CN109697905A - A kind of exam paper marking system - Google Patents

A kind of exam paper marking system Download PDF

Info

Publication number
CN109697905A
CN109697905A CN201710981868.8A CN201710981868A CN109697905A CN 109697905 A CN109697905 A CN 109697905A CN 201710981868 A CN201710981868 A CN 201710981868A CN 109697905 A CN109697905 A CN 109697905A
Authority
CN
China
Prior art keywords
result
answer
comments
stroke
module
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Pending
Application number
CN201710981868.8A
Other languages
Chinese (zh)
Inventor
陈滢朱
杨宁
卢鑫
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Shenzhen Eaglesoul Technology Co Ltd
Original Assignee
Shenzhen Eaglesoul Technology Co Ltd
Priority date (The priority date 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 date listed.)
Filing date
Publication date
Application filed by Shenzhen Eaglesoul Technology Co Ltd filed Critical Shenzhen Eaglesoul Technology Co Ltd
Priority to CN201710981868.8A priority Critical patent/CN109697905A/en
Priority to PCT/CN2017/111811 priority patent/WO2019075820A1/en
Publication of CN109697905A publication Critical patent/CN109697905A/en
Pending legal-status Critical Current

Links

Classifications

    • GPHYSICS
    • G09EDUCATION; CRYPTOGRAPHY; DISPLAY; ADVERTISING; SEALS
    • G09BEDUCATIONAL OR DEMONSTRATION APPLIANCES; APPLIANCES FOR TEACHING, OR COMMUNICATING WITH, THE BLIND, DEAF OR MUTE; MODELS; PLANETARIA; GLOBES; MAPS; DIAGRAMS
    • G09B7/00Electrically-operated teaching apparatus or devices working with questions and answers
    • G09B7/02Electrically-operated teaching apparatus or devices working with questions and answers of the type wherein the student is expected to construct an answer to the question which is presented or wherein the machine gives an answer to the question presented by a student
    • GPHYSICS
    • G06COMPUTING; CALCULATING OR COUNTING
    • G06FELECTRIC DIGITAL DATA PROCESSING
    • G06F18/00Pattern recognition
    • GPHYSICS
    • G06COMPUTING; CALCULATING OR COUNTING
    • G06FELECTRIC DIGITAL DATA PROCESSING
    • G06F40/00Handling natural language data
    • GPHYSICS
    • G06COMPUTING; CALCULATING OR COUNTING
    • G06QINFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
    • G06Q50/00Systems or methods specially adapted for specific business sectors, e.g. utilities or tourism
    • G06Q50/10Services
    • G06Q50/20Education
    • GPHYSICS
    • G06COMPUTING; CALCULATING OR COUNTING
    • G06VIMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
    • G06V30/00Character recognition; Recognising digital ink; Document-oriented image-based pattern recognition
    • G06V30/10Character recognition
    • G06V30/32Digital ink
    • G06V30/333Preprocessing; Feature extraction
    • G06V30/347Sampling; Contour coding; Stroke extraction
    • GPHYSICS
    • G06COMPUTING; CALCULATING OR COUNTING
    • G06VIMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
    • G06V30/00Character recognition; Recognising digital ink; Document-oriented image-based pattern recognition
    • G06V30/10Character recognition
    • G06V30/32Digital ink
    • G06V30/36Matching; Classification
    • GPHYSICS
    • G09EDUCATION; CRYPTOGRAPHY; DISPLAY; ADVERTISING; SEALS
    • G09BEDUCATIONAL OR DEMONSTRATION APPLIANCES; APPLIANCES FOR TEACHING, OR COMMUNICATING WITH, THE BLIND, DEAF OR MUTE; MODELS; PLANETARIA; GLOBES; MAPS; DIAGRAMS
    • G09B7/00Electrically-operated teaching apparatus or devices working with questions and answers
    • GPHYSICS
    • G06COMPUTING; CALCULATING OR COUNTING
    • G06VIMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
    • G06V30/00Character recognition; Recognising digital ink; Document-oriented image-based pattern recognition
    • G06V30/10Character recognition

Abstract

The present invention provides a kind of exam paper marking system, comprising: first terminal equipment, comprising: the first display module, for showing the test question of the electronic test paper to student;First input module, for the answer result for student's input to the test question;Second terminal equipment, comprising: the second display module shows the electronic test paper with the answer result for being denoted as different colours to outside the teacher that gos over examination papers;Second input module reads and makes comments the electronic test paper for the teacher that gos over examination papers;Server is communicatively connected to the first terminal equipment, second terminal equipment and paper database respectively, comprising: first reads and makes comments module, and the answer of the test question of paper database purchase and the answer result are carried out automatic comparison and tentatively read and make comments result to generate;Fractional statistics module tentatively reads and makes comments result for receiving and counting this;Second reads and makes comments module, generates for the input according to the teacher that gos over examination papers and finally reads and makes comments result.Exam paper marking system efficiency with higher of the invention and enough precision.

Description

A kind of exam paper marking system
Technical field
The present invention relates to multimedia teaching fields to relate in particular to one kind more particularly to a kind of exam paper marking system The exam paper marking system of electronic test paper for multimedia teaching.
Background technique
Known, the examination that early stage carries out generallys use the mode for teacher's group signature of going over examination papers, and the teacher that gos over examination papers is straight by score It connects and is labeled on paper, after the completion of waiting all papers to correct, united again by manually copying score in summary Meter analysis.This craft mode of going over examination papers be it is very original, working efficiency is low, and achievement transcribing errors are also inevitable, is only suitable for Minimum scale and the insignificant examinations such as class, grade's examination.
With the development of science and technology gradually proposing the concept that automation is goed over examination papers.Mode known to a kind of is by using mark Remember the automatic scoring that paper carries out, or passes through image procossing as disclosed in Japanese documentation JP-A-7-199794 Device scores to answer sheet, is configured to described image processor to come distinguished symbol and mark by using Digital Image Processing Remember and these symbols and label are compared to score to answer sheet with correct option.And such as in Japanese documentation Disclosed in JP-A-11-224245 like that, where it is proposed such a points-scoring systems, and the system is configured to by making With special-purpose cushion and pen and detects the state of pen automatic scoring is carried out to answer sheet.
However, the above-mentioned prior art has the disadvantages that.That is, institute is public using mark sheet and in JP-A-7-199794 In the case that the image processor opened carries out automatic scoring, if answer and is-not symbol, character or label by text (as answered The problem of), then answer cannot be identified by answer recognition unit, therefore cannot score answer.Further, since will Points-scoring system disclosed in JP-A-11-224245 is configured to by using special-purpose cushion and pen and detects the state of pen and come to answer Paper carries out automatic scoring, it is therefore desirable to the specialized hardware of such as special-purpose cushion and pen, therefore its structure becomes complicated and expensive.
With the further development of computer networking technology and image processing techniques, automation, which is goed over examination papers, to be further evolved into Using computer marking.Computer marking is carried out by the way of paper and test paper separation, and objective item is appraised automatically by computer, Subjective item on computers scores respectively to the electronic image of examinee's test paper by network by the different teachers that gos over examination papers, finally Carry out core point and achievement verification automatically by computer system again.The teacher that marks examination papers can be effectively prevented using computer network grading Between influence each other, moreover it is possible to the error rate that will go over examination papers minimizes.Current is all utilized in for computer marking system Anchor point is set on answering card, and the answer sheet recognition device of computer marking system is identified by induction on special answering card Anchor point identifies different topic answer locations.As an example, a kind of calculating is disclosed in CN101866415A The answer sheet recognition device of machine marking system, including data read module and character recognition module, wherein the reading data mould Block is sent to the character recognition module for obtaining answer sheet picture information, and by the answer sheet picture information;The word Identification module is accorded with, for using optical character recognition technology, the character shape of the answer sheet picture information is translated into calculating The readable text information of machine and storage.As another example, a kind of examination based on OCR identification is disclosed in CN102890775A Score automatic identification processing method is rolled up, including the marking region on automatic program identification paper, is identified and is tried using OCR technique Roll up score.Directly the hand-written score on papery paper is handled in above-mentioned document, does not need to make dedicated answering card.So And in practice, it has been found that can only enough meet that there are the objective items of model answer (such as to select using optical character recognition technology Select topic, True-False etc.), but for needing then still obviously not accurately identifying and sentence come the topic of answer using character express It is disconnected.
It is also used as another example, a kind of terminal examination evaluation system, the system packet are disclosed in CN104900101A Include server, student terminal system and the teacher's terminal system that interaction is realized by internet, wherein contain in server memory Exam pool and exam pool answer, the server send operation topic and examination paper according to the examination order that teacher's terminal system is issued To aobvious outside student terminal system, and input student's answer in student terminal system by user, then on pass server back, service Device receives student's answer and is read and made comments according to exam pool answer, then will read and make comments result and be sent to examination terminal system and student terminal system System.The system can be realized Automatic Read Overmarginalia, and provide go over examination papers speed and quality of instruction.Then, the system is not first still The topic that answer is literal expression can be read and made comments.Further, it is difficult to ensure due to automating the quality read and made comments, in the essence read and made comments Real needs are not able to satisfy in terms of exactness and accuracy.
As an improvement example, a kind of teaching machine based on electromagnetic handwritten plate is disclosed in CN106887168A Network on-line teaching system, including teacher's terminal and student terminal include to read and make comments module in teacher's terminal, described to read and make comments mould Block includes reading and making comments module and Automatic Read Overmarginalia module manually.It is described manually read and make comments module for teacher manually to student reply content into Row verification, Automatic Read Overmarginalia module are that system automatically replies student and answer compares verification.Specifically, using including original Handwriting recognition module and speech recognition module acquire the voice of person's handwriting and microphone or microphone of the handwriting pad with time writer, logical It crosses data processing module the reply message of identification is compared with correct option and go over examination papers.Manually read and make comments module be then used as it is auxiliary It helps and student's answer content is checked manually for teacher.Although introduced in the document manually read and make comments module for into Capable verification, but the hardware that the actually described system itself is inputted is prohibitively expensive, and for oneself of person's handwriting and voice It is dynamic to read and make comments the topic for being inherently unable to satisfy and reading and making comments that answer is literal expression, therefore the system is still unable to satisfy real needs.
As a kind of trial of the automatic marking processing to character property examination question, by Iflytek limited liability company as Shen A kind of answer recognition methods and system for character property examination question, the method packet are disclosed in the CN104463101A to ask someone It includes: obtaining character property script image;One or more answer character strings to be identified are obtained from the answer image;It is based on General acoustic model carries out handwriting recognition to the answer character string to be identified, obtains the first recognition result;Obtain character property The answer information of examination question;According to the answer information of first recognition result and the character property examination question, adaptive acoustic is constructed Model;Handwriting recognition is carried out to the answer character string to be identified using the adaptive acoustic model, obtains finally identifying knot Fruit.By above technical scheme, the recognition accuracy of character property objective item can effectively improve, and then improve go over examination papers efficiency and standard True property.
However, above paper automatic marking processing system do not meet the teacher that gos over examination papers in practice read and make comments habit, exist Human-computer interaction is poor, read and make comments process design with actually go over examination papers habit be separated the case where, the effect of actual use can not make us full Meaning.Further, in actually examination test, paper is different according to the subject test or field and generally comprises different topics Type, such as examine in and must examine subject with college entrance examination: for Chinese language and foreign language, it is indispensable can there are problems that for example, fill, Or the topics type such as discussion topic.The feature of reading and making comments of these topic types is: 1. and be not present no model answer, and be usually by multiple keys Key for Reference composed by word (so-called to branch or adopt branch);2. fixed format is often not present in the result of answering of student And template, there are very big othernesses, there is a situation where following such: although the result of answering of student is stating upper and standard Answer is different, but has actually grasped key answering principles substantially, at this time often according to stringent comparison in the prior art It is judged as mistake, makes the accuracy of test by serious influence;3. read and make comments paper, in addition to needing to provide examination Outside score, it is also desirable to retain the trace of reading and making comments for the teacher that gos over examination papers, such as teacher is for the note of student's answer and comment etc., So that student is it will be appreciated that the shortcoming of itself, while it also can satisfy the needs that teacher comments on paper.
Applicant it is hereby stated that, what is stated in above section is only the technology contents that are known of applicant.Above content is only Provide only the prior art for being related to the relevant background information of the disclosure but not necessarily constituting the application.
It is that present multimedia tutoring system is opened in consideration of it, how to provide a kind of exam paper marking system to meet current demand Originator problem anxious to be resolved.
Summary of the invention
At least one of deficiency exists in the prior art in order to overcome, the present invention provides a kind of exam paper marking system, For reading and making comments student to the electronic test paper answer comprising at least one test question from paper database as a result, the paper The system of reading and making comments includes: first terminal equipment, comprising: the first display module, for showing the test of the electronic test paper to student Topic;First input module, for the answer result for student's input to the test question;Second terminal equipment, comprising: second is aobvious Show module, the electronic test paper with the answer result for being denoted as different colours is shown to outside the teacher that gos over examination papers;Second input module, The electronic test paper is read and made comments for the teacher that gos over examination papers;Server is communicatively connected to the first terminal equipment, second terminal respectively Equipment and paper database, the server include first and read and make comments module, by the answer of the test question of paper database purchase Automatic comparison, which is carried out, with the answer result tentatively reads and makes comments result to generate;Fractional statistics module, it is described for receiving and counting Tentatively read and make comments result;Second reads and makes comments module, generates for the input according to the teacher that gos over examination papers and finally reads and makes comments result.
Exam paper marking system according to the present invention can assist going over examination papers teacher fast contrast testing student answer result with Difference between Key for Reference so that improve electronic test paper reads and makes comments speed and efficiency, and reduces the work for the teacher that gos over examination papers It measures and has also taken into account the accuracy for reading and making comments electronic test paper.Also, it is long that exam paper marking system of the invention is more in line with the teacher that gos over examination papers Habit that the phase formed go over examination papers.
Preferably, at least one described test question preferably includes subjective item and objective item, and described first reads and makes comments mould Block includes the complete comparing unit for reading and making comments the objective item and the character string comparison unit for reading and making comments the subjective item.
The present invention to subjective item can compare automatically and be read and made comments in the case where keeping centainly reading and making comments accuracy as a result, Speed and efficiency are read and made comments to improve electronic test paper.
Preferably, the character string comparison unit preferably includes: the first transform subblock, is used for the answer result Multiple character strings and the character string of Key for Reference of subjective item be converted into word sequence respectively;Second transform subblock, being used for will The word sequence is converted into corresponding set of words respectively;First operation submodule, for calculating the min-hash of each set of words Min-hash distance between value and set of words;Classification submodule, for being classified according to min-hash distance to word sequence; And the second operation submodule, for carrying out shared string operations in same class.
Quickly judge whether difference is excessive between answer result and Key for Reference by min-hash algorithm first as a result, It identifies with the obvious dissimilar or not approximate answer of Key for Reference as a result, carrying out shared character string to effectively reduce the later period The character string quantity of operation can be effectively saved character compared with the character string comparison unit of unused min-hash algorithm The calculation amount of string comparison operation and the calculation resources of occupancy.Secondly, by the judgement of answer result and the degree of approximation of Key for Reference It is converted into the operation of the shared length of the min-hash distance and shared string operations between set of words, rather than is used The optical mask identification technology being widely used in the prior art, so as to avoid need in the prior art high hardware cost and The deficiency of poor discrimination.
Preferably, the set of words of each answer result and the ginseng of subjective item are further calculated by the first operation submodule Min-hash distance between examining the set of words of answer two-by-two, by classification submodule by the min-hash of the set of words with Key for Reference Distance is less than word sequence corresponding to the set of words of first threshold and is divided into same class, and min-hash distance is greater than first threshold Set of words corresponding to the character string of answer result be denoted as the first color;Subsequent second operation submodule min-hash away from From shared string operations below being carried out in the same class for being less than first threshold: making the word sequence of Key for Reference in the class For the first word sequence, the shared character string of first word sequence and other word sequences in such is calculated separately, wherein gained The ratio of the string length of the length and Key for Reference of the shared character string arrived is greater than corresponding to the word sequence of second threshold The character string of answer result is denoted as the second color, and the length of obtained shared character string and the character string of Key for Reference are long The character string that the ratio of degree is less than answer result corresponding to the word sequence of second threshold is denoted as third color.
As a result, on the basis of being calculated with min-hash algorithm and shared character string, by will be approximate with Key for Reference The different answer result of degree is labeled as different colors, quickly understands and identify the feelings of answer result convenient for the teacher that gos over examination papers Condition reads and makes comments the efficiency finally read and made comments and annotated in module second to effectively improve.
Preferably, the first threshold be set of words minimum hash digit 50%, the second threshold, preferably It is 80%.
It is read and made comments thus, it is possible to carry out automatic comparison to subjective item in the case where keeping centainly reading and making comments accuracy.
Preferably, it second reads and makes comments module and is configured to receive the input from the second terminal equipment in this way: corresponding to The answer of the character string of the first color and the second color is denoted as a result, described second, which reads and makes comments module, allows to go over examination papers teacher via described Second terminal equipment adds their confirmation and/or is annotated that read and make comments module via second outside to be formed to the result of tentatively reading and making comments Result is finally read and made comments in output;Corresponding to the character string for being denoted as third color answer as a result, described second read and make comments module permit Xu Yuejuan teacher tentatively reads and makes comments result and is modified and/or annotated to form warp via the second terminal equipment to described By second read and make comments that module exports outward finally read and make comments result.
Thus, it is possible to according to the difference of answer result and Key for Reference degree of approximation come targeted to tentatively result is read and made comments Add their confirmation or modify, sufficiently accurate finally read and make comments result to export outward.
Preferably, the exam paper marking system is that per pass subjective item is equipped with affiliated score upper bound, when teacher's warp of going over examination papers By the second terminal equipment to it is described tentatively read and make comments fractional value that result modify and exceed the score upper bound when, the examination Volume reads and makes comments system and forbids the operation from the second terminal equipment and alarmed.
Thus, it is possible to which the teacher that go over examination papers will not be goed over examination papers work and mistakenly assert the class subjective item due to heavy Upper limit score value, thus maximum likelihood avoid the teacher that gos over examination papers subjective a possibility that making a mistake.
Preferably, the first terminal equipment receives to read and make comments finally reading and making comments as a result, and via described for module from second It shows outside first display module to student.
Thereby, it is possible to the traces of reading and making comments of the outer aobvious teacher that gos over examination papers, such as teacher is for the note of student's answer and comment etc. Deng so that student is it will be appreciated that the shortcoming of itself, while also can satisfy the needs that teacher comments on paper.
Preferably, wherein second input module is the handwriting recognition list for recording the handwriting input trace for the teacher that gos over examination papers Member, second display module are the touch screen with hand-write input function.
The present invention can be inputted as a result, in hand-written mode, and be more in line with the teacher that gos over examination papers in this way reads and makes comments habit.
Preferably, the hand-writing recognition unit is configured to execute the hand-written input method of reduplicated word, and the reduplicated word is hand-written Input method includes the following steps:
Firstly, the hand-writing recognition unit detects that finger contacts this event of touch screen, the start recording teacher that gos over examination papers is defeated The stroke person's handwriting of the character entered, and the handwriting area transferred on the touchscreen shows corresponding stroke track;Until its finger from Open touch screen, the end of input of current stroke;
Then, the hand-writing recognition unit judged according to following judgment basis the stroke of current handwriting input with before Whether the stroke through inputting belongs to the same character: first determining whether current stroke is newly to write pen on the right side of preceding unicursal It draws, is write if it is not, then further judging whether current stroke is folded on trace with the stroke of handwriting input before, if so, Then further judge that current stroke and the folded of stroke of handwriting input write whether degree is greater than predetermined threshold, if so, recognizing The same character is not belonging to the stroke inputted before for the stroke of current handwriting input;
If the hand-writing recognition unit judge stroke currently entered and the stroke inputted before be belong to it is same Character, then the hand-writing recognition unit persistently records the handwriting input trace for the teacher that gos over examination papers, until judging to obtain currently entered Stroke no longer belongs to the same character with the stroke inputted before, at this time by the handwritten traces of current record write into Row Text region, and recognition result is exported on the touchscreen, recognition result is then subjected to dimmed processing;
If it is same that the hand-writing recognition unit judges that stroke currently entered is not belonging to the stroke inputted before Character, then judging, which whether there is on touch screen, has write and had carried out identifying processing and carried out the handwritten word of dimmed processing Symbol then eliminates the character for having carried out identifying processing if there is there is such hand-written character;Continuing will be except when preceding pen All strokes except picture constitute a handwriting characters, carry out Text region and export recognition result on the touchscreen, so Recognition result is subjected to dimmed processing afterwards.
The present invention can allow the teacher that gos over examination papers that need not pause when writing kinds of characters as a result, can be with the multiple hands of continuous writing Write characters and identified accordingly and display processing, to improve the efficiency of handwriting input.
Detailed description of the invention
To describe the technical solutions in the embodiments of the present invention more clearly, institute in being described below to the embodiment of the present invention 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 some implementations of the invention Example, for those of ordinary skill in the art, without creative efforts, can also implement according to the present invention The content of example and these attached drawings obtain other attached drawings.
Fig. 1 is the general frame schematic diagram of exam paper marking system of the invention;
Fig. 2 is the first of the exam paper marking system of the invention architecture diagram for reading and making comments module;
Fig. 3 is the architecture diagram of character string comparison unit of the invention;With
Fig. 4 is the flow chart that character string of the invention compares automatically.
Reference signs list
10 first terminal equipment, 101 first 102 first display module of input module, 20 second terminal equipment 201 second 202 second display module of input module, 30 server 301 first is read and made comments 302 fractional statistics module 303 second of module and is read and made comments 3,011 321 first transform subblock of complete 3012 character string comparison unit of comparison unit of module, 322 second transform subblock 323 first operation submodules 324 classification 325 second operation submodule of submodule, 31 paper database
Specific embodiment
Below in conjunction with attached drawing, technical scheme in the embodiment of the invention is clearly and completely described, it is clear that institute The embodiment of description is only a part of the embodiment of the present invention, instead of all the embodiments.Usually described in attached drawing here It can arrange and design with a variety of different configurations with the component of the embodiment of the present invention shown.Therefore, below in attached drawing The detailed description of the embodiment of the present invention of middle offer is not intended to limit the range of claimed invention, but only table Show selected embodiment of the invention.Based on the embodiment of the present invention, those skilled in the art are not making creative work Under the premise of every other embodiment obtained, shall fall within the protection scope of the present invention.
With reference to Fig. 1, the general frame schematic diagram of exam paper marking system according to the present invention is shown.The paper batch System is read for reading and making comments student to the result of the electronic test paper answer comprising at least one test question.Specifically, the paper The system of reading and making comments includes consisting of: first terminal equipment 10, is communicatively connected to the first terminal at second terminal equipment 20 respectively The server 30 of equipment 10 and second terminal equipment 20, and the paper database 31 with the server 30 communication connection, In, the first terminal equipment 10 includes at least: for the first input module for student's input to the answer result of test question 101 with the first display module 102 for showing the test question of the electronic test paper to student, second terminal equipment 20 is then at least Include: the second input module 201 for reading and making comments the electronic test paper for the teacher that gos over examination papers is shown with to outside the teacher that gos over examination papers with answer knot Second display module 202 of the electronic test paper of fruit, and server 30 includes at least and carries out the answer of test question and answer result Automatic comparison tentatively reads and makes comments the first of result and reads and makes comments module 301, receives and count the score for tentatively reading and making comments result to generate It statistical module 302 and is generated according to the input for the teacher that gos over examination papers and finally reads and makes comments the second of result and read and make comments module 303.
In the present invention, first terminal equipment 10 and second terminal equipment 20 may, for example, be and is connected to via cable network The desktop computer of server 30, naturally it is also possible to be the various portable terminals communicated to connect via wireless network and server 30 End device, such as, but not limited to: first terminal equipment 10 and second terminal equipment 20 can be tablet computer, smart phone etc., And it can directly be communicated to connect with server 30 via radio network techniques such as 4G, bluetooth or WiFi.Take one's test Life can log in the exam paper marking system via network and in the form of account number cipher, take one's test and understand answering for oneself Inscribe situation.
According to a particular embodiment of the invention, the student to take one's test can be via the first terminal equipment 10 simultaneously To input in a manner of account or student number, and by the code entry exam paper marking system, so that the test of electronic test paper is carried out, and the One terminal device 10 is to be communicatively connected to paper database 31 by server 30, to receive electricity from the paper database 31 The sub- paper and student for Gong taking one's test answers.Certainly, alternatively, the electronic test paper is it is also contemplated that be It is provided by server 30 itself, i.e., server 30 connects paper database 31 to obtain electronic test paper, then by the electronic test paper The student that first terminal equipment 10 Gong takes one's test is sent to answer.
In a specific embodiment of the present invention, the electronic test paper includes at least one quiz, and the above quiz is selected From in quiz stored in paper database 31, naturally it is also possible to it is envisioned that can also be by server 30 from paper database Different quizs is selected in 31 to form the electronic test paper.Here, at least one test question in the electronic test paper is excellent Selection of land includes objective item and subjective item two types.Wherein, in the present invention, term " objective item " refers to the test The topic type of the model answer of topic, the result of the topic type read and made comments only include two kinds of answer right and wrong as a result, as non-limit Qualitative example, the topic type be for example non-topic, multiple-choice question, choose topic etc..Meanwhile " subjective item " refers to reference at least one example The Key for Reference of for example character property gives the topic type read and made comments with certain subjectivity, wherein the Key for Reference for example can be by more A keyword (i.e. so-called to branch or adopt branch) composition, wherein the result read and made comments may comprising answer correct, mistake with At least three kinds of unknown (can not judge whether correctly or incorrectly) etc. as a result, as non-limiting example, the topic type is for example It is simple answer, Essay Writing topic and gap-filling questions etc..
Referring next to Fig. 1, via the aobvious electronic test paper from server 30 outside the first display module 102 and take one's test Student can be via the first input module 101 existing comprising subjective item and objective item to electronic test paper of the first terminal equipment 10 Interior quiz is answered and answer result is uploaded to server 30, and then, the first of server 30 reads and makes comments module 301 can According to the Key for Reference for corresponding to each quiz, the comparison of Lai Jinhang answer result, and then obtains and tentatively read and make comments result.
In the present invention, the answer of quiz can be also stored in paper database 31 with each quiz, in the present invention Specific embodiment in read and make comments module 301 by tentatively read and made comments automatically first and read out, according to each quiz by participating in Quiz answer given by the student of test as a result, first read and make comments module 301 automatically read and make comments the student that takes one's test to Whether the answer result of each quiz out is correct.The process of Automatic Read Overmarginalia according to the present invention described in detail below: it is taking First in business device 30 reads and makes comments complete comparison unit 3011 of the module 301 with complete comparing function and there is character string to compare The character string comparison unit 3012 of function.
Fig. 2 is referred to, wherein Fig. 2 is to schematically depict the according to the present invention first composition for reading and making comments module 301.Such as Shown in Fig. 2, described first, which reads and makes comments module 301, further includes complete comparison unit 3011, character string comparison unit 3012.Below By English test electronic test paper for, quiz can be roughly divided into have appraise according to model answer go over examination papers it is objective Topic, and with subjective item both topic types goed over examination papers of taking the circumstances into consideration appraise according to Key for Reference, wherein have according to reference to answering Case is taken the circumstances into consideration appraise the subjective item goed over examination papers, and the whether correct judgment mode of answer has certain fuzzy space.Citing and Speech, for the topics type such as simply inscribing, Key for Reference is often made of multiple keywords (i.e. so-called to branch or adopt branch), Even if the answer result of the student to take one's test is above not quite similar with Key for Reference in answer statement, if but having contained described It is multiple to branch or to adopt branch, and state upper logical, can be considered that answer is correct, even if answer statement may with answer It is very big to inscribe result gap.
According to a particular embodiment of the invention, complete comparison unit 3011 is used to compare answering for the objective item in each quiz Topic as a result, and generate correspond to objective item accurately read and make comments result, wherein the objective entitled topic type with model answer, Such as the objective entitled single choice, as long as having read the option information that the student that takes one's test is chosen, by the option information Whether answer correct with the scheduled model answer test question that compares that you can get it and gives corresponding score, subsequent institute The score for stating objective item will be sent to fractional statistics module 302.
Further, for the answer of character property objective item, due to the student that takes one's test knowledge point grasp ability and Without fixed format, the answer result for being input to character string comparison unit 3012 exists for the difference of answer level and answer itself The different situation of answering mode.It really may be implemented accurately to automate using character string comparison progress operation and go over examination papers, but count Calculation amount is still excessive, therefore has and advanced optimize to promote the room for improvement of operation efficiency.
As more preferred mode, in electronic test paper, the character string comparison unit 3012 will take one's test student Answer result compared with Key for Reference obtain the subjective item about the test tentatively read and make comments as a result, and described Indicate the error degree with Key for Reference in answer result respectively in different colors.Specifically, test student is first checked for Answer as a result, differentiating sentence paragraph using comma or fullstop to form multiple character strings, then by the multiple character string Transcription comparison is carried out with the character string of Key for Reference to read and make comments, when answer result is clearly present error with Key for Reference, depending on It for answer apparent error, will be shown with red-label, conversely, when not having obvious errors, it is correct to be considered as answer, will be shown with Green Marker Show, to can not judge whether correctly or incorrectly to indicate with yellow.It is described in detail below with reference to preferable example.
Still by taking English letter answer as an example, as shown in figure 3, the character string comparison unit 3012 includes: the first conversion submodule Block 321, for the character string of multiple character strings of the answer result for the student that takes one's test and the Key for Reference of subjective item will to be used as It is converted into word sequence respectively;Second transform subblock 322, for word sequence to be converted into corresponding set of words respectively;First fortune Operator module 323, the min-hash distance between minimum hash and set of words for calculating each set of words;Classification submodule Block 324, for being classified according to min-hash distance to word sequence;And the second operation submodule 325, in same class In carry out shared string operations.It should be pointed out that in the complete comparison unit 3011 and character string comparison unit 3012 Each functional module being implemented in combination with by hardware, software or hardware and software, thereby executing following according to this hair Bright method and steps.
In the present invention, " word sequence " refers to the sequence of word;Correspondingly, " set of words " refers to the set of word.In other words, sequence Column and the constitution element of set are all words.The difference of the two is that the element in sequence can repeat and must have sequentially, and Order of elements is not considered in set and element is without repetition.As an example, for example, test student submit answer result word Symbol string are as follows: " the Chinese nation gets the great rejuvenation ";At the same time, the word of Key for Reference Symbol string is " the great rejuvenation of the Chinese nation ".So, as shown in figure 4, first in step The word sequence A for obtaining the character string of answer result after the participle operation of the first transform subblock 321 in rapid S100 are as follows: { the, Chinese, nation, gets, the, great, rejuvenation }.The character string of Key for Reference is by the first conversion Keyword word sequence B:{ the, great, rejuvenation, of, the obtained after the participle operation of device 321, Chinese, nation }.Certainly as known to those skilled in the art, character string of the invention is right not only just for English word It also can be all applicable in the mixed word etc. of the word or Chinese and English of Chinese.
It in the present invention, then in step s 200 can be by the more of the second transform subblock self-test in 322 future student The word sequence of the Key for Reference of a answer result and the subjective item is converted into corresponding set of words, in conversion process, weight Existing word of appearing again only retains one, such as " the ".To make above example convert after obtain set of words A the, Chinese, nation, gets, great, rejuvenation };Set of words B be then the, great, rejuvenation, Of, Chinese, nation }.
Then, each of the above set of words is calculated using hash algorithm by the first operation submodule 323 in step S300 Minimum hash, the minimum hash can be used for calculating the Jaccard similarity between two set of words.Those skilled in the art Member knows that so-called Jaccard similarity refers to for two set A, B, then Jaccard (A, B)=| A ∩ B |/| A ∪ B |.? On the basis of this, the first operation submodule 323 calculates the set of words for each answer result that test student submits and subjective item refers to The set of words of answer two-by-two between min-hash distance.In the present invention, minimum hash is the numerical value (such as 32 of regular length Position).In this 32, position is identical and the number for the position that numerical value is different is the min-hash distance between two set of words, For middle min-hash apart from smaller, the Jaccard similarity of the two set of words is higher.By the min-hash distance being calculated with Preset first threshold is compared, and min-hash distance is less than word order corresponding to the set of words of the first threshold Column are classified as same class, and in the present invention, as non-limiting example, the first threshold can be the digit of minimum hash 50%.Certainly, those skilled in the art, which can according to need, sets the threshold value, this is understood that obtain.
Then, in step S400, in multiple answer results for submitted the student to take one's test by classification submodule 324 , the word sequence for meeting min-hash distance in the word sequence being converted into and being less than this condition of first threshold remain, it is right In being unsatisfactory for answer of the min-hash distance less than this condition of first threshold, the answer result of student's input can be considered as Completely and the keyword (i.e. point scoring or adopting branch) of Key for Reference is dissimilar or not approximate, therefore character string comparison unit The answer result of the student is directly automatically judged to answer apparent error and is indicated with red-label and be sent to the by 3012 Two display modules 202, and the test question is not given point or the information of score is not sent to fractional statistics module 302.It is right In the word sequence for meeting min-hash distance condition, its word sequence corresponding with Key for Reference is divided into a class.
Finally, in step S500, by the second operation submodule 325 arbitrarily selected from above same class one as The word sequence of student's answer result and the word sequence of Key for Reference carry out the operation of shared character string.Below to belong in same class A student the word sequence of answer result and the word sequence of Key for Reference for carry out shared character string Calculating process.For simplicity by above English word respectively with the schematical mark of letter.
In the first situation, it is assumed that the word order of student's answer result is classified as { A, B, A, D, E, F, G }, the word of Key for Reference Sequence is { A, B, B, D, E, F, G }.By sharing string operations, the word sequence of student's answer result and the word order of Key for Reference The shared character string of column is { A, B, D, E, F, G }, length 6.Since the length of the word sequence of Key for Reference is 7, it is known that shared The lenth ratio of the word sequence of string length and Key for Reference is greater than preset threshold value (such as 80%).In this case, Can be considered as student input answer result it is identical with keyword (i.e. point scoring or the adopting branch) of Key for Reference or Person is completely approximate, and it is obvious correct simultaneously that the answer result of the student is directly automatically judged to answer by character string comparison unit 3012 And the word order of student's answer result is classified as outside { A, B, A, D, E, F, G } progress and shows and is included by shared character string therein Character shown with Green Marker and be sent to the second display module 202, further also the corresponding score of the test question is believed Breath is sent to fractional statistics module 302.
In another case, assume that the word order of student's answer result is classified as { A, B, A, D, E, F, G }, the word order of Key for Reference It is classified as { A, B, B, C, E, F, G }.By sharing string operations, the word sequence of student's answer result and the word sequence of Key for Reference Shared character string be { A, B, E, F, G }, length 5.Since the length of the word sequence of Key for Reference is 7, it is known that shared character The lenth ratio of the word sequence of string length and Key for Reference is less than preset threshold value.In this case, character string comparison unit 3012 it is considered that the answer result of student input and keyword (i.e. point scoring or adopting branch) part phase of Key for Reference Same or part is approximate, i.e., the answer result of the described student belongs to the scope that can not judge whether correctly or incorrectly.In this feelings Under condition, by the word order of student's answer result be classified as { A, B, A, D, E, F, G } carry out it is outer aobvious and by shared character string institute therein The character for including shows with yellow flag and is sent to the second display module 202, and the score of the test question half is believed Breath can be sent to fractional statistics module 302.
From the foregoing, it will be observed that complete comparison unit 3011 and character string comparison unit 3012 in module 301 are read and made comments by first, Can be realized to automatically, rapidly answer to the student that takes one's test to electronic test paper the case where read and make comments, wherein for visitor Sight topic can accurately provide score, and for subjective item, preliminary screening can be provided according to answer situation and is criticized for the first time Read as a result, and the answer result for the student that takes one's test is marked with different colors, thus go over examination papers religion convenient for the later period Teacher's finally reads and makes comments work.Further, according to the preferred embodiment of min-hash algorithm utilized above, with unused min-hash The character string comparison unit of algorithm is compared, and can rapidly be judged between answer result and Key for Reference with the presence or absence of apparent poor It is different, so as to be effectively saved the calculation amount of character string comparison operation and the calculation resources of occupancy.
From the above, in the present invention, finally reading and making comments work includes finally reading and making comments work and to student to subjective item The note of answer and annotation, and retain reading and making comments to show outside trace and checking to the student to take one's test for the teacher that gos over examination papers.Specifically For, as shown in Figure 1, the teacher that gos over examination papers can log on server 30 via second terminal equipment 20, utilize the second input module 201 and second read and make comments module 112 to read and make comments for the first time result and first read and make comments the score that module 301 provides read and made comments again and It obtains and finally reads and makes comments result.In the present invention, finally reading and making comments result may be that first read and make comments in result of modification is judged for example The answer of subjective item whether review again correctly or incorrectly, can also be to be criticized to the answer result and thinking of each subjective item Note, naturally it is also possible to be that amendment first reads and makes comments the score that module 301 provides to generate and finally read and make comments as a result, can also further incite somebody to action It is stored in server 30 with the electronic test paper for finally reading and making comments result and/or answer annotation for the reading of first terminal equipment 10 It takes.Specifically, go over examination papers teacher by the answer result of each subjective item again row check correct or annotate after finally read and made comments As a result, therefore, the recycling first terminal equipment 10 of the student to take one's test logs on server 30, and via the first display mould Block 102 shows annotation given by result (containing the electronic test paper for finally reading and making comments score) and the person of going over examination papers after reading and making comments.
In the present invention, second input module 201 is preferably the handwriting input trace that can recorde the teacher that gos over examination papers Hand-writing recognition unit 201.In the present invention, the hand-writing recognition unit 201 of adaptation is selected according to the type of server 30.Tool For body, when server 30 is desktop computer, the hand-writing recognition unit 201 can utilize wired mode (such as USB connection Line) or wireless mode (such as infrared ray, bluetooth, WiFi, NFC wireless communication technique) with as desktop computer server into Row communication connection, and in the case, the hand-writing recognition unit 201 can be selected from handwriting pad, photography assembly or electronic pen etc. Has the handwriting recognition component of handwriting record and discriminating function.Further, the hand-writing recognition unit 201 preferably has memory Function is for storage handwriting recognition data.At the same time, when second terminal equipment 20 is movable termination, for example, smart phone Or when tablet computer, the hand-writing recognition unit 201 can use the capacitive touch sensors of the movable termination with And matched handwriting recognition engine, i.e., religion of going over examination papers directly is recorded in the sliding of the mobile terminal by the finger of the teacher that gos over examination papers The handwriting input trace of teacher.Then, divide the handwriting input trace and by handwriting recognition engine identification character to be transmitted to screen Curtain carries out outer aobvious.
In a preferred embodiment, by way of the shared string operations in character string comparison unit 3012 Through the Key for Reference of the answer result for the student that takes one's test and each subjective item has been carried out preliminary comparison.For via character Correct answer of the comparison unit 3012 labeled as the complete wrong of red and labeled as green go here and there as a result, going over examination papers teacher at this time only It needs to operate hand-writing recognition unit 201 to add their confirmation to the result of going over examination papers of character string comparison unit 3012.At this time due to religion of going over examination papers Teacher only needs substantially to browse can be confirmed by the requirement whether answer of the part of Green Marker has met Key for Reference, To greatly reduce the labor intensity for the teacher that gos over examination papers.Further, it is answered for what label in the second display module 202 was Topic as a result, the teacher that gos over examination papers can annotate answer result according to answer situation, such as in simple topic in label with it is described Related source of the relevant knowledge point of problem to the knowledge point in textbook, above information can be led to by the teacher that gos over examination papers It crosses hand-writing recognition unit 201 to be input in server 30, and is further used as additional information storage into electronic test paper.
The answer for being for the label that can not judge whether correctly or incorrectly is as a result, the teacher that gos over examination papers needs emphasis to be subject to It reads and makes comments.When reading and making comments, exam paper marking system of the invention can assign the teacher that the gos over examination papers permission for reading and making comments the subjective item again.Tool For body, exam paper marking system of the invention is that per pass subjective item is equipped with score upper bound belonging to it, for example, per pass letter answer Score upper bound be 10 points.Label correctly or incorrectly, which is, to be judged whether for those character string comparison units 3012 Answer as a result, being taken the circumstances into consideration in the score section of 0-10 point by going over examination papers teacher according to Key for Reference to be given a mark and/or write batch Note.When exam paper marking system discovery is greater than 10 via the fractional value that the teacher that gos over examination papers that hand-writing recognition unit 201 inputs provides, institute It states exam paper marking system and forbids the operation from second input equipment 20 and sending " can not operate " or " beyond score The warning of the upper limit ", to prompt to go over examination papers, teacher does not meet pre-defined rule to score value for the simple answer.Via to per pass subjectivity Topic is equipped with score upper bound, so that the teacher that gos over examination papers will not be goed over examination papers work and mistakenly assert the class due to heavy The upper limit score value of subjective item, thus maximum likelihood avoid the teacher that gos over examination papers subjective a possibility that making a mistake.
It is further preferred that in order to improve the speed and that retains the teacher that gos over examination papers go over examination papers trace for example of going over examination papers for the teacher that gos over examination papers Point comment or annotation need to effectively improve in the writing input efficiency and stringent identification for example, of hand-writing recognition unit 201 The input method of the character of text, therefore input method hand-written it is preferable to use reduplicated word in the present invention, this input method are to allow For user after writing a character, the character late of handwriting input can fold the handwriting input for writing on the top of first character Method.The hand-written input method of reduplicated word according to the present invention, go over examination papers teacher can in the useful space of handwriting input any position It sets and carries out continuous handwriting input, to more be bonded the habit of going over examination papers for the teacher that gos over examination papers.
Specifically, hand-writing recognition unit 201 of the invention is configured to allow go over examination papers teacher uninterruptedly carry out it is any number of Character continuous hand-written simultaneously carries out corresponding identifying processing, to effectively promote the efficiency for writing input.It below will be by readding It is that example describes this hair come the handwriting input trace for recording the teacher that gos over examination papers that the finger of teacher, which is rolled up, in the sliding of the mobile terminal The hand-written input method of bright reduplicated word.Certainly, as known to those skilled in the art, the hand-written input method of above-mentioned reduplicated word is similarly Suitable for carrying out handwriting input in a manner of writing pencil and this mode should also be as being considered as claimed content.
The hand-written input method of the reduplicated word of multiword symbol continuous writing according to the present invention, is detected by hand-writing recognition unit 201 The different strokes of first character and second character out, when the teacher that gos over examination papers starts to fold and writes second character, preferably automatically The display color for having write the first character finished is dimmed, to facilitate the teacher that gos over examination papers from 202 supernatant of the second display module The trace of the current character of oneself institute's handwriting input is seen clearly.It can not stop in this way and continuously arbitrarily write multiple words Symbol.It is when handwriting input third character, the viewing area of the first character inputted before from the second display module 202 is automatic It disappears, and second character inputted before being allowed to is automatically dimmed, thus realizes final continuous formality input, it below will be specific Introduce each step of the method:
Go over examination papers first teacher finger contact in the present invention be, for example, touch screen the second display module 202, start it is defeated Enter stroke, at this moment hand-writing recognition unit 201 can be detected finger contact this event of touch screen, start recording go over examination papers teacher input Character stroke person's handwriting.Next, the hand-writing recognition unit 201 records religion of going over examination papers when finger moves on the touchscreen The handwriting input trace of teacher, and transfer handwriting area on the touchscreen and show corresponding stroke track, this embodiment it In, the stroke color of current hand-written character is indicated with red;(as known to those skilled in the art, it is also possible to which other color expression is worked as The color of preceding handwritten stroke, red are intended merely to be more in line with the habit of going over examination papers for the teacher that gos over examination papers, and replace the color of handwritten stroke simultaneously The substantive content of the content of present invention is not influenced);The teacher that gos over examination papers keeps handwriting input until its finger leaves touch screen, current stroke End of input.
Then, hand-writing recognition unit 201 judges whether the stroke of current handwriting input belongs to the stroke inputted before In the same character.Specific decision logic and method are as follows: either Chinese Hanzi character or English word character, stroke Input is all sequence from left to right, such as is usually the combinde rqdical character of tiled configuration in chinese character, wherein left side is needs The radical first inputted.Based on this discovery, the stroke for judging current handwriting input of the application and the stroke inputted before Whether first criterion of same character is belonged to are as follows: judges whether current stroke is newly to write pen on the right side of preceding unicursal It draws, if it is determined that being that stroke is write in the new of right side, it may be considered that stroke currently entered and the stroke inputted before are to belong to In the same character.If it is determined that being not located at the right side of preceding unicursal, then in view of certain Chinese characters are, for example, up-down structure Combinde rqdical character, such as " basin " word are made of " being divided to " and " ware word bottom " two radicals, then carry out the judgement of Second Criterion condition: judgement is worked as Whether preceding stroke, which is folded on trace with the stroke of handwriting input before, is write, if not folded on trace write, can be recognized It with the stroke before inputted is to belong to the same character for stroke currently entered.If it is determined that current stroke is therewith The stroke of preceding handwriting input on trace fold writes, then carry out the judgement of third criterion: judge current stroke with it is hand-written The folded of the stroke of input writes degree, writes degree less than 60% if folded, it may be considered that stroke currently entered with before The stroke of input is to belong to the same character., whereas if folded degree of writing is greater than 60%, it may be considered that stroke currently entered It is to be not belonging to the same character with the stroke inputted before.
Then, judge that the stroke that stroke currently entered has been inputted with before is to belong to together in hand-writing recognition unit 201 In the case where one character, further determine whether that there are also new strokes to input, if there is also have new stroke input, institute The handwriting input trace that hand-writing recognition unit 201 persistently records the teacher that gos over examination papers is stated, and handwriting area always on the touchscreen is aobvious Show corresponding stroke track, until to show that stroke currently entered no longer belongs to the stroke inputted before same for judgement The handwritten traces of current record write are carried out Text region at this time, and export recognition result on the touchscreen by character, so Recognition result is subjected to dimmed processing afterwards.
Meanwhile if judging that the stroke that stroke currently entered has been inputted with before does not belong in hand-writing recognition unit 201 In the case where the same character, then judging, which whether there is on touch screen, has write and had carried out identifying processing and become The hand-written character of dark treatment, if having had existed such hand-written character, by the word for having carried out identifying processing Symbol is eliminated.Continue by other than current stroke all strokes constitute a handwriting characters, carry out Text region and Recognition result is exported on touch screen, and recognition result is then subjected to dimmed processing.
It can be seen that the input method hand-written via reduplicated word of the invention, so that the teacher that gos over examination papers carries out book on the touchscreen When writing, continuous reduplicated word handwriting input is carried out according to the habit that the teacher that gos over examination papers writes naturally, it can be continuous hand-written defeated incessantly Enter multiple texts, the handwritten traces that hand-writing recognition unit 201 automatically inputs the teacher that gos over examination papers carry out identification output, thus effectively The efficiency for improving handwriting input.
On this basis, read and make comments the operation at module 303 second via teacher is goed over examination papers, can to tentatively go over examination papers result and The score being stored in fractional statistics module 302 is read and made comments again and is obtained and finally read and made comments result and go over examination papers given by teacher Annotation.The recycling first terminal equipment 10 of the student to take one's test as a result, logs on server 30 and is read and made comments with reading second Information above in module 303, and finally reading and making comments after reading and making comments is shown to the student to take one's test via the first display module 102 As a result (contain the electronic test paper for finally reading and making comments score) and the teacher that gos over examination papers given by annotation.
To sum up, exam paper marking system according to the present invention can first be answered with standard by complete comparison unit 3011 The answer result of the objective item of case is strictly compared, and via character string comparison unit 3012 to by multiple crucial phrases At the answer result that provides of student of subjective item of Key for Reference compare, thus tentatively goed over examination papers result and its point Number, at the same the answer result of subjective item is indicated according to different situations with different colors and show answer result with Shared character string between Key for Reference.Then, going over examination papers teacher can be to tentatively going over examination papers result and its score is read and made comments again, example Such as is checked, is confirmed, modified or marked in detail to Key for Reference or the subjective item compare, thus It obtains finally to read and make comments result and be back to outside first terminal equipment and show to the student to take one's test.Via automatically to objective item progress It compares completely, and the mode that shared character string is compared and indicated is carried out to subjective item, the labour that can reduce the teacher that gos over examination papers is strong Degree also takes into account the accuracy goed over examination papers simultaneously.
The technical scheme of the present invention has been explained in detail above with reference to the attached drawings, although preferred implementation of the invention has been described Example, once a person skilled in the art knows basic creative concepts, then other change can be made to these embodiments More and modify.So it includes preferred embodiment and all changes for falling into the scope of the invention that the following claims are intended to be interpreted as More and modify.
Better embodiment of the invention is described above, it is intended to so that spirit of the invention is more clear and convenient for managing Solution, is not meant to limit the present invention, all within the spirits and principles of the present invention, modification, replacement, the improvement made should all Within the protection scope that appended claims of the invention is summarized.

Claims (10)

1. a kind of exam paper marking system, for reading and making comments student to the electronics comprising at least one test question from paper database Paper answer as a result, the exam paper marking system includes:
First terminal equipment, comprising:
First display module, for showing the test question of the electronic test paper to student;
First input module, for the answer result for student's input to the test question;
Second terminal equipment, comprising:
Second display module shows the electronic test paper with the answer result for being denoted as different colours to outside the teacher that gos over examination papers;
Second input module reads and makes comments the electronic test paper for the teacher that gos over examination papers;
Server is communicatively connected to the first terminal equipment, second terminal equipment and paper database, the service respectively Device includes
First reads and makes comments module, by the answer of the test question of paper database purchase and the answer result carry out automatic comparison with It generates and tentatively reads and makes comments result;
Fractional statistics module tentatively reads and makes comments result for receiving and counting described;
Second reads and makes comments module, generates for the input according to the teacher that gos over examination papers and finally reads and makes comments result.
2. exam paper marking system according to claim 1, wherein
At least one described test question preferably includes subjective item and objective item, and described first reads and makes comments module including for criticizing Read the complete comparing unit of the objective item and the character string comparison unit for reading and making comments the subjective item.
3. exam paper marking system according to claim 1, wherein
The character string comparison unit, it is preferable that include:
First transform subblock, for dividing multiple character strings of the answer result and the character string of the Key for Reference of subjective item Word sequence is not converted into it;
Second transform subblock, for the word sequence to be converted into corresponding set of words respectively;
First operation submodule, the min-hash distance between minimum hash and set of words for calculating each set of words;
Classification submodule, for being classified according to min-hash distance to word sequence;And
Second operation submodule, for carrying out shared string operations in same class.
4. exam paper marking system according to claim 3, wherein it is further,
The set of words two of the set of words of each answer result and the Key for Reference of subjective item is calculated by the first operation submodule The min-hash distance of set of words with Key for Reference is less than first threshold by classification submodule by the min-hash distance between two Set of words corresponding to word sequence be divided into same class, and min-hash distance is greater than corresponding to the set of words of first threshold The character string of answer result is denoted as the first color;Then,
Following shared character string fortune is carried out in the same class that min-hash distance is less than first threshold by the second operation submodule It calculates: using the word sequence of Key for Reference as the first word sequence in such, calculating separately in first word sequence and such The shared character string of other word sequences, wherein the ratio of the string length of the length and Key for Reference of obtained shared character string The character string that value is greater than answer result corresponding to the word sequence of second threshold is denoted as the second color, and obtained shared word The ratio for according with the length of string and the string length of Key for Reference is less than answer result corresponding to the word sequence of second threshold Character string is denoted as third color.
5. exam paper marking system according to claim 4, wherein
The first threshold is the 50% of the digit of the minimum hash of set of words, the second threshold, preferably 80%.
6. exam paper marking system according to claim 1, wherein
Second, which reads and makes comments module, is configured to receive the input from the second terminal equipment in this way: corresponding to being denoted as first The answer of the character string of color and the second color is as a result, described second reads and makes comments module and allow to go over examination papers teacher via the second terminal equipment The result of tentatively reading and making comments is added their confirmation and/or annotated to be formed via the second final batch for reading and making comments that module exports outward Read result;Corresponding to the character string for being denoted as third color answer as a result, described second read and make comments module allow to go over examination papers teacher warp It is tentatively read and made comments result to described by the second terminal equipment and is modified and/or annotated to be formed and read and make comments mould via second What block exported outward finally reads and makes comments result.
7. exam paper marking system according to claim 6, wherein
The exam paper marking system is that per pass subjective item is equipped with affiliated score upper bound, when the teacher that gos over examination papers is whole via described second End equipment to it is described tentatively read and make comments fractional value that result modify beyond the score upper bound when, the exam paper marking system taboo Operation only from the second terminal equipment is simultaneously alarmed.
8. exam paper marking system according to claim 6, wherein
The first terminal equipment receives to read and make comments finally reading and making comments as a result, and via first display module for module from second It shows outside to student.
9. exam paper marking system according to claim 1, wherein
Second input module is the hand-writing recognition unit for recording the handwriting input trace for the teacher that gos over examination papers, the second display mould Block is the touch screen with hand-write input function.
10. exam paper marking system according to claim 1, wherein
The hand-writing recognition unit is configured to execute the hand-written input method of reduplicated word, and the hand-written input method of the reduplicated word includes Following steps:
Firstly, the hand-writing recognition unit detects that finger contacts this event of touch screen, start recording gos over examination papers what teacher inputted The stroke person's handwriting of character, and the handwriting area transferred on the touchscreen shows corresponding stroke track;Until its finger leaves touching Touch screen, the end of input of current stroke;
Then, the hand-writing recognition unit judged according to following judgment basis the stroke of current handwriting input with it is defeated before Whether the stroke entered belongs to the same character: first determine whether current stroke is newly to write stroke on the right side of preceding unicursal, It is write if it is not, then further judging whether current stroke is folded on trace with the stroke of handwriting input before, if so, into One step judges that current stroke and the folded of stroke of handwriting input write whether degree is greater than predetermined threshold, if so, thinking to work as The stroke of preceding handwriting input and the stroke inputted before are not belonging to the same character;
If the hand-writing recognition unit judges that stroke currently entered and the stroke inputted before are to belong to the same character, Then the hand-writing recognition unit persistently records the handwriting input trace for the teacher that gos over examination papers, until judgement obtains stroke currently entered not Belong to the same character with the stroke inputted before again, the handwritten traces of current record write are subjected to text at this time Identification, and recognition result is exported on the touchscreen, recognition result is then subjected to dimmed processing;
If the hand-writing recognition unit judges that stroke currently entered and the stroke inputted before are not belonging to the same character, Then judging, which whether there is on touch screen, has write and had carried out identifying processing and carried out the hand-written character of dimmed processing, if There are such hand-written characters, then eliminate the character for having carried out identifying processing;Continuing will be other than current stroke All strokes constitute a handwriting characters, carry out Text region and exporting recognition result on the touchscreen, then will knowledge Other result carries out dimmed processing.
CN201710981868.8A 2017-10-20 2017-10-20 A kind of exam paper marking system Pending CN109697905A (en)

Priority Applications (2)

Application Number Priority Date Filing Date Title
CN201710981868.8A CN109697905A (en) 2017-10-20 2017-10-20 A kind of exam paper marking system
PCT/CN2017/111811 WO2019075820A1 (en) 2017-10-20 2017-11-20 Test paper reviewing system

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN201710981868.8A CN109697905A (en) 2017-10-20 2017-10-20 A kind of exam paper marking system

Publications (1)

Publication Number Publication Date
CN109697905A true CN109697905A (en) 2019-04-30

Family

ID=66173065

Family Applications (1)

Application Number Title Priority Date Filing Date
CN201710981868.8A Pending CN109697905A (en) 2017-10-20 2017-10-20 A kind of exam paper marking system

Country Status (2)

Country Link
CN (1) CN109697905A (en)
WO (1) WO2019075820A1 (en)

Cited By (11)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN110910043A (en) * 2019-12-04 2020-03-24 广州云蝶科技有限公司 Analysis method of test paper information
CN111367451A (en) * 2020-02-29 2020-07-03 南京红松信息技术有限公司 Method and system for network scoring by handwriting
CN111597931A (en) * 2020-04-30 2020-08-28 杭州学同科技有限公司 Machine marking method based on edge calculation
CN112597990A (en) * 2020-12-22 2021-04-02 安徽鸿程光电有限公司 Judging and reading method and system of handwriting formula, terminal device and storage medium
CN113420983A (en) * 2021-06-23 2021-09-21 科大讯飞股份有限公司 Writing evaluation method, device, equipment and storage medium
CN113486786A (en) * 2021-07-05 2021-10-08 江苏鼎峰信息技术有限公司 Automatic operation correcting system
CN113643582A (en) * 2021-10-14 2021-11-12 南京极域信息科技有限公司 Multi-source wireless interactive feedback system
CN114550181A (en) * 2022-02-10 2022-05-27 珠海读书郎软件科技有限公司 Method, device and medium for identifying question
CN116343242A (en) * 2023-05-30 2023-06-27 山东一品文化传媒有限公司 Real-time examination and reading method and system based on image data
CN116416629A (en) * 2023-06-12 2023-07-11 北京量子伟业信息技术股份有限公司 Electronic file generation method, device, equipment and medium
CN116610774A (en) * 2023-07-20 2023-08-18 河北鑫考科技股份有限公司 High-efficiency intelligent online paper reading auxiliary method and system

Families Citing this family (27)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN110096708B (en) * 2019-04-30 2023-12-01 科大讯飞股份有限公司 Calibration set determining method and device
CN110245860B (en) * 2019-06-13 2022-08-23 桂林电子科技大学 Automatic scoring method based on virtual experiment platform
CN110321837B (en) * 2019-07-02 2023-05-30 南京怀宇科技有限公司 Test question score identification method, device, terminal and storage medium
CN110633767A (en) * 2019-08-06 2019-12-31 安徽学云教育科技有限公司 Online examination paper reading abnormity screening method and system
CN110795997B (en) * 2019-09-19 2023-07-28 平安科技(深圳)有限公司 Teaching method and device based on long-short-term memory and computer equipment
CN110991277B (en) * 2019-11-20 2023-09-22 湖南检信智能科技有限公司 Multi-dimensional multi-task learning evaluation system based on deep learning
CN111179670B (en) * 2019-12-27 2022-09-16 广东德诚科教有限公司 Method and device for quantifying physical and electrical experiment result, terminal and storage medium
CN110968277B (en) * 2019-12-30 2023-04-11 福建天晴数码有限公司 Answer sheet generation method
CN111274262B (en) * 2020-01-06 2022-08-26 成都信息工程大学 SQLServer automatic question judging system
CN111638807A (en) * 2020-04-29 2020-09-08 上海元趣信息技术有限公司 Learning auxiliary system based on intelligent handwriting pen
CN111666864B (en) * 2020-06-01 2023-03-24 合肥工业大学 Automatic paper marking method based on computer vision
CN111738885A (en) * 2020-06-28 2020-10-02 闵亨锋 Student class attending quality evaluation method and system based on wearable device
CN111898972A (en) * 2020-07-08 2020-11-06 东莞理工学院城市学院 Overseas personnel introduction and evaluation system with evaluation system for human resources
CN112765646B (en) * 2020-12-22 2023-05-30 南京新知艺测科技有限公司 Method and system for recording painting and calligraphy and anonymously reading paper in artistic evaluation
CN112749692A (en) * 2020-12-30 2021-05-04 广州宏途教育网络科技有限公司 Intelligent reading and amending system
CN113191344A (en) * 2021-04-22 2021-07-30 读书郎教育科技有限公司 Automatic scoring method for intelligent desk lamp and intelligent desk lamp
CN113128469A (en) * 2021-05-13 2021-07-16 北京一起教育科技有限责任公司 Paper teaching text processing method and device and electronic equipment
CN113298491A (en) * 2021-05-18 2021-08-24 郑州千慧传媒信息技术有限公司 Logic method for art test paper evaluation
CN113343650A (en) * 2021-06-23 2021-09-03 武汉悦学帮网络技术有限公司 Batch reading method and device, electronic equipment and storage medium
CN113408520B (en) * 2021-07-16 2023-12-12 北京南昊科技股份有限公司 Marking method, marking device and storage medium
CN113705438A (en) * 2021-08-27 2021-11-26 河北风口飞网络科技有限公司 Examination and examination paper marking method and system based on multiple data acquisition modes
CN113723323A (en) * 2021-09-02 2021-11-30 北京云蝶智学科技有限公司 Test paper information acquisition method and device
CN113723543B (en) * 2021-09-02 2024-01-09 成都东方闻道科技发展有限公司 Automatic cutting question display method and device, computer equipment and storage medium
CN114368534B (en) * 2022-01-13 2023-06-30 陕西师范大学 Intelligent test paper distributing and collecting machine
CN114407536B (en) * 2022-01-16 2023-03-21 北京云思智学科技有限公司 Control method of automatic test question correcting and mark retaining device
CN114418806B (en) * 2022-03-25 2022-07-12 北京一起航帆科技有限公司 Data processing method and device and electronic equipment
CN116308934B (en) * 2023-05-15 2024-01-16 深圳市采集科技有限公司 Auxiliary equipment for correcting operation and method for correcting operation by using same

Citations (9)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN1801244A (en) * 2004-12-01 2006-07-12 株式会社东芝 System, server, and program for providing examination
US20090006510A1 (en) * 2007-06-29 2009-01-01 Raza Microelectronics, Inc. System and method for deflate processing within a compression engine
CN102156609A (en) * 2010-12-10 2011-08-17 上海合合信息科技发展有限公司 Overlap handwriting input method
CN102508598A (en) * 2011-10-09 2012-06-20 北京捷通华声语音技术有限公司 Method and device for gradually blanking character strokes
CN102799767A (en) * 2012-06-27 2012-11-28 苏州奇可思信息科技有限公司 Marking software system
CN103236196A (en) * 2013-04-08 2013-08-07 明博教育科技有限公司 Electronic assignment system and method allowing for automatic distributing, receiving, marking and counting
CN104794948A (en) * 2015-04-20 2015-07-22 西安青柠电子信息技术有限公司 Automatic scoring system and method for applying same
US20150332600A1 (en) * 2014-03-31 2015-11-19 Varun Aggarwal Method and system for building and scoring situational judgment tests
CN106610965A (en) * 2015-10-21 2017-05-03 北京瀚思安信科技有限公司 Text string common sub sequence determining method and equipment

Family Cites Families (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20070160971A1 (en) * 2006-01-12 2007-07-12 Caldera Paul F Method for Automated Examination Testing and Scoring
CN106781784A (en) * 2017-01-04 2017-05-31 王骁乾 A kind of intelligence correction system

Patent Citations (9)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN1801244A (en) * 2004-12-01 2006-07-12 株式会社东芝 System, server, and program for providing examination
US20090006510A1 (en) * 2007-06-29 2009-01-01 Raza Microelectronics, Inc. System and method for deflate processing within a compression engine
CN102156609A (en) * 2010-12-10 2011-08-17 上海合合信息科技发展有限公司 Overlap handwriting input method
CN102508598A (en) * 2011-10-09 2012-06-20 北京捷通华声语音技术有限公司 Method and device for gradually blanking character strokes
CN102799767A (en) * 2012-06-27 2012-11-28 苏州奇可思信息科技有限公司 Marking software system
CN103236196A (en) * 2013-04-08 2013-08-07 明博教育科技有限公司 Electronic assignment system and method allowing for automatic distributing, receiving, marking and counting
US20150332600A1 (en) * 2014-03-31 2015-11-19 Varun Aggarwal Method and system for building and scoring situational judgment tests
CN104794948A (en) * 2015-04-20 2015-07-22 西安青柠电子信息技术有限公司 Automatic scoring system and method for applying same
CN106610965A (en) * 2015-10-21 2017-05-03 北京瀚思安信科技有限公司 Text string common sub sequence determining method and equipment

Non-Patent Citations (1)

* Cited by examiner, † Cited by third party
Title
KINGYOSHI: "经济史论文查重率", 《百度文库》 *

Cited By (17)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN110910043A (en) * 2019-12-04 2020-03-24 广州云蝶科技有限公司 Analysis method of test paper information
CN111367451A (en) * 2020-02-29 2020-07-03 南京红松信息技术有限公司 Method and system for network scoring by handwriting
CN111597931A (en) * 2020-04-30 2020-08-28 杭州学同科技有限公司 Machine marking method based on edge calculation
CN112597990A (en) * 2020-12-22 2021-04-02 安徽鸿程光电有限公司 Judging and reading method and system of handwriting formula, terminal device and storage medium
CN113420983A (en) * 2021-06-23 2021-09-21 科大讯飞股份有限公司 Writing evaluation method, device, equipment and storage medium
CN113420983B (en) * 2021-06-23 2024-04-12 科大讯飞股份有限公司 Writing evaluation method, device, equipment and storage medium
CN113486786B (en) * 2021-07-05 2024-04-09 江苏鼎峰信息技术有限公司 Automatic operation modifying system
CN113486786A (en) * 2021-07-05 2021-10-08 江苏鼎峰信息技术有限公司 Automatic operation correcting system
CN113643582A (en) * 2021-10-14 2021-11-12 南京极域信息科技有限公司 Multi-source wireless interactive feedback system
CN114550181B (en) * 2022-02-10 2023-01-10 珠海读书郎软件科技有限公司 Method, device and medium for identifying question
CN114550181A (en) * 2022-02-10 2022-05-27 珠海读书郎软件科技有限公司 Method, device and medium for identifying question
CN116343242B (en) * 2023-05-30 2023-08-11 山东一品文化传媒有限公司 Real-time examination and reading method and system based on image data
CN116343242A (en) * 2023-05-30 2023-06-27 山东一品文化传媒有限公司 Real-time examination and reading method and system based on image data
CN116416629A (en) * 2023-06-12 2023-07-11 北京量子伟业信息技术股份有限公司 Electronic file generation method, device, equipment and medium
CN116416629B (en) * 2023-06-12 2023-08-29 北京量子伟业信息技术股份有限公司 Electronic file generation method, device, equipment and medium
CN116610774A (en) * 2023-07-20 2023-08-18 河北鑫考科技股份有限公司 High-efficiency intelligent online paper reading auxiliary method and system
CN116610774B (en) * 2023-07-20 2023-09-26 河北鑫考科技股份有限公司 High-efficiency intelligent online paper reading auxiliary method and system

Also Published As

Publication number Publication date
WO2019075820A1 (en) 2019-04-25

Similar Documents

Publication Publication Date Title
CN109697905A (en) A kind of exam paper marking system
CN109710590B (en) Error problem book generation method and device
KR101671693B1 (en) Method for Studying User Incorrect Concentration
CN111144191B (en) Font identification method, font identification device, electronic equipment and storage medium
CN106781784A (en) A kind of intelligence correction system
CN105427696A (en) Method for distinguishing answer to target question
CN107016022A (en) Teaching difficulty big data analysis system
CN101866415A (en) Answer sheet recognition device and method of computer marking system
CN104881480B (en) Annotation method and device based on database
CN106898173A (en) The online review system in real time of students' work
WO2014141414A1 (en) Learning process visualization system and visualization processing device, and visualization method
CN110059978B (en) Teacher evaluation system based on cloud computing auxiliary teaching evaluation
CN106373447A (en) Intelligent paper marking system and method
CN105787834A (en) Intelligent student management system
CN110059636B (en) Method and system for checking and correcting learning homework of students
CN112116840B (en) Job correction method and system based on image recognition and intelligent terminal
Scrimgeour Issues and approaches to literacy development in Chinese second language classrooms
CN110826872A (en) Method, system and device for generating examination result analysis report
JP2016085284A (en) Program, apparatus and method for estimating evaluation level with respect to learning item on the basis of person's remark
Boland et al. Creating an annotated corpus for sentiment analysis of german product reviews
Spoon et al. Can we (and should we) use AI to detect Dyslexia in children’s handwriting?
KR101996247B1 (en) Method and apparatus of diagnostic test
Kobayashi et al. Basic research on a handwritten note image recognition system that combines two OCRs
CN115544226B (en) Similarity recognition method based on multi-mode emotion analysis
CN106339965A (en) Learning situation analysis method

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
RJ01 Rejection of invention patent application after publication

Application publication date: 20190430

RJ01 Rejection of invention patent application after publication