includes three nodes, is respectively as follows: foreign language
The sentence vector of the term vector of current word, the hidden layer vector of previous moment, foreign language sentence in sentence；The output layer includes one
Node is the correction term vector of current word in foreign language sentence.
11. according to the described in any item systems of claim 7 to 10, which is characterized in that the system also includes:
Miscue module judges text after the error correction for showing after the error correction before text in the display module
It is whether identical as the foreign language sentence；If it is not, then carrying out miscue.
12. system according to claim 11, which is characterized in that the miscue module specifically in the following ways into
Row miscue:
In text after showing the error correction, the error correction is shown hereinafter using highlighting or being different from other colors for showing content
This；Or
In text after showing the error correction, while showing the foreign language sentence, and mark after the error correction in text with it is described
Different content in foreign language sentence；Or
In text after showing the error correction, show prompt information to user；Or
In text after showing the error correction, voice broadcast prompt information.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201510691638.9A CN106610930B (en) | 2015-10-22 | 2015-10-22 | Foreign language writing methods automatic error correction method and system |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201510691638.9A CN106610930B (en) | 2015-10-22 | 2015-10-22 | Foreign language writing methods automatic error correction method and system |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN106610930A CN106610930A (en) | 2017-05-03 |
| CN106610930B true CN106610930B (en) | 2019-09-03 |
Family
ID=58610486
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201510691638.9A Active CN106610930B (en) | 2015-10-22 | 2015-10-22 | Foreign language writing methods automatic error correction method and system |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN106610930B (en) |
Families Citing this family (22)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN107341143B (en) * | 2017-05-26 | 2020-08-14 | 北京奇艺世纪科技有限公司 | Sentence continuity judgment method and device and electronic equipment |
| CN107357775A (en) | 2017-06-05 | 2017-11-17 | 百度在线网络技术（北京）有限公司 | The text error correction method and device of Recognition with Recurrent Neural Network based on artificial intelligence |
| CN107451106A (en) * | 2017-07-26 | 2017-12-08 | 阿里巴巴集团控股有限公司 | Text method and device for correcting, electronic equipment |
| JP7031101B2 (en) * | 2017-08-03 | 2022-03-08 | リンゴチャンプ インフォメーション テクノロジー （シャンハイ） カンパニー， リミテッド | Methods, systems and tangible computer readable devices |
| CN108197107A (en) * | 2017-12-29 | 2018-06-22 | 秦男 | Data processing method |
| CN108595410B (en) * | 2018-03-19 | 2023-03-24 | 小船出海教育科技（北京）有限公司 | Automatic correction method and device for handwritten composition |
| CN108519973A (en) * | 2018-03-29 | 2018-09-11 | 广州视源电子科技股份有限公司 | Character spelling detection method, system, computer equipment and storage medium |
| CN108491392A (en) * | 2018-03-29 | 2018-09-04 | 广州视源电子科技股份有限公司 | Method, system, computer device and storage medium for correcting character spelling errors |
| CN108563632A (en) * | 2018-03-29 | 2018-09-21 | 广州视源电子科技股份有限公司 | Method, system, computer device and storage medium for correcting character spelling errors |
| CN108595432B (en) * | 2018-04-28 | 2022-02-08 | 江苏医像信息技术有限公司 | Medical document error correction method |
| CN109376360B (en) * | 2018-10-29 | 2023-10-20 | 广东小天才科技有限公司 | Method and device for assisting in learning language |
| CN109670040B (en) * | 2018-11-27 | 2024-04-05 | 平安科技（深圳）有限公司 | Writing assistance method and device, storage medium and computer equipment |
| CN111435408B (en) * | 2018-12-26 | 2023-04-18 | 阿里巴巴集团控股有限公司 | Dialog error correction method and device and electronic equipment |
| KR102199835B1 (en) * | 2018-12-31 | 2021-01-07 | 주식회사 엘솔루 | System for correcting language and method thereof, and method for learning language correction model |
| CN111523305A (en) * | 2019-01-17 | 2020-08-11 | 阿里巴巴集团控股有限公司 | Text error correction method, device and system |
| CN111523306A (en) * | 2019-01-17 | 2020-08-11 | 阿里巴巴集团控股有限公司 | Text error correction method, device and system |
| CN110929875A (en) * | 2019-10-12 | 2020-03-27 | 平安国际智慧城市科技股份有限公司 | Intelligent language learning method, system, device and medium based on machine learning |
| CN112541342B (en) * | 2020-12-08 | 2022-07-22 | 北京百度网讯科技有限公司 | Text error correction method and device, electronic equipment and storage medium |
| CN112560430B (en) * | 2020-12-25 | 2024-04-02 | 北京百度网讯科技有限公司 | Error correction method and device for numerical content in text and electronic equipment |
| CN113254574A (en) * | 2021-03-15 | 2021-08-13 | 河北地质大学 | Method, device and system for auxiliary generation of customs official documents |
| WO2022240358A1 (en) * | 2021-05-11 | 2022-11-17 | National University Of Singapore | System and method for training a culturally-specific assisting language learning model |
| CN113033186B (en) * | 2021-05-31 | 2021-08-17 | 江苏联著实业股份有限公司 | Error correction early warning method and system based on event analysis |
Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102122507A (en) * | 2010-01-08 | 2011-07-13 | 龚澍 | Speech error detection method by front-end processing using artificial neural network (ANN) |
| CN104346035A (en) * | 2013-08-06 | 2015-02-11 | 联想（新加坡）私人有限公司 | Indicating automatically corrected words |
| CN104462085A (en) * | 2013-09-12 | 2015-03-25 | 腾讯科技（深圳）有限公司 | Method and device for correcting search keywords |
| CN104484377A (en) * | 2014-12-09 | 2015-04-01 | 百度在线网络技术（北京）有限公司 | Generating method and device of substitute dictionaries |
| CN104882139A (en) * | 2015-05-28 | 2015-09-02 | 百度在线网络技术（北京）有限公司 | Voice synthesis method and device |
Family Cites Families (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US10373047B2 (en) * | 2014-02-28 | 2019-08-06 | Educational Testing Service | Deep convolutional neural networks for automated scoring of constructed responses |
-
2015
- 2015-10-22 CN CN201510691638.9A patent/CN106610930B/en active Active
Patent Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102122507A (en) * | 2010-01-08 | 2011-07-13 | 龚澍 | Speech error detection method by front-end processing using artificial neural network (ANN) |
| CN104346035A (en) * | 2013-08-06 | 2015-02-11 | 联想（新加坡）私人有限公司 | Indicating automatically corrected words |
| CN104462085A (en) * | 2013-09-12 | 2015-03-25 | 腾讯科技（深圳）有限公司 | Method and device for correcting search keywords |
| CN104484377A (en) * | 2014-12-09 | 2015-04-01 | 百度在线网络技术（北京）有限公司 | Generating method and device of substitute dictionaries |
| CN104882139A (en) * | 2015-05-28 | 2015-09-02 | 百度在线网络技术（北京）有限公司 | Voice synthesis method and device |
Non-Patent Citations (1)
| Title |
|---|
| Techniques for Automatically Correcting Words in Text;KAREN KUKICH;《ACM Computing Surveys》;19921231;第24卷(第4期);第377-433页 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN106610930A (en) | 2017-05-03 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN106610930B (en) | Foreign language writing methods automatic error correction method and system | |
| CN106448675B (en) | Method and system for correcting recognition text | |
| CN103970765B (en) | Correct mistakes model training method, device and text of one is corrected mistakes method, device | |
| CN107563498B (en) | Image description method and system based on visual and semantic attention combined strategy | |
| CN106503255B (en) | Method and system for automatically generating article based on description text | |
| CN105654945B (en) | Language model training method, device and equipment | |
| Yu et al. | Chinese spelling error detection and correction based on language model, pronunciation, and shape | |
| CN106534548B (en) | Voice error correction method and device | |
| CN107562918A (en) | A kind of mathematical problem knowledge point discovery and batch label acquisition method | |
| Thai et al. | Exploring document-level literary machine translation with parallel paragraphs from world literature | |
| US20110208508A1 (en) | Interactive Language Training System | |
| CN105760359A (en) | Question processing system and method thereof | |
| CN108595427B (en) | Subjective question scoring method and device, readable storage medium and electronic equipment | |
| Almazova et al. | Computer assisted individual approach to acquiring foreign vocabulary of students major | |
| Hanoulle et al. | The efficacy of terminology-extraction systems for the translation of documentaries | |
| Wang et al. | Are perfect transcripts necessary when we analyze classroom dialogue using AIoT? | |
| Volkova et al. | Lightly supervised learning of procedural dialog systems | |
| Yeh et al. | Chinese word spelling correction based on rule induction | |
| Wu et al. | Chinese grammatical error diagnosis system based on hybrid model | |
| Zhou et al. | On Development History of Australia’s Language Policy and the Enlightenment to China’s Foreign Language Education | |
| CN102955842A (en) | Multi-feature-fused controlling method for recognizing Chinese organization name | |
| CN107783965B (en) | Sentence pattern structure-based translation method and device | |
| Wang | An English listening and speaking ability training system based on binary decision tree | |
| Jo et al. | Effective computer‐assisted pronunciation training based on phone‐sensitive word recommendation | |
| Agrawal et al. | Do Text Simplification Systems Preserve Meaning? A Human Evaluation via Reading Comprehension |
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 | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |