CN103294660A - Automatic English composition scoring method and system - Google Patents
Automatic English composition scoring method and system Download PDFInfo
- Publication number
- CN103294660A CN103294660A CN2012100499824A CN201210049982A CN103294660A CN 103294660 A CN103294660 A CN 103294660A CN 2012100499824 A CN2012100499824 A CN 2012100499824A CN 201210049982 A CN201210049982 A CN 201210049982A CN 103294660 A CN103294660 A CN 103294660A
- Authority
- CN
- China
- Prior art keywords
- mark
- composition
- dimension
- appraisal result
- waiting
- 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.)
- Granted
Links
- 239000000203 mixture Substances 0.000 title claims abstract description 155
- 238000013077 scoring method Methods 0.000 title claims abstract description 21
- 238000004364 calculation method Methods 0.000 claims abstract description 13
- 238000012545 processing Methods 0.000 claims description 25
- 238000004458 analytical method Methods 0.000 claims description 17
- 230000008878 coupling Effects 0.000 claims description 5
- 238000010168 coupling process Methods 0.000 claims description 5
- 238000005859 coupling reaction Methods 0.000 claims description 5
- 238000012549 training Methods 0.000 claims description 5
- 238000000034 method Methods 0.000 description 17
- 238000010586 diagram Methods 0.000 description 4
- 238000005516 engineering process Methods 0.000 description 3
- 238000011156 evaluation Methods 0.000 description 3
- 230000008569 process Effects 0.000 description 3
- 238000013473 artificial intelligence Methods 0.000 description 2
- 238000012417 linear regression Methods 0.000 description 2
- 238000012986 modification Methods 0.000 description 2
- 230000004048 modification Effects 0.000 description 2
- 229910001369 Brass Inorganic materials 0.000 description 1
- 238000013459 approach Methods 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 230000008901 benefit Effects 0.000 description 1
- 230000015572 biosynthetic process Effects 0.000 description 1
- 239000010951 brass Substances 0.000 description 1
- 230000006870 function Effects 0.000 description 1
- 238000010801 machine learning Methods 0.000 description 1
- 230000007246 mechanism Effects 0.000 description 1
- 238000003058 natural language processing Methods 0.000 description 1
- 238000003786 synthesis reaction Methods 0.000 description 1
- 238000012360 testing method Methods 0.000 description 1
Images
Abstract
The invention provides an automatic English composition scoring method and an automatic English composition scoring system. The automatic English composition scoring method comprises the following steps: acquiring a composition to be scored; analyzing the composition to be scored to obtain a plurality of dimensionalities; according to each dimensionality, performing scoring calculation on the composition to be scored to obtain a plurality of scoring results; and weighting the scoring results to obtain a final scoring result of the composition to be scored. By the automatic English composition scoring method and the automatic English composition scoring system, the scoring result of the English composition can be obtained more objectively.
Description
Technical field
The present invention relates to electronic information industry artificial intelligence field, refer to a kind of english composition automatic scoring method and system especially.
Background technology
Existing english composition methods of marking, basic ideas are by various natural language technology, from english composition to be marked, extract text feature, with artificial scoring and machine learning and artificial intelligence approach, relation between text feature and the mark is learnt, carried out automatic scoring by text feature and the relation between the mark of gained.
In the process of automatic scoring, as independent variable, adopt equation of linear regression to obtain last appraisal result the text feature of said extracted, this mode is to adopt the regression formula of an integral body to obtain final appraisal result.The appraisal result that obtains like this has bigger gap with artificial appraisal result, can not reflect the height of composition level objectively.And can't provide the detailed suggestion of correcting.
Summary of the invention
The technical problem to be solved in the present invention provides a kind of english composition automatic scoring method and system, and it is comparatively objective to obtain appraisal result, can reflect the level of english composition comparatively truly.
For solving the problems of the technologies described above, embodiments of the invention provide a kind of english composition automatic scoring method, comprising:
Obtain and treat the scoring composition;
The described composition of waiting to mark is analyzed, obtained a plurality of dimensions;
According to each dimension, score calculation is carried out in the described composition of waiting to mark, obtain a plurality of appraisal result;
Described a plurality of appraisal result are weighted processing, obtain the final appraisal result of the described composition of waiting to mark.
Wherein, the described composition of waiting to mark is analyzed, the step that obtains a plurality of dimensions comprises:
Lexical analysis, the analysis of sentence, structure of an article analysis and content analysis are carried out in the described composition of waiting to mark, obtained sub-dimension, the sub-dimension about sentence, the sub-dimension of the structure of an article and the sub-dimension of content about vocabulary;
Respectively the sub-dimension of described vocabulary, the sub-dimension of sentence, the sub-dimension of the structure of an article and the sub-dimension of content are divided, obtain a plurality of dimensions.
Wherein, described dimension is specific targets data of scoring composition, and the final appraisal result of described specific targets data and described scoring composition has linear dependence, monotonicity, independence, diversionary and/or balance.
Wherein, according to each dimension, score calculation is carried out in the described composition of waiting to mark, the step that obtains a plurality of appraisal result comprises:
Obtain the specific targets data of each dimension correspondence of the described composition of waiting to mark;
Pass through formula: appraisal result=first obtains the appraisal result of described dimension correspondence with reference to mark+(second with reference to mark-first with reference to mark) * (described specific targets data-first reference index data)/first reference index data;
Wherein, described second with reference to mark greater than first with reference to mark, the first reference index data are described first specific targets data with reference to the mark correspondence, and described first is the appraisal result of the described dimension correspondence of the scoring composition in the training set with reference to mark and described second with reference to mark.
Wherein, described first is arbitrary value between the 0-99.99 with reference to mark, and described second is arbitrary value between the 1-100 with reference to mark.
Wherein, described first is 0,60,80 or 90 with reference to mark, and described second is 60,80,90 or 100 with reference to mark.
Wherein, described a plurality of appraisal result are weighted processing, the step that obtains the final appraisal result of the described composition of waiting to mark comprises:
Obtain to be described each dimension predetermined weights value;
According to the weighted value of the described dimension of each appraisal result in described a plurality of appraisal result and this appraisal result correspondence, obtain the final appraisal result of the described composition of waiting to mark.
Wherein, according to the weighted value of the described dimension of each appraisal result in described a plurality of appraisal result and this appraisal result correspondence, the step that obtains the final appraisal result of the described composition of waiting to mark comprises:
Pass through formula: the weighted value of the appraisal result * dimension 2 of the weighted value+dimension 2 of the appraisal result * dimension 1 of final appraisal result=dimension 1+... the weighted value of appraisal result * dimension n of+dimension n obtains described final appraisal result; Wherein, described n is positive integer.
Wherein, the weighted value of described each dimension correspondence is adjustable weighted value.
Wherein, said method also comprises:
Sentence is carried out in the described composition of waiting to mark cut apart, obtain a plurality of sentences;
From described each sentence, extract the language point;
Language point in described language point and the corpus is mated, if coupling is unsuccessful, provide in the described sentence comment at this language point.
Wherein, said method also comprises: for a plurality of sentences that belong to a paragraph, according to the comment at the language point of described a plurality of sentences, provide wait the to mark comment of composition of the comment of described paragraph and/or entire chapter.
Embodiments of the invention also provide a kind of English composition automatic scoring system, comprising:
Acquisition module is used for obtaining waiting to mark and writes a composition;
First processing module is used for the described composition of waiting to mark is analyzed, and obtains a plurality of dimensions;
Second processing module is used for according to each dimension, and score calculation is carried out in the described composition of waiting to mark, and obtains a plurality of appraisal result;
The 3rd processing module is used for described a plurality of appraisal result are weighted processing, obtains the described final appraisal result of marking and writing a composition waited.
The beneficial effect of technique scheme of the present invention is as follows:
In the such scheme, analyze by treating the scoring composition, obtain to reflect a plurality of dimensions of the evaluation index of this composition of waiting to mark, and treat that to described scoring composition carries out score calculation respectively by these a plurality of dimensions, obtain a plurality of appraisal result, again these a plurality of appraisal result are weighted processing, obtain the final appraisal result of this composition of waiting to mark; Than traditional composition methods of marking, the solution of the present invention can reflect the level of this composition of waiting to mark more objectively.
Description of drawings
Fig. 1 is the schematic flow sheet of english composition automatic scoring method of the present invention;
Fig. 2 is the synoptic diagram of a plurality of dimensions of the english composition in the method shown in Figure 1;
Fig. 3 is usually composition score distribution situation synoptic diagram;
Fig. 4 is the application interface synoptic diagram of the applied English composition automatic scoring system of method of the present invention;
Fig. 5 provides a kind of synoptic diagram of comment for method of the present invention;
Fig. 6 is the another kind of schematic flow sheet of method of the present invention;
Fig. 7 is the structural representation of English composition automatic scoring system of the present invention.
Embodiment
For making the technical problem to be solved in the present invention, technical scheme and advantage clearer, be described in detail below in conjunction with the accompanying drawings and the specific embodiments.
As shown in Figure 1, embodiments of the invention provide a kind of english composition automatic scoring method, comprising:
This embodiment of the present invention analyzes by treating the scoring composition, obtain to reflect a plurality of dimensions of the evaluation index of this composition of waiting to mark, and treat that to described scoring composition carries out score calculation respectively by these a plurality of dimensions, obtain a plurality of appraisal result, be that each dimension is all carried out score calculation, and obtain appraisal result at each dimension correspondence of this composition of waiting to mark, and again these a plurality of appraisal result are weighted processing, obtain the final appraisal result of this composition of waiting to mark; Than traditional composition methods of marking, the solution of the present invention can reflect the level of this composition of waiting to mark more objectively.
In another embodiment of the present invention, comprise on the basis of above-mentioned steps 11-14 that wherein, step 12 can comprise:
Step 121 is carried out lexical analysis, the analysis of sentence, structure of an article analysis and content analysis to the described composition of waiting to mark, and obtains sub-dimension, the sub-dimension about sentence, the sub-dimension of the structure of an article and the sub-dimension of content about vocabulary;
Step 122 is divided the sub-dimension of described vocabulary, the sub-dimension of sentence, the sub-dimension of the structure of an article and the sub-dimension of content respectively, obtains a plurality of dimensions.
Wherein, described sub-dimension is the achievement data of scoring composition, as the vocabulary achievement data of this scoring composition, sentence achievement data, piece of writing single structure achievement data and content achievement data; Described dimension is specific targets data of scoring composition, and the final appraisal result of described specific targets data and described scoring composition has linear dependence, monotonicity, independence, diversionary and/or balance.
Wherein, correlativity refers to: dimension must have correlativity with mark;
Monotonicity: the monotonicity of dimension must be identical or opposite with the monotonicity of mark, is the grammar mistake rate such as dimension, and the grammar mistake rate is more high, and mark is more low, is that monotonicity is opposite; If dimension is the grammer accuracy, monotonicity is identical like this, and the grammer accuracy is more high, and mark is more high;
Independence: the corresponding specific targets data of dimension must independent can be calculated, and can not depend on the specific targets data computation of other dimension correspondence;
Diversionary: some dimension can be designed to check and balance, the appraisal result height of this dimension correspondence, and the appraisal result of that dimension correspondence is just low, just checks and balance such as paragraph number and average two dimensions of segment length;
Balance: composition is the synthesis of a balance, has been difficult to perfect composition, so will note balance on dimension is selected.
Particularly, respectively to the sub-dimension of vocabulary, the sub-dimension of sentence, the sub-dimension of the structure of an article and the sub-dimension of content are divided, obtain a plurality of dimensions, can be as shown in Figure 2, in the embodiments of the invention, have 192 dimensions, as this paper sentence number (being a dimension about sentence), this paper word number (being a dimension about vocabulary), this paper token-category number (being a dimension about vocabulary) ... average word is long, length is greater than 9 word number percent etc., and certainly, it is 192 that the dimension number in the embodiments of the invention is not limited to, can also be more, also can be the part dimension of selecting as required wherein.
Certainly, when the scoring composition was marked, the dimension that participates in calculating was more many, and the final appraisal result that obtains is exactly more objective.
In another embodiment of the present invention, comprise that on the basis of above-mentioned steps 11-14, above-mentioned steps 13 can comprise:
Step 131 obtains the specific targets data of each dimension correspondence of the described composition of waiting to mark;
Step 132, pass through formula: appraisal result=first obtains the appraisal result of described dimension correspondence with reference to mark+(second with reference to mark-first with reference to mark) * (described specific targets data-first reference index data)/first reference index data;
Wherein, described second with reference to mark greater than first with reference to mark, the first reference index data are described first specific targets data with reference to the mark correspondence, and described first is the appraisal result of the described dimension correspondence of the scoring composition in the training set with reference to mark and described second with reference to mark.
Wherein, described first is arbitrary value between the 0-99.99 with reference to mark, and described second is arbitrary value between the 1-100 with reference to mark.Preferably, described first is 0,60,80 or 90 with reference to mark, and described second is 60,80,90 or 100 with reference to mark.
Among this embodiment of the present invention, being based on five point location point systems treats scoring composition and marks, the theme mark of a common class presents standardized normal distribution, as shown in Figure 3, be that two broad in the middle is little, intermediate section many and low divide and high score few, suppose to define hang down be divided into 60 fens, high score is 90 minutes, was equally divided into 80 fens, then account for the overwhelming majority in 60 numbers of assigning between 90 minutes, generally more than 90%;
Five point location point systems just are being based on this brass tacks, and the composition of waiting to mark is parsed into different dimensions, as average word long (AWL, below all abbreviate AWL as), mean sentence length, vocabulary richness etc., independently composition are given a mark at each dimension.
Such as AWL, from a collection of training set of accomplishing fluently branch, train parameter as shown in the table respectively:
| Mark | Average word long (AWL) |
| 0 | 3.50 |
| 60 | 4.10 |
| 80 | 4.56 |
| 90 | 5.10 |
| 100 | 6.00 |
Then in the marking process, if the AWL value of one piece of composition of waiting to mark is 4.72, then this piece wait to mark dimension score computation process of writing a composition on AWL is as follows:
At first 4.72 between 4.56 and 5.10, then score should 80 minutes and 90 minutes between;
Being calculated as follows of concrete score: the appraisal result=80+ of AWL correspondence (90-80) * (4.72-4.56)/4.56=80.35; Wherein, here first be 80 with reference to mark, second is 90 with reference to mark, the specific targets data of the AWL correspondence of the current composition of waiting to mark are 4.72, the first reference index data are 80 fens corresponding AWL in the above-mentioned table, be 4.56, the appraisal result that obtains the AWL dimension correspondence of this composition of waiting to mark is 80.35; Each dimension for this composition of waiting to mark after all calculating according to above-mentioned computing formula, obtains a plurality of appraisal result, these a plurality of appraisal result is weighted syntheticly respectively again, obtains and treats the final appraisal result of scoring composition.
Wherein, in above-described embodiment, step 14 can comprise:
Step 141 obtains to be described each dimension predetermined weights value;
Step 142 according to the weighted value of the described dimension of each appraisal result in described a plurality of appraisal result and this appraisal result correspondence, obtains the final appraisal result of the described composition of waiting to mark.
Particularly, as shown in Figure 4, each dimension is corresponding weighted value all, this weighted value is adjustable, namely after the appraisal result that obtains each dimension correspondence, when carrying out last weighting processing, can adjust or revise the weight of this each dimension correspondence;
Particularly, above-mentioned steps 142 can comprise:
Pass through formula: the weighted value of the appraisal result * dimension 2 of the weighted value+dimension 2 of the appraisal result * dimension 1 of final appraisal result=dimension 1+... the weighted value of appraisal result * dimension n of+dimension n obtains described final appraisal result; Wherein, described n is positive integer; The weighted value of described each dimension correspondence is adjustable weighted value.Like this concrete when calculating the final appraisal result of a certain composition of waiting to mark, teacher can be by adjusting weight to a certain dimension, in prior art, treat the scoring composition and calculate with a linear regression mode, this embodiment of the present invention can obtain more objectively final appraisal result.
In another embodiment of the present invention, comprise that on the basis of above-mentioned steps 11-14, described method can also comprise:
Step 15 is carried out sentence to the described composition of waiting to mark and is cut apart, and obtains a plurality of sentences;
Step 16 is extracted the language point from described each sentence;
Step 17 is mated the language point in described language point and the corpus, if coupling is unsuccessful, provides in the described sentence comment at this language point.
Preferably, can also comprise:
Step 18 for a plurality of sentences that belong to a paragraph, according to the comment at the language point of described a plurality of sentences, provides wait the to mark comment of composition of the comment of described paragraph and/or entire chapter.
Particularly, as shown in Figure 5 and Figure 6, treat scoring composition for one piece, at first system's type identification of can writing a composition, processings of making pauses in reading unpunctuated ancient writings then namely treated that scoring is write a composition and cut apart described, obtains a plurality of sentences; Carry out grammatical analysis then, namely extract the language point from described each sentence, i.e. described language point and language point in the corpus are mated in knowledge point, if coupling gets nowhere, provide in the described sentence comment at this language point; Such as, in this sentence problem such as verb mismatch is arranged, and with in the corpus correctly or more genuine collocation point out by the mode of comment, can allow the student like this when obtaining appraisal result, the comment feedback that the system that can also obtain gives is to improve the theme level.
Preferably, after the comment that obtains each language point, can obtain the comment of this sentence according to the comment of each the language point in this sentence, after the comment that obtains each sentence, can obtain the comment of paragraph according to the comment of this each sentence, and can be further according to the comment of paragraph, the comment of sentence and/or the comment of language point, obtain the general comment of the whole composition of waiting to mark, and point out, in order to can improve student's composition level.
Wherein, language point is: the metastable element of some in the english article, as collocation, word piece, elements such as sentence pattern pattern; Summarize the fundamental type of wrong language point by these metastable elements: misapply as word; The misuse of phrase module; Mismatch; The misuse of regular collocation pattern etc.; The language point and the paragraph storehouse that have comprised all english articles in the corpus.Preferably, this corpus is continuous updating in real time, and like this, when providing final appraisal result, the relevant comment that provides (comprise sentence comment, section is commented and general comment) also is real-time continuous updating, is conducive to the student like this and improves the composition level according to these comments.
The described method of the above embodiment of the present invention, can be applied to an online composition points-scoring system, by contrast student's mark composition and the method for English corpus waited, correct student's English automatically and write a composition, automatically calculate the composition score, and the quality of theme estimated, point out problem, and provide suggestion for revision, the student can revise exercise repeatedly according to suggestion, submit to and analyze, improve writing level in this way, improve the English exam achievement; The present invention simultaneously also has " English usage tunnel degree " analyzing and testing function, and whether can detect theme genuine, helps the English learner to grasp correct english expression, reduces the use of Chinglish.
As shown in Figure 7, embodiments of the invention also provide a kind of English composition automatic scoring system, comprising:
The 3rd processing module 64 is used for described a plurality of appraisal result are weighted processing, obtains the described final appraisal result of marking and writing a composition waited.
This system embodiment of the present invention is same, analyze by treating the scoring composition, obtain to reflect a plurality of dimensions of the evaluation index of this composition of waiting to mark, and treat that to described scoring composition carries out score calculation respectively by these a plurality of dimensions, obtain a plurality of appraisal result, again these a plurality of appraisal result are weighted processing, obtain the final appraisal result of this composition of waiting to mark; Than traditional composition methods of marking, the solution of the present invention can reflect the level of this composition of waiting to mark more objectively.
Preferably, first processing module 62 specifically is used for, lexical analysis, the analysis of sentence, structure of an article analysis and content analysis are carried out in the described composition of waiting to mark, obtained sub-dimension, the sub-dimension about sentence, the sub-dimension of the structure of an article and the sub-dimension of content about vocabulary; Respectively the sub-dimension of vocabulary, the sub-dimension of sentence, the sub-dimension of the structure of an article and the sub-dimension of content are divided, obtain a plurality of independently dimensions.
Wherein, described dimension is specific targets data of scoring composition, and the final appraisal result of described specific targets data and described scoring composition has linear dependence, monotonicity, independence, diversionary and/or balance.
Wherein, described second processing module 63 specifically is used for, and obtains the specific targets data of each dimension correspondence of the described composition of waiting to mark; Pass through formula: appraisal result=first obtains the appraisal result of described dimension correspondence with reference to mark+(second with reference to mark-first with reference to mark) * (described specific targets data-first reference index data)/first reference index data; Wherein, described second with reference to mark greater than first with reference to mark, the first reference index data are described first specific targets data with reference to the mark correspondence, and described first is the appraisal result of the described dimension correspondence of the scoring composition in the training set with reference to mark and described second with reference to mark.
Wherein, described first is arbitrary value between the 0-99.99 with reference to mark, and described second is arbitrary value between the 1-100 with reference to mark.
Wherein, described first is 0,60,80 or 90 with reference to mark, and described second is 60,80,90 or 100 with reference to mark.
Preferably, described the 3rd processing module 64 specifically is used for, and obtains to be described each dimension predetermined weights value; According to the weighted value of the described dimension of each appraisal result in described a plurality of appraisal result and this appraisal result correspondence, obtain the final appraisal result of the described composition of waiting to mark; Particularly, pass through formula: the weighted value of the appraisal result * dimension 2 of the weighted value+dimension 2 of the appraisal result * dimension 1 of final appraisal result=dimension 1+... the weighted value of appraisal result * dimension n of+dimension n obtains described final appraisal result; Wherein, described n is positive integer.
Wherein, the weighted value of described each dimension correspondence is adjustable weighted value.
Preferably, said system can also comprise: the comment module, and be used for that sentence is carried out in the described composition of waiting to mark and cut apart, obtain a plurality of sentences; From described each sentence, extract the language point; Language point in described language point and the corpus is mated, if coupling is unsuccessful, provide in the described sentence comment at this language point.
Preferably, described comment module can also be used for for a plurality of sentences that belong to a paragraph, according to the comment at the language point of described a plurality of sentences, provides the comment of described paragraph and/or entire chapter and waits the comment of marking and writing a composition.
Said system of the present invention is that a kind of english composition intelligence based on natural language processing is corrected system, be applied to an english composition and correct net automatically, mainly towards education sector, English is on-line automatic corrects service for the English Instructor of primary school, middle school, university, postgraduate and English teaching mechanism provides, and helps the student to improve writing level.Utilize english composition to correct net automatically and from 192 dimensions comprehensive scoring is carried out in theme, alleviate the workload of teacher comment composition; Utilize english composition to correct net automatically theme is carried out commenting by fullstop, write situation from grammer, semantic angle prompting theme, help the student to improve the composition mark; Help the English learner to grasp correct english expression mode, reduce the use of Chinglish; Utilize english composition to correct net automatically and realize that teacher's personal knowledge point is shared.
The above is preferred implementation of the present invention; should be pointed out that for those skilled in the art, under the prerequisite that does not break away from principle of the present invention; can also make some improvements and modifications, these improvements and modifications also should be considered as protection scope of the present invention.
Claims (12)
1. an english composition automatic scoring method is characterized in that, comprising:
Obtain and treat the scoring composition;
The described composition of waiting to mark is analyzed, obtained a plurality of dimensions;
According to each dimension, score calculation is carried out in the described composition of waiting to mark, obtain a plurality of appraisal result;
Described a plurality of appraisal result are weighted processing, obtain the final appraisal result of the described composition of waiting to mark.
2. english composition automatic scoring method according to claim 1 is characterized in that, the described composition of waiting to mark is analyzed, and the step that obtains a plurality of dimensions comprises:
Lexical analysis, the analysis of sentence, structure of an article analysis and content analysis are carried out in the described composition of waiting to mark, obtained sub-dimension, the sub-dimension about sentence, the sub-dimension of the structure of an article and the sub-dimension of content about vocabulary;
Respectively the sub-dimension of described vocabulary, the sub-dimension of sentence, the sub-dimension of the structure of an article and the sub-dimension of content are divided, obtain a plurality of dimensions.
3. english composition automatic scoring method according to claim 1 and 2, it is characterized in that, described dimension is specific targets data of scoring composition, and the final appraisal result of described specific targets data and described scoring composition has linear dependence, monotonicity, independence, diversionary and/or balance.
4. english composition automatic scoring method according to claim 3 is characterized in that, according to each dimension, score calculation is carried out in the described composition of waiting to mark, and the step that obtains a plurality of appraisal result comprises:
Obtain the specific targets data of each dimension correspondence of the described composition of waiting to mark;
Pass through formula: appraisal result=first obtains the appraisal result of described dimension correspondence with reference to mark+(second with reference to mark-first with reference to mark) * (described specific targets data-first reference index data)/first reference index data;
Wherein, described second with reference to mark greater than first with reference to mark, the first reference index data are described first specific targets data with reference to the mark correspondence, and described first is the appraisal result of the described dimension correspondence of the scoring composition in the training set with reference to mark and described second with reference to mark.
5. english composition automatic scoring method according to claim 4 is characterized in that, described first is arbitrary value between the 0-99.99 with reference to mark, and described second is arbitrary value between the 1-100 with reference to mark.
6. english composition automatic scoring method according to claim 5 is characterized in that, described first is 0,60,80 or 90 with reference to mark, and described second is 60,80,90 or 100 with reference to mark.
7. english composition automatic scoring method according to claim 4 is characterized in that, described a plurality of appraisal result are weighted processing, and the step that obtains the final appraisal result of the described composition of waiting to mark comprises:
Obtain to be described each dimension predetermined weights value;
According to the weighted value of the described dimension of each appraisal result in described a plurality of appraisal result and this appraisal result correspondence, obtain the final appraisal result of the described composition of waiting to mark.
8. english composition automatic scoring method according to claim 7, it is characterized in that, according to the weighted value of the described dimension of each appraisal result in described a plurality of appraisal result and this appraisal result correspondence, the step that obtains the final appraisal result of the described composition of waiting to mark comprises:
Pass through formula: the weighted value of the appraisal result * dimension 2 of the weighted value+dimension 2 of the appraisal result * dimension 1 of final appraisal result=dimension 1+... the weighted value of appraisal result * dimension n of+dimension n obtains described final appraisal result; Wherein, described n is positive integer.
9. according to claim 7 or 8 described english composition automatic scoring methods, it is characterized in that the weighted value of described each dimension correspondence is adjustable weighted value.
10. english composition automatic scoring method according to claim 1 is characterized in that, also comprises:
Sentence is carried out in the described composition of waiting to mark cut apart, obtain a plurality of sentences;
From described each sentence, extract the language point;
Language point in described language point and the corpus is mated, if coupling is unsuccessful, provide in the described sentence comment at this language point.
11. english composition automatic scoring method according to claim 10, it is characterized in that, also comprise: for a plurality of sentences that belong to a paragraph, according to the comment at the language point of described a plurality of sentences, provide wait the to mark comment of composition of the comment of described paragraph and/or entire chapter.
12. an English composition automatic scoring system is characterized in that, comprising:
Acquisition module is used for obtaining waiting to mark and writes a composition;
First processing module is used for the described composition of waiting to mark is analyzed, and obtains a plurality of dimensions;
Second processing module is used for according to each dimension, and score calculation is carried out in the described composition of waiting to mark, and obtains a plurality of appraisal result;
The 3rd processing module is used for described a plurality of appraisal result are weighted processing, obtains the described final appraisal result of marking and writing a composition waited.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201210049982.4A CN103294660B (en) | 2012-02-29 | 2012-02-29 | A kind of english composition automatic scoring method and system |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201210049982.4A CN103294660B (en) | 2012-02-29 | 2012-02-29 | A kind of english composition automatic scoring method and system |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN103294660A true CN103294660A (en) | 2013-09-11 |
| CN103294660B CN103294660B (en) | 2015-09-16 |
Family
ID=49095555
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201210049982.4A Active CN103294660B (en) | 2012-02-29 | 2012-02-29 | A kind of english composition automatic scoring method and system |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN103294660B (en) |
Cited By (16)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103824482A (en) * | 2014-03-18 | 2014-05-28 | 西华大学 | Online examination system and method |
| CN105183712A (en) * | 2015-08-27 | 2015-12-23 | 北京时代焦点国际教育咨询有限责任公司 | Method and apparatus for scoring English composition |
| CN105183713A (en) * | 2015-08-27 | 2015-12-23 | 北京时代焦点国际教育咨询有限责任公司 | English composition automatic correcting method and system |
| CN105448151A (en) * | 2015-12-23 | 2016-03-30 | 冀付军 | Method for carrying out teaching formative evaluation by using Wi-Fi laptop computer |
| CN107506360A (en) * | 2016-06-14 | 2017-12-22 | 科大讯飞股份有限公司 | A kind of essay grade method and system |
| CN108040213A (en) * | 2017-12-05 | 2018-05-15 | 北京小米移动软件有限公司 | The method and apparatus of shooting image |
| CN108280065A (en) * | 2017-01-05 | 2018-07-13 | 广州讯飞易听说网络科技有限公司 | A kind of foreign language text evaluation method and device |
| CN108595410A (en) * | 2018-03-19 | 2018-09-28 | 小船出海教育科技（北京）有限公司 | The automatic of hand-written composition corrects method and device |
| CN109272262A (en) * | 2018-11-26 | 2019-01-25 | 广州努比互联网科技有限公司 | A kind of analysis method of natural language feature |
| CN109285404A (en) * | 2018-10-25 | 2019-01-29 | 安徽创见未来教育科技有限公司 | A kind of English composition automatic scoring system |
| CN109376355A (en) * | 2018-10-08 | 2019-02-22 | 上海起作业信息科技有限公司 | English word and sentence screening technique, device, storage medium and electronic equipment |
| CN110688839A (en) * | 2019-07-02 | 2020-01-14 | 安徽七天教育科技有限公司 | Method for automatically reviewing English composition based on test paper |
| CN110795538A (en) * | 2019-10-30 | 2020-02-14 | 腾讯科技（深圳）有限公司 | Text scoring method based on artificial intelligence and related equipment |
| CN111914532A (en) * | 2020-09-14 | 2020-11-10 | 北京阅神智能科技有限公司 | Chinese composition scoring method |
| CN112527968A (en) * | 2020-12-22 | 2021-03-19 | 大唐融合通信股份有限公司 | Composition review method and system based on neural network |
| CN111914532B (en) * | 2020-09-14 | 2024-05-03 | 北京阅神智能科技有限公司 | Chinese composition scoring method |
Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN1700200A (en) * | 2005-05-30 | 2005-11-23 | 梁茂成 | English composition automatic scoring system |
| US20060194183A1 (en) * | 2005-02-28 | 2006-08-31 | Yigal Attali | Method of model scaling for an automated essay scoring system |
| CN102279844A (en) * | 2011-08-31 | 2011-12-14 | 中国科学院自动化研究所 | Method and system for automatically testing Chinese composition |
-
2012
- 2012-02-29 CN CN201210049982.4A patent/CN103294660B/en active Active
Patent Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20060194183A1 (en) * | 2005-02-28 | 2006-08-31 | Yigal Attali | Method of model scaling for an automated essay scoring system |
| CN1700200A (en) * | 2005-05-30 | 2005-11-23 | 梁茂成 | English composition automatic scoring system |
| CN102279844A (en) * | 2011-08-31 | 2011-12-14 | 中国科学院自动化研究所 | Method and system for automatically testing Chinese composition |
Cited By (22)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103824482A (en) * | 2014-03-18 | 2014-05-28 | 西华大学 | Online examination system and method |
| CN105183712A (en) * | 2015-08-27 | 2015-12-23 | 北京时代焦点国际教育咨询有限责任公司 | Method and apparatus for scoring English composition |
| CN105183713A (en) * | 2015-08-27 | 2015-12-23 | 北京时代焦点国际教育咨询有限责任公司 | English composition automatic correcting method and system |
| CN105448151B (en) * | 2015-12-23 | 2018-07-24 | 冀付军 | A method of carrying out teaching forming evaluation using Wi-Fi notebooks |
| CN105448151A (en) * | 2015-12-23 | 2016-03-30 | 冀付军 | Method for carrying out teaching formative evaluation by using Wi-Fi laptop computer |
| CN107506360A (en) * | 2016-06-14 | 2017-12-22 | 科大讯飞股份有限公司 | A kind of essay grade method and system |
| CN107506360B (en) * | 2016-06-14 | 2020-09-11 | 科大讯飞股份有限公司 | Article scoring method and system |
| CN108280065A (en) * | 2017-01-05 | 2018-07-13 | 广州讯飞易听说网络科技有限公司 | A kind of foreign language text evaluation method and device |
| CN108280065B (en) * | 2017-01-05 | 2021-12-14 | 广州讯飞易听说网络科技有限公司 | Foreign text evaluation method and device |
| CN108040213A (en) * | 2017-12-05 | 2018-05-15 | 北京小米移动软件有限公司 | The method and apparatus of shooting image |
| CN108040213B (en) * | 2017-12-05 | 2020-11-27 | 北京小米移动软件有限公司 | Method and apparatus for photographing image and computer-readable storage medium |
| CN108595410A (en) * | 2018-03-19 | 2018-09-28 | 小船出海教育科技（北京）有限公司 | The automatic of hand-written composition corrects method and device |
| CN109376355A (en) * | 2018-10-08 | 2019-02-22 | 上海起作业信息科技有限公司 | English word and sentence screening technique, device, storage medium and electronic equipment |
| CN109285404A (en) * | 2018-10-25 | 2019-01-29 | 安徽创见未来教育科技有限公司 | A kind of English composition automatic scoring system |
| CN109272262A (en) * | 2018-11-26 | 2019-01-25 | 广州努比互联网科技有限公司 | A kind of analysis method of natural language feature |
| CN109272262B (en) * | 2018-11-26 | 2022-04-01 | 广州努比互联网科技有限公司 | Method for analyzing natural language features |
| CN110688839A (en) * | 2019-07-02 | 2020-01-14 | 安徽七天教育科技有限公司 | Method for automatically reviewing English composition based on test paper |
| CN110795538A (en) * | 2019-10-30 | 2020-02-14 | 腾讯科技（深圳）有限公司 | Text scoring method based on artificial intelligence and related equipment |
| CN110795538B (en) * | 2019-10-30 | 2023-09-26 | 腾讯科技（深圳）有限公司 | Text scoring method and related equipment based on artificial intelligence |
| CN111914532A (en) * | 2020-09-14 | 2020-11-10 | 北京阅神智能科技有限公司 | Chinese composition scoring method |
| CN111914532B (en) * | 2020-09-14 | 2024-05-03 | 北京阅神智能科技有限公司 | Chinese composition scoring method |
| CN112527968A (en) * | 2020-12-22 | 2021-03-19 | 大唐融合通信股份有限公司 | Composition review method and system based on neural network |
Also Published As
| Publication number | Publication date |
|---|---|
| CN103294660B (en) | 2015-09-16 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN103294660B (en) | A kind of english composition automatic scoring method and system | |
| Lee et al. | Effects of using machine translation to mediate the revision process of Korean university students’ academic writing | |
| Ahangari et al. | Exploring the teachers’ use of spoken corrective feedback in teaching Iranian EFL learners at different levels of proficiency | |
| CN108319581B (en) | Natural language sentence evaluation method and device | |
| Tandikombong et al. | Grammatical errors in the English translation made by the students of English study program of UKI Toraja | |
| CN108280065B (en) | Foreign text evaluation method and device | |
| CN106856095A (en) | The voice quality evaluating system that a kind of phonetic is combined into syllables | |
| Gomaa et al. | Arabic short answer scoring with effective feedback for students | |
| Yoke et al. | the use of online corrective feedback in academic writing by L1 Malay learners. | |
| Weiss et al. | Assessing sentence readability for German language learners with broad linguistic modeling or readability formulas: When do linguistic insights make a difference? | |
| Latifah et al. | An Analysis of Translation Shift in Novel Shine by Jessica Jung and Its Translation | |
| Putrayasa | Syntactical analysis of essays | |
| Kunilovskaya | How far do we agree on the quality of translation? | |
| Yoon et al. | A comparison of grammatical proficiency measures in the automated assessment of spontaneous speech | |
| Li et al. | An effective automated essay scoring system using support vector regression | |
| Ariely et al. | First steps towards nlp-based formative feedback to improve scientific writing in hebrew | |
| Dascalu et al. | Reflecting comprehension through French textual complexity factors | |
| Koby | The ATA flowchart and framework as a differentiated error-marking scale in translation teaching | |
| Nedungadi et al. | Unsupervised word sense disambiguation for automatic essay scoring | |
| CN107436863A (en) | The evaluating method and device of English discourse readability degree | |
| Hashemi et al. | A Comparison of MT Errors and ESL Errors. | |
| Bhat et al. | Shallow analysis based assessment of syntactic complexity for automated speech scoring | |
| Panjaitan et al. | An Analysis of Grammatical Errors from Indonesian into English in Translating Descriptive Texts | |
| Rama et al. | Are pre-trained text representations useful for multilingual and multi-dimensional language proficiency modeling? | |
| Flanagan et al. | Classification and clustering english writing errors based on native language |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| C14 | Grant of patent or utility model | ||
| GR01 | Patent grant | ||
| C41 | Transfer of patent application or patent right or utility model | ||
| TR01 | Transfer of patent right |
Effective date of registration: 20170214 Address after: 100876 Beijing city Haidian District Xitucheng Road No. 10, building 21, room 216 Patentee after: Beijing word Network Technology Co.,Ltd. Address before: 100876 Beijing, Beijing University of Posts and Telecommunications,, white building, room 216 Patentee before: Zhang Yue |
|
| TR01 | Transfer of patent right | ||
| TR01 | Transfer of patent right |
Effective date of registration: 20180209 Address after: Hankou Road, Gulou District of Nanjing city in Jiangsu province 210093 Nanjing University No. 22 Beiyuan Gulou campus teaching building ten building B1015 room Patentee after: NANJING CIKU NETWORK INFORMATION TECHNOLOGY CO.,LTD. Address before: 100876 Beijing city Haidian District Xitucheng Road No. 10, building 21, room 216 Patentee before: Beijing word Network Technology Co.,Ltd. |
|
| TR01 | Transfer of patent right | ||
| TR01 | Transfer of patent right |
Effective date of registration: 20200507 Address after: 7F, high tech Zone building, No.1, anqiao Road, Hanjiang District, Yangzhou City, Jiangsu Province Patentee after: Jiangsu Internet Technology Co.,Ltd. Address before: Hankou Road, Gulou District of Nanjing city in Jiangsu province 210093 Nanjing University No. 22 Beiyuan Gulou campus teaching building ten building B1015 room Patentee before: NANJING CIKU NETWORK INFORMATION TECHNOLOGY Co.,Ltd. |
|
| TR01 | Transfer of patent right | ||
| TR01 | Transfer of patent right |
Effective date of registration: 20240305 Address after: Room B1015, 10th Floor, North Campus Teaching Complex Building, Nanjing University Gulou Campus, No. 22 Hankou Road, Gulou District, Nanjing City, Jiangsu Province, 210093 Patentee after: NANJING CIKU NETWORK INFORMATION TECHNOLOGY CO.,LTD. Country or region after: China Address before: Room 7F, High tech Zone Building, No.1 Anqiao Road, Hanjiang District, Yangzhou City, Jiangsu Province, 225100 Patentee before: Jiangsu Internet Technology Co.,Ltd. Country or region before: China |