阅读说明：本技术 一种中文作文句评输出方法 (Chinese sentence evaluation output method ) 是由 杨林 雷思东 于 2021-08-31 设计创作，主要内容包括：本发明涉及一种中文作文句评输出方法,通过基于不同体裁的作文的评价标准构建作文的分体裁检测维度表；对目标作文进行分句并确定每句的句子主干；获取分体裁检测维度表中和目标作文体裁对应的检测项,基于检测项对每句的句子主干进行维度检测；对检测出的每句的问题进行探测,并基于句评评语库给出每句相应的评语。实现了基于句子修辞手法、技法、常见问题等方面检测的评价方法,扩大了评价维度能够根据文章体裁的不同采用不同的评价维度和策略,进而提升了评价的全面性和准确率。(The invention relates to a Chinese sentence evaluation output method, which constructs a composition split-type cutting detection dimension table based on the evaluation standards of compositions of different genres; dividing the target composition into sentences and determining a sentence backbone of each sentence; acquiring detection items corresponding to the detection dimension table of the segments and the target composition and the style, and performing dimension detection on the sentence trunks of each sentence based on the detection items; and detecting the detected problem of each sentence, and giving out a corresponding comment of each sentence based on the sentence comment library. The evaluation method based on sentence retrieval methods, skills, common problems and other aspects detection is realized, the evaluation dimension is enlarged, different evaluation dimensions and strategies can be adopted according to different article genres, and further the comprehensiveness and accuracy of evaluation are improved.)

1. A Chinese sentence evaluation output method is characterized by comprising the following steps:

constructing a composition split-type cutting detection dimension table based on evaluation criteria of compositions of different genres;

dividing the target composition into sentences and determining a sentence backbone of each sentence;

acquiring detection items corresponding to the target composition and the language in the split-type cutting detection dimension table, and performing dimension detection on sentence trunks of each sentence based on the detection items;

and detecting the detected problem of each sentence, and giving out a corresponding comment of each sentence based on the sentence comment library.

2. The Chinese composition sentence evaluation output method of claim 1, wherein the genre detection dimension table includes at least dimension detection items corresponding to written human composition, written scenery class description, narrative class description, form class description, imagination class description, narrative text, application text and discussion paper.

3. The chinese composition sentence evaluation output method of claim 2, wherein the dividing the target composition and determining the sentence stem of each sentence comprises:

and determining the core of the sentence based on syntactic analysis, and determining parts having a dominance relation, a moving object relation, a mediating object relation and an association relation with the core to form a main part of the sentence.

4. The Chinese sentence evaluation output method of claim 3, wherein the dimension detection item includes a character profile, and the method for detecting the character profile includes:

constructing a word bank for representing a person, a word bank for representing the body part of the person and a word bank for representing the appearance description, and determining the appearance description of the person if the sentence components meet the following conditions:

the words representing the body parts of the characters and the words representing the people are in the main stems of sentences of the sentences, and a centering relation is formed between the words representing the people and the words representing the body parts;

if no centering relation exists, searching words in association relation with words representing human body parts, and determining the centering relation;

the words representing the appearance description are in the main stems of sentences of the sentences, and the words representing people and the words representing the appearance description form a main-predicate relationship or a central line relationship;

if there is no main and auxiliary relation or central line relation, searching and expressing the words of the character body part with the associated relation, and determining the main and auxiliary relation.

5. The Chinese sentence evaluation output method of claim 3, wherein the dimension detection item includes character description, and the character description detection method includes:

constructing a word stock representing human and a word stock representing character description, and determining the character description of the human if sentence components meet the following conditions:

the words representing the human and the words representing the characters are in the sentence trunks of the sentences, and the words representing the human and the words representing the characters have a cardinal relationship;

if there is no main-meaning relation, looking up and expressing the words with relation to the person's words, and determining the main-meaning relation.

6. The Chinese sentence evaluation output method of claim 3, wherein the dimension detection item includes a person psychological description, and the method for detecting the person psychological description includes:

constructing a word stock representing the inner heart of the person, a word stock representing the body part, a word stock representing the person and a word stock representing the psychological depiction, and determining the psychological depiction of the person if the sentence components meet the following conditions:

the words representing the human mind and the words representing the human are in the main stems of sentences of the sentences, and the words representing the human mind and the words representing the human form a centering relationship;

if no centering relation exists, searching for words which have an association relation with the words representing the body parts, and determining the centering relation;

the words representing the psychological description and the words representing the human are in the main stems of sentences of the sentences, and the words representing the psychological description and the words representing the human have a main-meaning relationship;

if the main and predicate relations do not exist, searching and expressing words of the human body part with the association relation, and determining the main and predicate relations;

and if the words representing the psychological depiction have a guest-moving relationship, determining words representing persons of the corresponding main-predicate relationship if the words representing the psychological depiction do not have a verb relationship.

7. The Chinese sentence evaluation and output method of claim 3, wherein the dimension detection item includes repeated and interlocked retrieval, and wherein the repeated retrieval detection method includes:

dividing the sentence into clauses according to commas and semicolons, and if the sentence contains two continuous identical clauses; and if the two clauses are not in the quotation marks, the clauses are repeatedly repaired;

the detection method of the linkage correction comprises the following steps: a sentence is divided into clauses according to commas, and if the clauses contain two continuous components beginning with the same verb, the sentence is a linkage correction.

8. The Chinese sentence evaluation output method of claim 3, wherein the dimension detection items include a sense description, the sense description includes an auditory sense, an olfactory sense, a gustatory sense, and a tactile sense, and the detection method of the sense description includes:

and constructing a corresponding perception descriptor library, wherein if the corresponding perception descriptor in the sentence is in the trunk of the sentence and the number of words is more than 1, the sentence is a perception description.

9. The chinese sentence evaluation output method of claim 3, wherein the dimension detection items include metaphor projects, and the metaphor project detection method includes:

nouns or pronouns in the sentences and words with verb parts of speech form a main-meaning relationship;

the word with verb part of speech and the noun form an actor relationship or an intermediary relationship, no negative word is arranged before the word with verb part of speech, and the number of the nouns forming the actor relationship and the intermediary relationship is less than 2.

10. The Chinese sentence evaluation output method of claim 3, wherein the dimension detection item includes a question description, and the question description detection method includes:

and performing corresponding regular matching on the basis of corresponding problem words to determine the problem sentences.

Technical Field

The invention belongs to the technical field of language processing, and particularly relates to a Chinese sentence evaluation output method.

Background

The increasing perfection of the knowledge representation method provides a technical reserve for constructing a bottom knowledge base, the increasing maturity of a syntactic analysis technology in natural language processing enables us to analyze sentence components of a sentence more accurately, and the increasing strength of a deep learning algorithm provides a more advanced text knowledge representation method, so that the us can extract deeper semantic features besides extracting text shallow features.

The existing sentence evaluation output scheme is realized by a word matching mode, the matching rule of the adopted sentence evaluation output scheme is simple, the number of detection points is small, the comparison is extensive, the sentence evaluation output is relatively thin, and the accuracy rate is not high.

Disclosure of Invention

In order to solve the problems of thin sentence list and low accuracy rate of sentence assessment in the prior art, the invention provides a Chinese sentence assessment output method which has the characteristics of more comprehensive assessment, higher accuracy rate and the like.

The Chinese sentence evaluation output method according to the specific embodiment of the invention comprises the following steps:

constructing a composition split-type cutting detection dimension table based on evaluation criteria of compositions of different genres;

dividing the target composition into sentences and determining a sentence backbone of each sentence;

acquiring detection items corresponding to the target composition and the language in the split-type cutting detection dimension table, and performing dimension detection on sentence trunks of each sentence based on the detection items;

and detecting the detected problem of each sentence, and giving out a corresponding comment of each sentence based on the sentence comment library.

Further, the genre detection dimension table at least comprises dimension detection items corresponding to written human composition, written scenery class description, narrative class description, form class description, imagination class description, narrative text, application text and narrative paper.

Further, the sentence dividing the target composition and determining the sentence stem of each sentence includes:

and determining the core of the sentence based on syntactic analysis, and determining parts having a dominance relation, a moving object relation, a mediating object relation and an association relation with the core to form a main part of the sentence.

Further, the dimension detection item includes a person profile, and the detection method of the person profile includes:

constructing a word bank for representing a person, a word bank for representing the body part of the person and a word bank for representing the appearance description, and determining the appearance description of the person if the sentence components meet the following conditions:

the words representing the body parts of the characters and the words representing the people are in the main stems of sentences of the sentences, and a centering relation is formed between the words representing the people and the words representing the body parts;

if no centering relation exists, searching words in association relation with words representing human body parts, and determining the centering relation;

the words representing the appearance description are in the main stems of sentences of the sentences, and the words representing people and the words representing the appearance description form a main-predicate relationship or a central line relationship;

if there is no main and auxiliary relation or central line relation, searching and expressing the words of the character body part with the associated relation, and determining the main and auxiliary relation.

Further, the dimension detection item includes a character sketch, and the detection method of the character sketch includes:

constructing a word stock representing human and a word stock representing character description, and determining the character description of the human if sentence components meet the following conditions:

the words representing the human and the words representing the characters are in the sentence trunks of the sentences, and the words representing the human and the words representing the characters have a cardinal relationship;

if there is no main-meaning relation, looking up and expressing the words with relation to the person's words, and determining the main-meaning relation.

Furthermore, the dimension detection item includes a person psychography, and the detection method of the person psychography includes:

constructing a word stock representing the inner heart of the person, a word stock representing the body part, a word stock representing the person and a word stock representing the psychological depiction, and determining the psychological depiction of the person if the sentence components meet the following conditions:

the words representing the human mind and the words representing the human are in the main stems of sentences of the sentences, and the words representing the human mind and the words representing the human form a centering relationship;

if no centering relation exists, searching for words which have an association relation with the words representing the body parts, and determining the centering relation;

the words representing the psychological description and the words representing the human are in the main stems of sentences of the sentences, and the words representing the psychological description and the words representing the human have a main-meaning relationship;

if the main and predicate relations do not exist, searching and expressing words of the human body part with the association relation, and determining the main and predicate relations;

and if the words representing the psychological depiction have a guest-moving relationship, determining words representing persons of the corresponding main-predicate relationship if the words representing the psychological depiction do not have a verb relationship.

Furthermore, the dimension detection item includes repeated and linked retrieval, wherein the detection method of repeated retrieval includes:

dividing the sentence into clauses according to commas and semicolons, and if the sentence contains two continuous identical clauses; and if the two clauses are not in the quotation marks, the clauses are repeatedly repaired;

the detection method of the linkage correction comprises the following steps: a sentence is divided into clauses according to commas, and if the clauses contain two continuous components beginning with the same verb, the sentence is a linkage correction.

Further, the dimension detection item comprises a sense description, the sense description comprises hearing, smell, taste and touch, and the detection method of the sense description comprises the following steps:

and constructing a corresponding perception descriptor library, wherein if the corresponding perception descriptor in the sentence is in the trunk of the sentence and the number of words is more than 1, the sentence is a perception description.

Furthermore, the dimension detection item comprises metaphorical words, and the detection method of the metaphorical words comprises the following steps:

nouns or pronouns in the sentences and words with verb parts of speech form a main-meaning relationship;

the word with verb part of speech and the noun form an actor relationship or an intermediary relationship, no negative word is arranged before the word with verb part of speech, and the number of the nouns forming the actor relationship and the intermediary relationship is less than 2.

Further, the dimension detection item includes question description, and the detection method of the question description includes:

and performing corresponding regular matching on the basis of corresponding problem words to determine the problem sentences.

The invention has the beneficial effects that: constructing a composition split-type cutting detection dimension table based on evaluation criteria of compositions of different genres; dividing the target composition into sentences and determining a sentence backbone of each sentence; acquiring detection items corresponding to the detection dimension table of the segments and the target composition and the style, and performing dimension detection on the sentence trunks of each sentence based on the detection items; and detecting the detected problem of each sentence, and giving out a corresponding comment of each sentence based on the sentence comment library. The evaluation method based on sentence retrieval methods, skills, common problems and other aspects detection is realized, the evaluation dimension is enlarged, different evaluation dimensions and strategies can be adopted according to different article genres, and further the comprehensiveness and accuracy of evaluation are improved.

Drawings

In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings used in the description of the embodiments or the prior art will be briefly described below, it is obvious that the drawings in the following description are only some embodiments of the present invention, and for those skilled in the art, other drawings can be obtained according to the drawings without creative efforts.

FIG. 1 is a flow chart of a method for Chinese sentence evaluation output in accordance with an exemplary embodiment;

FIG. 2 is a block diagram of a split detection dimension table provided in accordance with an exemplary embodiment;

fig. 3 is a partial content of a question table of a sentence evaluation corpus provided according to an exemplary embodiment.

Detailed Description

In order to make the objects, technical solutions and advantages of the present invention more apparent, the technical solutions of the present invention will be described in detail below. It is to be understood that the described embodiments are merely exemplary of the invention, and not restrictive of the full scope of the invention. All other embodiments, which can be derived by a person skilled in the art from the examples given herein without any inventive step, are within the scope of the present invention.

Referring to fig. 1, an embodiment of the present invention provides a method for evaluating and outputting chinese sentences, which specifically includes the following steps:

101. constructing a composition split-type cutting detection dimension table based on evaluation criteria of compositions of different genres;

referring to fig. 2, evaluation criteria and rules can be established from multiple aspects such as revisions, techniques, deficiencies, etc. as a priori knowledge; the technical diagnosis dimension changes along with article types, such as human writing skill detection of human actions, human appearances, human languages, human psychology, human expression, human characters, five-sense portrayal, conclusion sentences, and skillful association words, insufficiency (mismatching, colloquialization and the like), and the rest is not detected, the human writing skill detection five-sense portrayal, the rest is not detected, and finally a split-type detection dimension table is formed.

102. Dividing the target composition into sentences and determining a sentence backbone of each sentence;

the relationship of each part in the sentence can be determined according to the corresponding syntactic analysis, and then the parts which jointly form the main part of the sentence are determined.

103. Acquiring detection items corresponding to the detection dimension table of the segments and the target composition and the style, and performing dimension detection on the sentence trunks of each sentence based on the detection items;

104. and detecting the detected problem of each sentence, and giving out a corresponding comment of each sentence based on the sentence comment library.

The sentence evaluation database can be used for detecting the problems of the sentences based on the problem table shown in figure 3 given in the sentence evaluation database, detecting the problems listed in the table for one sentence, constructing different types of sentence evaluation databases, and giving corresponding comments based on the results of sentence retrieval, skill and problem detection.

In specific implementation, deep learning can be combined to realize automatic generation of Chinese composition comments, more comprehensive and higher-accuracy review results are provided aiming at detection of sentence retrieval methods, skills, common problems and the like, and the weaknesses of students can be pertinently improved after writing, so that the improvement is achieved.

In some embodiments of the present invention, the plurality of different dimension detection items include dimension detection items corresponding to a writing of a human composition, a writing of a scene class description, a narrative class description, a shape class description, a fantasy class description, a narrative text, an application text, and a treatise. Firstly, judging the genre of a composition, finding a corresponding genre detection dimension table according to the genre of the composition, dividing the composition into sentences, and carrying out dimension detection in the dimension table shown in fig. 2 on each sentence.

Wherein the sentence dividing of the target composition and the determination of the sentence backbone of each sentence comprise:

and determining the core of the sentence based on syntactic analysis, and determining parts having a dominance relation, a moving object relation, a mediating object relation and an association relation with the core to form a main part of the sentence.

According to the syntactic analysis, the HED of the sentence can be found first, and then the parts having the relations of SBV (main and subordinate relation), VOB (moving object relation), POB (intervening object relation) and COO (association relation) with the HED (core relation) are found to jointly form the main part of the sentence.

In other embodiments of the present invention, the dimension detection item includes a person profile, and the method for detecting the person profile includes:

constructing a word bank for representing a person, a word bank for representing the body part of the person and a word bank for representing the appearance description, and determining the appearance description of the person if the sentence components meet the following conditions:

the words representing the body parts of the characters and the words representing the people are in the main stems of sentences of the sentences, and a centering relation is formed between the words representing the people and the words representing the body parts;

if no centering relation exists, searching words in association relation with words representing human body parts, and determining the centering relation;

the words representing the appearance description are in the main stems of sentences of the sentences, and the words representing people and the words representing the appearance description form a main-predicate relationship or a central line relationship;

if there is no main and auxiliary relation or central line relation, searching and expressing the words of the character body part with the associated relation, and determining the main and auxiliary relation.

In the specific implementation, a word of a human body part is firstly expressed and the word is in a main part; a centering relation is formed between the words representing the human and the words representing the body parts, and the words representing the human are also in the trunk part; if the centering matching is not directly found, but the word representing the part has a word with a relationship with the centering matching, finding whether the word with the relationship has the centering relationship; then represents an outlining word and this word is in the stem part; a main-predicate relation is formed between a word representing a person and a word representing appearance, and the word representing the person is also in a main part; if the main-predicate relation is not directly found, but the word representing the part has a word with a relation with the main-predicate relation, finding whether the word with the relation has a relation with the main-predicate relation or not; and finally, if the word representing the person appearance has an association relation, detecting the word in the same way as above so as to determine the person appearance.

The detection method of the character description comprises the following steps:

constructing a word stock representing human and a word stock representing character description, and determining the character description of the human if sentence components meet the following conditions:

the words representing the human and the words representing the characters are in the sentence trunks of the sentences, and the words representing the human and the words representing the characters have a cardinal relationship;

if there is no main-meaning relation, looking up and expressing the words with relation to the person's words, and determining the main-meaning relation.

In the sentence, a main-meaning relation is formed between a word representing a person and a word representing a character, and the word representing the person is also in a main part; if the main and predicate relations are not found directly, but the word representing the part has a word with a relation with the main and predicate relations, the main and predicate relations are found; and detecting whether negative words exist in the front word and the rear word of the character description word.

The detection method for the psychological depiction of the person comprises the following steps:

constructing a word stock representing the inner heart of the person, a word stock representing the body part, a word stock representing the person and a word stock representing the psychological depiction, and determining the psychological depiction of the person if the sentence components meet the following conditions:

the words representing the human mind and the words representing the human are in the main stems of sentences of the sentences, and the words representing the human mind and the words representing the human form a centering relationship;

if no centering relation exists, searching for words which have an association relation with the words representing the body parts, and determining the centering relation;

the words representing the psychological description and the words representing the human are in the main stems of sentences of the sentences, and the words representing the psychological description and the words representing the human have a main-meaning relationship;

if the main and predicate relations do not exist, searching and expressing words of the human body part with the association relation, and determining the main and predicate relations;

and if the words representing the psychological depiction have a guest-moving relationship, determining words representing persons of the corresponding main-predicate relationship if the words representing the psychological depiction do not have a verb relationship.

First, a word indicating the inner core of a character and this word is in the stem part; a centering relation is formed between the words representing the human and the words representing the body parts, and the words representing the human are also in the trunk part; if the centering matching is not directly found, but the word representing the part has a word with a relationship with the word, finding whether the word with the relationship has a centering relationship with the word; then represents a psychologically written word and this word is in the trunk portion; a main-predicate relation is formed between a word representing a person and a word representing psychological depiction, and the word representing the person is also in a main part; if the main and predicate relations are not found directly, but the word representing the part has a word with a relation with the main and predicate relations, finding out whether the word with the relation has the main and predicate relations; if the word expressed by the psychology has a moving object relationship, if the part of speech of the word is a verb, directly returning an error, and if not, judging whether a character word having a main-meaning relationship with the word exists or not.

The detection method of repeated revising comprises the following steps:

dividing the sentence into clauses according to commas and semicolons, and if the sentence contains two continuous identical clauses; and if the two clauses are not in the quotation marks, the clauses are repeatedly repaired;

the detection method of the linkage correction comprises the following steps: a sentence is divided into clauses according to commas, and if the clauses contain two continuous components beginning with the same verb, the sentence is a linkage correction.

The sensory description comprises auditory sensation, olfactory sensation, gustatory sensation and tactile sensation, and the detection method of the sensory description comprises the following steps:

and constructing a corresponding perception descriptor library, wherein if the corresponding perception descriptor in the sentence is in the trunk of the sentence and the number of words is more than 1, the sentence is a perception description.

For example: the auditory tracing detection method comprises the following steps: and constructing an acoustic description word library, wherein a word which represents the acoustic description is in the sentence, the word number of the word is more than 1, and the word is in the trunk of the sentence.

The olfaction tracing detection method comprises the following steps: and constructing an olfactory description word library, wherein a word which represents the olfactory description exists in the sentence, the word number is more than 1, and the word is in the trunk of the sentence.

The taste profile detection method comprises the following steps: constructing a taste description word bank, wherein a sentence has a word representing the smell description and the word number is more than 1, and the word is in the main stem of the sentence.

The tactile depiction detection method comprises the following steps: constructing a tactile tracing word library, wherein a sentence has a word representing odor tracing and the word number is more than 1, and the word is in the trunk of the sentence.

The metaphor detection method comprises the following steps:

nouns or pronouns in the sentences and words with verb parts of speech form a main-meaning relationship;

the word with verb part of speech and the noun form an actor relationship or an intermediary relationship, no negative word is arranged before the word with verb part of speech, and the number of the nouns forming the actor relationship and the intermediary relationship is less than 2.

Specifically, a noun or pronoun constitutes an SBV relationship with 'like', and 'like' of the part of the verb, and this noun is not 'meaning', 'i', 'powerful', 'daughter', 'blood', 'son', 'guilder', 'so-minded'; verbs of 'like', etc. and nouns constitute VOB or POB relationships; the terms 'like', etc. are not preceded by the negation of the terms 'not', 'no', 'not', no modification; the number of nouns forming VOB and POB relation behind the metaphor is less than 2; the ontology part of speech is not a pronoun.

The detection method described by the question sentence comprises the following steps:

and performing corresponding regular matching on the basis of corresponding problem words to determine the problem sentences.

For example, the question exclamation detection method includes: constructing a psychological description word bank and an emotional description word bank, wherein the sentences contain words representing psychological description or emotional description and contain! (ii) a The sentence does not contain: "" etc. to identify a utterance.

The question-back detection method comprises the following steps:

will "(how can | how difficult | how can | why not) (-)? The regular matching is used for detecting the question.

The method for detecting the expression of the last words emotion comprises the following steps: the last sentence of the article; in? Or! Ending; one of the words represents a psychological depiction.

The emotion expression detection method comprises the following steps: constructing words of 'like', 'adme', 'appreciate', 'love', etc. which represent the word stock of people and are positioned in the main stem of the sentence; a main-predicate relation is formed between a word representing a person and the word, and the word representing the person is also in a main part; if the main and predicate relations are not found directly, but the word representing the part has a word with a relation with the main and predicate relations, the main and predicate relations are found; if the word has a moving-guest relationship, if the part of speech of the word is a verb, directly returning to False, and if not, judging whether a character word having a main-predicate relationship with the word exists.

Continuous motion description detection method: finding a verb and locating in the main stem; finding a noun forming a main-predicate relation with the noun, wherein the noun is positioned at the main stem of the sentence; finding two verbs which form an association relation with the verbs, wherein the two verbs do not form a subject-predicate relation with other words; the distance between the three verbs is less than 10; any two of the three verbs are different; and the verb is a verb representing the action of the person

The method for detecting the pseudonym comprises the following steps: and constructing an acoustic word library, wherein the part of speech is marked as 'O', or words in the acoustic word library exist in the sentence.

The method for detecting the ranking sentence comprises the following steps: clauses are divided by commas or semicolons, the number of identical words after clause division is greater than 1 and the index position deviation in the sentence is less than 2.

The exaggeration detection method comprises the following steps: and constructing an exaggerated word bank, wherein the sentences contain words in the exaggerated word bank.

A theory detection method is explained: for [ "i (.. Is there a | A Is it? | A A "(# is) only present [,". Is there a | A Is it? | A H, (. Is there a | A Is it? | A After that, i feel (. Is there a | A Is it? | A "," (#) is also really not easy "," nevertheless, thought carefully (. Is there a | A Is it? | A -, "what to taste (. Is there a | A Is it? | A H, i understand (. Is there a | A Is it? | A "] these representations are regularly matched to determine statements that set forth the reasoning.

The related word detection method comprises the following steps: and constructing a related word bank, wherein front and back matching type related words in the related word bank appear in the sentence.

The scene blending detection method comprises the following steps: and constructing a scene description word bank and a psychological description word bank, wherein the sentences simultaneously contain words comprising scene description and psychological description, and the words are positioned in the trunks of the sentences.

The scene description detection method comprises the following steps: and constructing a scene description word library, wherein the sentences contain the scene description words, and the word number of the words exceeds two words.

Color description detection method: constructing a color description word library, wherein words representing color description exist in sentences; and the word number is more than 1; and this word is at the sentence stem.

The description detection method of the character lattice quality comprises the following steps: constructing a character description word stock, wherein the word stock for representing the shape is constructed to firstly represent a word described by the character and the word is in a trunk part; a main-meaning relationship is formed between a pronoun of a representation object and a word of a representation character, and the pronoun is also arranged in the main part; if the main and predicate relations are not found directly, but the word representing the part has a word with a relation with the main and predicate relations, finding out whether the word with the relation has the main and predicate relations; and detecting whether negative words exist in the front word and the rear word of the character description word.

Therefore, a set of rules is formed to drive prior knowledge, various types of word banks at the bottom layer are supported, and dependency syntactic analysis and a deep learning algorithm are combined to perform feature extraction, so that various types of revising, techniques and problem detection are performed, and further the comprehensiveness and accuracy of evaluation are improved.

It will be understood by those skilled in the art that all or part of the steps carried by the method for implementing the above embodiments may be implemented by hardware related to instructions of a program, which may be stored in a computer readable storage medium, and when the program is executed, the program includes one or a combination of the steps of the method embodiments.

In addition, functional units in the embodiments of the present invention may be integrated into one processing module, or each unit may exist alone physically, or two or more units are integrated into one module. The integrated module can be realized in a hardware mode, and can also be realized in a software functional module mode. The integrated module, if implemented in the form of a software functional module and sold or used as a stand-alone product, may also be stored in a computer readable storage medium.

The storage medium mentioned above may be a read-only memory, a magnetic or optical disk, etc.

In the description herein, references to the description of the term "one embodiment," "some embodiments," "an example," "a specific example," or "some examples," etc., mean that a particular feature, structure, material, or characteristic described in connection with the embodiment or example is included in at least one embodiment or example of the invention. In this specification, the schematic representations of the terms used above do not necessarily refer to the same embodiment or example. Furthermore, the particular features, structures, materials, or characteristics described may be combined in any suitable manner in any one or more embodiments or examples.

What has been described above includes examples of one or more embodiments. It is, of course, not possible to describe every conceivable combination of components or methodologies for purposes of describing the aforementioned embodiments, but one of ordinary skill in the art may recognize that many further combinations and permutations of various embodiments are possible. Accordingly, the embodiments described herein are intended to embrace all such alterations, modifications and variations that fall within the scope of the appended claims. Furthermore, to the extent that the term "includes" is used in either the detailed description or the claims, such term is intended to be inclusive in a manner similar to the term "comprising" as "comprising" is interpreted when employed as a transitional word in a claim. Furthermore, any use of the term "or" in the specification of the claims is intended to mean a "non-exclusive or".

The above description is only for the specific embodiments of the present invention, but the scope of the present invention is not limited thereto, and any person skilled in the art can easily conceive of the changes or substitutions within the technical scope of the present invention, and all the changes or substitutions should be covered within the scope of the present invention. Therefore, the protection scope of the present invention shall be subject to the protection scope of the appended claims.