CN108711125A - A kind of blocks of knowledge centrad and difficulty quantization method - Google Patents
A kind of blocks of knowledge centrad and difficulty quantization method Download PDFInfo
- Publication number
- CN108711125A CN108711125A CN201810495007.3A CN201810495007A CN108711125A CN 108711125 A CN108711125 A CN 108711125A CN 201810495007 A CN201810495007 A CN 201810495007A CN 108711125 A CN108711125 A CN 108711125A
- Authority
- CN
- China
- Prior art keywords
- knowledge
- blocks
- difficulty
- centrad
- transition state
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION 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
- G06Q50/00—Systems or methods specially adapted for specific business sectors, e.g. utilities or tourism
- G06Q50/10—Services
- G06Q50/20—Education
- G06Q50/205—Education administration or guidance
Abstract
The invention discloses a kind of blocks of knowledge centrad and difficulty quantization methods, carry out the local importance of calculation knowledge unit with local centrality methods in terms of blocks of knowledge centrad, using local importance come the transfer of learning probability matrix P between calculation knowledge unit, in conjunction with blocks of knowledge and the markovian characteristic of absorbing state, blocks of knowledge is divided into transition state and absorbing state, calculate the probability matrix M that transition state blocks of knowledge i is transferred to the mean transferred degree matrix Q of transition state j before being absorbed by the blocks of knowledge of absorbing state and transition state is finally absorbed by absorbing state, obtain the calculation formula of centrad；The level horizontal of blocks of knowledge is first measured in terms of blocks of knowledge difficulty measurement, measures specific difficulty once again in same level level.The present invention has adequacy, necessity and high efficiency to the measurement of blocks of knowledge centrad and difficulty, can be widely applied to the practical application request that online education domain knowledge point importance and difficulty are distinguished.
Description
Technical field
The present invention relates to the calculating of complex network centrad, blocks of knowledge centrad and difficulty measurement, e-learning[Electronics
(change) learns ]Individualized learning bootstrap technique, more particularly to a kind of blocks of knowledge centrad and difficulty quantization method.
Background technology
With the requirement of fast development and the educational reform of network technology, network has become the important promotion of IT application in education sector
Power, network information resource rapid growth.Globalization, knowledge explosion, instant messaging, electronization are uniting and are overturning our work
And mode of learning.Currently, oneself unified logical construction is found in our Web education not yet, largely or utilize
Media are in clone, duplication school education.This gram for ignoring educational object community culture feature and the variation of Social Culture transitional period
Grand and duplication makes the surge of educational resource amount, and " the cognition overload " and " getting lost " that information overload is brought is than the letter under absence of information
Breath, which is hungered and thirst, to be less useful for learning.Research overcomes the core technology of " cognition overload " and " getting lost " in the educational resource that magnanimity increases
Problem is how to distinguish the centrad and difficulty of educational resource blocks of knowledge, to help user to have target to go to learn.And
The centrad and difficulty of blocks of knowledge play an important role in terms of the learning path for generating precise and high efficiency.Currently, centrad is measured
Research mainly in complex network is relatively more, and the measurement of the centrad of blocks of knowledge is fewer.Blocks of knowledge difficulty point
To study less direction in Content Difficulty and statistics difficulty, and current difficulty of knowledge points confirmation method.
Applicant retrieves following as follows with the relevant document of blocks of knowledge centrad difficulty quantization：
1. Zheng Qing is magnificent, yellow perfume monarch etc., a kind of generation method of navigation learning path on knowledge map, patent, and 2011
2.Duanbing Chena,Linyuan Lu,Identifying influential nodes in complex
networks,Physica A:Statistical Mechanics and its Applications,2011
3. the Sun Xue winters, a kind of teaching resource modeling method and system for supporting individualized teaching process to automatically generate, patent,
2017
4. Zheng Qing is magnificent, Dong Bo etc., the document subject matter partitioning based on domain knowledge map community structure, patent, and 2013
5. Wei all, Wang Chenchen etc. of pen, a method of learning dependence between extracting blocks of knowledge automatically from courseware text,
Patent, 2016.
The centrad that document 1 proposes calculates and difficulty measure.In terms of centrad, the calculating of local importance is
Simple to have used out-degree than in-degree, the key of centrad is exactly to have height link property, the out-degree of simple utilization blocks of knowledge
Ignore the link for other blocks of knowledge that thus blocks of knowledge is drawn than in-degree, it is clear that do so the calculating in centrad
There are major hidden dangers in terms of accuracy.In terms of difficulty, the measurement of difficulty level has only been carried out and to knowing under same difficulty level
The difficulty for knowing unit does not make measurement but, if there are many blocks of knowledge under same level, and difficulty gap is also big,
If being easy to prevent learner from according to their own situation preferably learning if not measuring.
Document 2 proposes a kind of measure of the local centrality of non-directed graph, and SIR assessment models is used in combination to go to test
Demonstrate,prove the validity of this method.The Knowledge Map of patent meaning of the present invention is a directed acyclic graph, the local that this article proposes
The measure of centrality can be not applied directly in directed acyclic graph, but important to the part of our blocks of knowledge
The measurement of degree has reference.
Document 3 is mainly unified teaching process description and resource description, supports quickly and accurately according to learner
It practises target and knowledge background carries out the selection of teaching resource, while automatically generating the teaching process optimized accordingly.Although can
The selection in resource is quickly and accurately carried out according to the learning objective of learner and background, but does not elaborate knowledge
The centrad of unit and the computational methods of difficulty.
Document 4 is to propose centrad and the method that blocks of knowledge document frequency is combined, and mainly uses SVM machine learning methods
Calculation knowledge unit center degree, and have ignored Content Difficulty.If by this method for blocks of knowledge difficulty measurement if that
There are the influence tests of many enchancement factors as a result, accurately can not really reflect the difficulty of blocks of knowledge.
Document 5 proposes a kind of method learning dependence between extracting blocks of knowledge automatically from courseware text.It gives
A set of universal model, by calculating criticality of the term to blocks of knowledge, to build the dependence between blocks of knowledge.It should
Method does not use Markov Chain method described in this patent to be designed, and is not related to the quantization of blocks of knowledge difficulty
Method.
Invention content
The invention discloses a kind of blocks of knowledge centrad and difficulty quantization methods, which is characterized in that in blocks of knowledge
In terms of heart degree：(1) carry out the local importance of calculation knowledge unit with local centrality methods：(2) part weight is utilized
Spend the transfer of learning probability matrix P between carrying out calculation knowledge unit；(3) blocks of knowledge and the markovian spy of absorbing state are combined
Property, blocks of knowledge is divided into transition state and absorbing state, transition state blocks of knowledge i is calculated and it is being absorbed by the blocks of knowledge of absorbing state
Before be transferred to the mean transferred degree matrix Q of transition state j and probability matrix M that transition state is finally absorbed by absorbing state：(4) it obtains
The calculation formula of centrad.In terms of blocks of knowledge difficulty measurement；(1) level horizontal of blocks of knowledge is first measured；(2) identical
Level horizontal measures specific difficulty once again.This method has adequacy, necessity to the measurement of blocks of knowledge centrad and difficulty
And high efficiency.
To achieve these objectives, the present invention, which adopts the following technical scheme that, is achieved：
A kind of blocks of knowledge centrad and difficulty measure, which is characterized in that comprise the steps of：
1, in terms of centrad：
(1) carry out the local importance of calculation knowledge unit with local centrality methods.
Knowledge Map is a directed acyclic graph (V, E), is denoted as KM；V=V1∪V2It is that ambit described in KM includes
Blocks of knowledge set, E be in V between blocks of knowledge learn dependence set；Here blocks of knowledge refers to having
The basic knowledge unit of complete ability to express, including Ding Yi ﹑ and Ding Li ﹑ algorithms；
The local importance W of blocks of knowledge iL(i) computational methods：
Γ (j) is the set for the blocks of knowledge (forerunner adds subsequent) that blocks of knowledge j is connected directly.
V (k) is the upper of the blocks of knowledge (forerunner adds subsequent) being connected directly with blocks of knowledge j and its lower single order blocks of knowledge
Swim blocks of knowledge (subsequent forerunner's blocks of knowledge).
(2) using local importance come the transfer of learning probability matrix P between calculation knowledge unit.
If U is currently learning i, his (she) successfully study understands the possibility p for recognizing consequent j in next stepijIt depends on
Following two points：
I.pijOnly with U currently to study understand that j states are related, and that has learnt is not related with U before study j,
Because the possibility that condition Un (j)=1 is set up if L (i)=1 exists；, whereas if L (i)=0, then Un (j)=0 must
It sets up；
II.pijWith the local importance W of jL(j) related, the more early grasp W of learner UL(j) bigger blocks of knowledge j,
More be conducive to subsequently learn；Therefore, the learning process on KM has Markov characteristics, then pijCalculation formula it is as follows：
(3) blocks of knowledge and the markovian characteristic of absorbing state are combined, blocks of knowledge is divided into transition state and absorbing state,
Transition state is exactly the set for the blocks of knowledge for recognizing consequent, that is, V1, absorbing state is the set for not recognizing consequent, also
It is V2.Calculate the mean transferred time that transition state blocks of knowledge i is transferred to transition state j before being absorbed by the blocks of knowledge of absorbing state
The probability matrix M that matrix number Q and transition state are finally absorbed by absorbing state：
CiIt is the measurement to blocks of knowledge i significance levels during learning on entire KM, with the total path for passing through i on KM
The desired value of item number is measured, and is as follows：
Step1:Because KM is a directed acyclic graph, the upper learning processes of KM are an Absorbing CO2s, and state is moved
It is as follows to move probability matrix P：
Wherein G Shis |V1|×|V1|Rank matrix, B Shis |V1|×|V2|Rank matrix, E Shis |V2|×|V2|The unit matrix of rank, 0
Shi |V2|×|V1|The null matrix of rank；The i-th row jth column element p of PijCalculation formula (4.1), indicate study understand energy after i
Success learns to understand the possibility of j；
Step2：Calculate the mean transferred degree matrix Q between transition state.
If gij kIt indicates before being absorbed by absorbing state, transition state knowledge list is transferred to from transition state blocks of knowledge by k steps
The probability of member, G (k) indicate the state transition probability matrix walked by k between transition state blocks of knowledge, then
G (k)=(gij k)=Gk(k=0,1,2L)
If qijIndicate that transition state i arrives the mean transferred number of transition state j, Q being averaged between transition state before being absorbed
Transfer number matrix.Then
Q=(E-R)-1 (4.2)
Step3：The probability matrix M that transition state is finally absorbed by absorbing state.
If mijIndicate that transition state i starts the probability absorbed by absorbing state j, Metzler matrix is indicated by transition state finally by absorbing state
The probability matrix of absorption.
M=Q*B (4.3)
(4) calculation formula of centrad is obtained.Measure CiOverall thought be to utilize the total path item number for passing through i on KM
Desired value.
The blocks of knowledge of transition state can not only move to transition state blocks of knowledge, can also move to the knowledge of absorbing state
Unit.The blocks of knowledge of absorbing state only can with absorptive transition state, so mean transferred degree matrix Q between calculating transition state and
After the probability matrix M that transition state is finally absorbed by absorbing state, the calculation formula of centrad can be obtained：
2. in terms of blocks of knowledge difficulty measurement：
(1) the level horizontal L of blocks of knowledge is first measuredi
a.LiMetric parameter, LiIt is a relative quantity, the learning difficulty of quantitative description blocks of knowledge is used for, if Ri=φ,
Blocks of knowledge i is concept most basic in KM it is assumed that other blocks of knowledge j, study understand that the difficulty of i is minimum in opposite KM, because
This regulation Li=0；, whereas if Ri≠ φ, it is desirable that conditionIt sets up, then provides i's
LiAny one cognition former piece than it will be higher by a rank；
B. L is measurediCalculation formula it is as follows：
(2) specific difficulty is measured once again in same level level using statistics difficulty.
Under identical difficulty level, we will make the statistics difficulty of blocks of knowledge with study feedback, to each knowledge list
Member will provide the exercise for having learnt this blocks of knowledge, we are by calculating learner's score come the difficulty of calculation knowledge unit.
For subjective item, its degree-of-difficulty factor is：
Wherein P is degree-of-difficulty factor, s be all learners do the topic score and, n is to do the total number of persons inscribed, and f is the topic
Full marks.
For objective item, because objective item only has several possibilities, we give judgement selection, while also because of this
A possibility is given, so with certain conjecture, so needing to be corrected its difficulty for objective item：
P is initial difficulty coefficient, and n is to do the total number of persons inscribed, m be do to number, k be objective item it is existing guess can
Energy property (such as there are four option so k=4 for multiple-choice question), cp is the final degree-of-difficulty factor corrected.
Compared with prior art, it is an advantage of the invention that in terms of centrad：By with the sides centrality local
Method carrys out the local importance of calculation knowledge unit, calculates study migration probability matrix, between transition state mean transferred degree matrix Q and
Transition state is finally finally calculated the centrad of blocks of knowledge by the probability matrix M that absorbing state absorbs, in terms of difficulty：In conjunction with interior
Hold difficulty and statistics difficulty comes together to measure the difficulty of blocks of knowledge.Thus the center for the blocks of knowledge for measuring out
Degree and difficulty precise and high efficiency, are of great significance to optimizing inquiry learning and navigation learning path.
Description of the drawings
Fig. 1 is that blocks of knowledge centrad measures flow；
Fig. 2 is the calculation process of blocks of knowledge part importance；
Fig. 3 is that V values calculate in the importance of blocks of knowledge part；
Fig. 4 is that blocks of knowledge difficulty measures flow.
Specific implementation mode
The specific implementation mode of the present invention is described below in conjunction with the accompanying drawings, preferably so as to those skilled in the art
Understand the present invention.Requiring particular attention is that in the following description, when known function and the detailed description of design perhaps
When can desalinate the main contents of the present invention, these descriptions will be ignored herein.
A kind of blocks of knowledge centrad and difficulty quantization method：
One, is in terms of blocks of knowledge centrad as shown in Figure 1, specifically including following steps：
Step1：In the Knowledge Map built by blocks of knowledge, the local importance of calculation knowledge unit.
Step2：By the transfer of learning probability between the local importance calculation knowledge unit of blocks of knowledge.
Step3：Mean transferred number between transition state blocks of knowledge is calculated by the transfer of learning probability between blocks of knowledge
Matrix.
Step4：It is finally absorbed by absorbing state by the transfer of learning probability between blocks of knowledge to calculate transition state blocks of knowledge
Probability matrix.
Step5：Finally obtain blocks of knowledge centrad calculation formula.
It will calculate and calculate according to above-mentioned flow sequential recitation specific embodiment, including blocks of knowledge part importance below
Mean transferred degree matrix between the computational algorithm of transfer of learning probability between method, blocks of knowledge, transition state blocks of knowledge calculates
The computational algorithm and blocks of knowledge centrad for the probability matrix that algorithm, transition state blocks of knowledge are finally absorbed by absorbing state calculate
Algorithm.
1. blocks of knowledge part importance computational algorithm
Local importance flow such as Fig. 1, in the Knowledge Map built by blocks of knowledge, the V of first calculation knowledge unit
Value, in the S values of calculation knowledge unit, finally obtains the value of blocks of knowledge part importance W.
The local importance W of blocks of knowledge iL(i) computational methods：
Γ (j) is the set for the blocks of knowledge (forerunner adds subsequent) that blocks of knowledge j is connected directly
V (k) is the upper of the blocks of knowledge (forerunner adds subsequent) being connected directly with blocks of knowledge j and its lower single order blocks of knowledge
Swim blocks of knowledge (subsequent forerunner's blocks of knowledge).
S (j) values be it is with the V values of the blocks of knowledge j blocks of knowledge being connected directly and.
WL(i) value be it is with the S values of the blocks of knowledge i blocks of knowledge being connected directly and.
The calculating of V values is slightly complicated, as shown in figure 3, its computational algorithm is as follows：
Step1：Any one blocks of knowledge i in Knowledge Map, first judges whether the blocks of knowledge has forerunner.If before having
Drive then turns Step5, turns Step2 if without forerunner.
Step2：For the blocks of knowledge of no forerunner, then blocks of knowledge i must have subsequent blocks of knowledge j (because of a subject
Blocks of knowledge in journey is unlikely to be an isolated point), then judge whether its subsequent blocks of knowledge j has and subsequent know whether there is or not subsequent
Know unit.
Step3：If nothing, the number of the subsequent blocks of knowledge of ν (i)=i
Step4：If so, the then number of the subsequent subsequent blocks of knowledge of the number+i of the subsequent blocks of knowledge of ν (i)=i
Step5：For there is the blocks of knowledge of forerunner, then judge that whether there is or not subsequent by i.
Step6：If without subsequent, the number of forerunner's blocks of knowledge of V (i)=i
Step7：If having subsequent, judge that whether there is or not subsequent for the subsequent blocks of knowledge of i.
Step8：If nothing, the number of the subsequent blocks of knowledge of the number-i of forerunner's blocks of knowledge of ν (i)=i
Step9：If so, then after the number+i of the subsequent blocks of knowledge of the number-i of forerunner's blocks of knowledge of ν (i)=i
After subsequent blocks of knowledge number
2. the computational algorithm of the transfer of learning probability between blocks of knowledge.
Using local importance come the transfer of learning probability matrix P between calculation knowledge unit.
Knowledge Map is a not directed acyclic graph, if U currently just in learning knowledge unit i, he (she) in next step at
Work(study understands the possibility p of cognition consequent jijDepending on following two points：
I.pijOnly with U currently to study understand that j states are related, and that has learnt is not related with U before study j,
Because the possibility that condition Un (j)=1 is set up if L (i)=1 exists；, whereas if L (i)=0, then Un (j)=0 must
It sets up；
II.pijWith the local importance W of jL(j) related, the more early grasp W of learner UL(j) bigger blocks of knowledge j,
More be conducive to subsequently learn；Therefore, the learning process on KM has Markov characteristics, then pijCalculation formula it is as follows：
3. the mean transferred degree matrix computational algorithm between transition state blocks of knowledge.
In conjunction with blocks of knowledge and the markovian characteristic of absorbing state, blocks of knowledge is divided into transition state and absorbing state, is counted
Calculate the mean transferred degree matrix that transition state blocks of knowledge i is transferred to transition state j before being absorbed by the blocks of knowledge of absorbing state
Q。
Because KM is a directed acyclic graph, the upper learning processes of KM are an Absorbing CO2, state transition probability
Matrix P is as follows：
Wherein R is that transition state blocks of knowledge moves to | between transition state blocks of knowledge;V1|×|V1|Rank probability matrix, B are
Transition state blocks of knowledge moves to | between absorbing state blocks of knowledge;V1|×|V2|Rank probability matrix, E Shis |V2|×|V2|The list of rank
Bit matrix, 0 Shi |V2|×|V1|The null matrix of rank；The i-th row jth column element p of PijCalculation formula (1.4), indicate study reason
It can successfully learn to understand the possibility of j after solution i；
If gij kIt indicates before being absorbed by absorbing state, transition state knowledge list is transferred to from transition state blocks of knowledge by k steps
The probability of member, G (k) indicate the state transition probability matrix walked by k between transition state blocks of knowledge, then
G (k)=(gij k)=Gk(k=0,1,2L)
If qijIndicate that transition state i arrives the mean transferred number of transition state j, Q being averaged between transition state before being absorbed
Transfer number matrix.Then
Q=(E-R)-1 (4.2)
4. the computational algorithm for the probability matrix that transition state blocks of knowledge is finally absorbed by absorbing state.
If mijIndicate that transition state i starts the probability absorbed by absorbing state j, Metzler matrix is indicated by transition state finally by absorbing state
The probability matrix of absorption.
M=(E-R)-1× B=Q × B (4.3)
5. blocks of knowledge centrad computational algorithm.CiIt is measured by the desired value of the total path item number of i on KM, so
There is the calculation formula of centrad.
The blocks of knowledge of transition state can not only move to transition state blocks of knowledge, can also move to the knowledge of absorbing state
Unit.The blocks of knowledge of absorbing state only can with absorptive transition state, so mean transferred degree matrix Q between calculating transition state and
After the probability matrix M that transition state is finally absorbed by absorbing state, the calculation formula of centrad can be obtained：
Two, are in terms of blocks of knowledge difficulty measurement, flow chart such as Fig. 4：
Step1：First measure the level horizontal L of blocks of knowledgei, whether first judgemental knowledge unit i have forerunner：
If having
Li=L(forerunner of i)+1 (2.0)
If without if
Li=0 (2.1)
Step2：The statistics difficulty of the blocks of knowledge of same level level is calculated using user's answer.
Under identical difficulty level, we will make the statistics difficulty of blocks of knowledge with study feedback, to each knowledge list
Member will provide the exercise for having learnt this blocks of knowledge, we are by calculating learner's score come the difficulty of calculation knowledge unit.
Step3：Judge topic types, is objective item or subjective item.It is that objective item turns Step4, no person Step5.
Step4：For objective item, because objective item only has several possibilities, we give judgement selection, while
Because this possibility is given, with certain conjecture, so needing to be corrected its difficulty for objective item：
P is initial difficulty coefficient, and n is to do the total number of persons inscribed, m be do to number, k be objective item it is existing guess can
Energy property (such as there are four option so k=4 for multiple-choice question), cp is the final degree-of-difficulty factor corrected.
For subjective item, its degree-of-difficulty factor is：
Wherein P is degree-of-difficulty factor, s be all learners do the topic score and, n is to do the total number of persons inscribed, and f is the topic
Full marks.
Although the illustrative specific implementation mode of the present invention is described above, in order to the technology of the art
Personnel understand the present invention, it should be apparent that the present invention is not limited to the range of specific implementation mode, to the common skill of the art
For art personnel, if various change the attached claims limit and determine the spirit and scope of the present invention in, these
Variation is it will be apparent that all utilize the innovation and creation of present inventive concept in the row of protection.
Claims (7)
1. a kind of blocks of knowledge centrad and difficulty quantization method, which is characterized in that include the following steps：
In terms of blocks of knowledge centrad：
S1：According to the local importance of Knowledge Map calculation knowledge unit；
S2：Transfer of learning probability matrix between calculation knowledge unit；
S3：Calculate the mean transferred that transition state blocks of knowledge i is transferred to transition state j before being absorbed by the blocks of knowledge of absorbing state
The probability matrix M that degree matrix Q and transition state are finally absorbed by absorbing state；
S4：Obtain the calculation formula of centrad；
In terms of blocks of knowledge difficulty measurement：
S5：Measure the level horizontal of blocks of knowledge；
S6：Measure the difficulty of the statistics of blocks of knowledge once again in same level level.
2. blocks of knowledge centrad according to claim 1 and difficulty quantization method, which is characterized in that blocks of knowledge i's
The computational methods of local importance are as follows：
Knowledge Map is a directed acyclic graph (V, E), is denoted as KM；V=V1∪V2It is the knowing of including of ambit described in KM
Know the set of unit, E is the set for learning dependence in V between blocks of knowledge；Here blocks of knowledge refer to have it is complete
The basic knowledge unit of ability to express, including Ding Yi ﹑ and Ding Li ﹑ algorithms；
The local importance W of blocks of knowledge iL(i) computational methods：
Γ (j) is the set for the blocks of knowledge (forerunner adds subsequent) that blocks of knowledge j is connected directly, and V (k) is straight with blocks of knowledge j
Connect upstream blocks of knowledge (subsequent forerunner's knowledge list of connected blocks of knowledge (forerunner adds subsequent) and its next section blocks of knowledge
Member).
3. blocks of knowledge centrad according to claim 1 and difficulty quantization method, which is characterized in that the blocks of knowledge
Between transfer of learning probability matrix computational methods, its step are as follows：
If currently learning i using local importance come the transfer of learning probability matrix P U between calculation knowledge unit, he
(she) successfully study understands the possibility p for recognizing consequent j in next stepijDepending on following two points：
I.pijOnly with U currently to study understand that j states are related, and that has learnt is not related with U before study j because,
The possibility that condition Un (j)=1 is set up if L (i)=1 exists；, whereas if L (i)=0, then Un (j)=0 must be set up；
II.pijWith the local importance W of jL(j) related, the more early grasp W of learner UL(j) bigger blocks of knowledge j, it is more advantageous
In follow-up study；Therefore, the learning process on KM has Markov characteristics, then pijCalculation formula it is as follows：
4. blocks of knowledge centrad according to claim 1 and difficulty quantization method, which is characterized in that the transition state is known
Know the probability matrix that mean transferred degree matrix Q and transition state blocks of knowledge between unit are finally absorbed by absorbing state blocks of knowledge
The computational methods of M include the following steps：
step1：In conjunction with blocks of knowledge and the markovian characteristic of absorbing state, blocks of knowledge is divided into transition state and absorbing state,
Transition state is exactly the set for the blocks of knowledge for recognizing consequent, that is, V1, absorbing state is the set for not recognizing consequent, also
It is V2；
step2：Because KM is a directed acyclic graph, the upper learning processes of KM are an Absorbing CO2s, and state transition is general
Rate matrix P is as follows：
The i-th row jth column element p of PijCalculation formula (4.1), indicate study understand i after can successfully learn to understand the possibility of j
Property；
step3：Calculate the mean transferred degree matrix Q between transition state
Q=(E-R)-1(4.2)；
step4：The probability matrix M that transition state is finally absorbed by absorbing state
M=Q*B (4.3).
5. blocks of knowledge centrad according to claim 1 and difficulty quantization method, which is characterized in that the blocks of knowledge
Centrad CiIt is the measurement to blocks of knowledge i significance levels during learning on entire KM, with the total path for passing through i on KM
The desired value of item number is measured, and the calculation formula of centrad is as follows：
6. blocks of knowledge centrad according to claim 1 and difficulty quantization method, which is characterized in that the blocks of knowledge
The measure of level horizontal is as follows：
step1:LiMetric parameter, LiIt is a relative quantity, the learning difficulty of quantitative description blocks of knowledge is used for, if Ri=φ,
Blocks of knowledge i is concept most basic in KM it is assumed that other blocks of knowledge j, study understand that the difficulty of i is minimum in opposite KM, because
This regulation Li=0；, whereas if Ri≠ φ, it is desirable that conditionIt sets up, then provides i's
LiAny one cognition former piece than it will be higher by a rank；
Step2. L is measurediCalculation formula it is as follows：
7. blocks of knowledge centrad according to claim 1 and difficulty quantization method, which is characterized in that the same level
Measure statistical difficulty method is as follows again for horizontal blocks of knowledge：
Under identical difficulty level, we will measure blocks of knowledge difficulty with study feedback, to each blocks of knowledge, will provide
The exercise of this blocks of knowledge is learnt, we are by calculating learner's score come the difficulty of calculation knowledge unit；
For subjective item, its difficulty value is：
Wherein P is difficulty value, s be all learners do the topic score and, n is to do the total number of persons inscribed, and f is the full marks of the topic；
For objective item, because objective item only has several possibilities, we give judgement selection, while also because this can
Energy property is given, so with certain conjecture, so needing to be corrected its difficulty for objective item：
P is initial difficulty, and n is to do the total number of persons inscribed, m be do to number, k is the existing possibility (example guessed of objective item
As there are four option so k=4 for multiple-choice question), cp is the final difficulty corrected.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201810495007.3A CN108711125A (en) | 2018-07-24 | 2018-07-24 | A kind of blocks of knowledge centrad and difficulty quantization method |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201810495007.3A CN108711125A (en) | 2018-07-24 | 2018-07-24 | A kind of blocks of knowledge centrad and difficulty quantization method |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN108711125A true CN108711125A (en) | 2018-10-26 |
Family
ID=63868704
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201810495007.3A Pending CN108711125A (en) | 2018-07-24 | 2018-07-24 | A kind of blocks of knowledge centrad and difficulty quantization method |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN108711125A (en) |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110533187A (en) * | 2019-09-05 | 2019-12-03 | 河南师范大学 | A kind of knowledge quantization modulation and intelligent tutoring method |
| CN112508334A (en) * | 2020-11-06 | 2021-03-16 | 华中师范大学 | Personalized paper combining method and system integrating cognitive characteristics and test question text information |
-
2018
- 2018-07-24 CN CN201810495007.3A patent/CN108711125A/en active Pending
Cited By (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110533187A (en) * | 2019-09-05 | 2019-12-03 | 河南师范大学 | A kind of knowledge quantization modulation and intelligent tutoring method |
| CN112508334A (en) * | 2020-11-06 | 2021-03-16 | 华中师范大学 | Personalized paper combining method and system integrating cognitive characteristics and test question text information |
| CN112508334B (en) * | 2020-11-06 | 2023-09-01 | 华中师范大学 | Personalized paper grouping method and system integrating cognition characteristics and test question text information |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN108573628A (en) | The method that H-NTLA based on study track is recommended with extension knowledge point set | |
| CN106780224A (en) | A kind of Modeling Teaching of Mathematics learning system | |
| McGeorge et al. | Assessing lesbian, gay, and bisexual affirmative training in couple and family therapy: Establishing the validity of the faculty version of the affirmative training inventory | |
| Pamungkas et al. | Increasing Interest and Learning Outcomes of Elementary School Students in Style Material Through Blended Learning | |
| Yuqiao et al. | Construction of distance education classroom in architecture specialty based on internet of things technology | |
| Tseng et al. | Students' self-regulated learning, online information evaluative standards and online academic searching strategies | |
| CN108711125A (en) | A kind of blocks of knowledge centrad and difficulty quantization method | |
| CN107169903A (en) | Learning behavior evaluation method and system based on college teaching big data | |
| Han | A fuzzy logic and multilevel analysis-based evaluation algorithm for digital teaching quality in colleges and universities | |
| CN108573051A (en) | Knowledge point collection of illustrative plates based on big data analysis | |
| Liang et al. | Evaluation Method of Mixed Teaching Efficiency of College Teachers Based on Kirkpatrick Model | |
| Feaster et al. | Serious toys: three years of teaching computer science concepts in K-12 classrooms | |
| Hare et al. | Optimize student learning via random forest-based adaptive narrative game | |
| Wang | Exploration on the operation status and optimization strategy of networked teaching of physical education curriculum based on AI algorithm | |
| Iskander et al. | Outreach to K–12 teachers: Workshop in instrumentation, sensors, and engineering | |
| Zhang et al. | Design and practice of arduino experiments for" E&I" oriented education | |
| CN108846579A (en) | Knowledge quantity calculation method and system for subject knowledge | |
| Shi et al. | Research and design of teaching evaluation system based on fuzzy model | |
| Sun | Construction of the interactive educational knowledge graph and classification of student groups | |
| Baker | Critical Thinking in the Physics Curriculum | |
| Guo | A MOOC online teaching quality evaluation method based on fuzzy algorithm | |
| Li et al. | The Best Summer Job Selection Based on AHP | |
| TWI608454B (en) | A method of determinating interesting and the application of this method completed by the career research platform | |
| Moon et al. | Generative Interpretation: Toward Human-Like Evaluation for Educational Question-Answer Pair Generation | |
| Wong | An extended Singapore mathematics curriculum framework |
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 | ||
| WD01 | Invention patent application deemed withdrawn after publication |
Application publication date: 20181026 |
|
| WD01 | Invention patent application deemed withdrawn after publication |