CN107273490A - A kind of combination mistake topic recommendation method of knowledge based collection of illustrative plates - Google Patents
A kind of combination mistake topic recommendation method of knowledge based collection of illustrative plates Download PDFInfo
- Publication number
- CN107273490A CN107273490A CN201710449002.2A CN201710449002A CN107273490A CN 107273490 A CN107273490 A CN 107273490A CN 201710449002 A CN201710449002 A CN 201710449002A CN 107273490 A CN107273490 A CN 107273490A
- Authority
- CN
- China
- Prior art keywords
- examination question
- mrow
- knowledge
- msub
- recommendation
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Granted
Links
- 238000000034 method Methods 0.000 title claims abstract description 20
- 238000013507 mapping Methods 0.000 claims abstract description 19
- 238000012360 testing method Methods 0.000 claims abstract description 17
- 239000011159 matrix material Substances 0.000 claims abstract description 10
- 238000001914 filtration Methods 0.000 claims abstract description 9
- 238000004364 calculation method Methods 0.000 claims abstract description 3
- 239000013598 vector Substances 0.000 claims description 19
- 238000005516 engineering process Methods 0.000 claims description 13
- 230000006870 function Effects 0.000 claims description 8
- 238000012549 training Methods 0.000 claims description 6
- HUTDUHSNJYTCAR-UHFFFAOYSA-N ancymidol Chemical compound C1=CC(OC)=CC=C1C(O)(C=1C=NC=NC=1)C1CC1 HUTDUHSNJYTCAR-UHFFFAOYSA-N 0.000 claims description 5
- 238000005259 measurement Methods 0.000 claims description 3
- 230000000052 comparative effect Effects 0.000 claims description 2
- 230000002708 enhancing effect Effects 0.000 claims description 2
- 238000011156 evaluation Methods 0.000 claims description 2
- 238000011478 gradient descent method Methods 0.000 claims description 2
- 238000003475 lamination Methods 0.000 claims description 2
- 238000005070 sampling Methods 0.000 claims description 2
- 238000013519 translation Methods 0.000 claims description 2
- 238000001228 spectrum Methods 0.000 claims 1
- 239000000284 extract Substances 0.000 abstract 2
- 238000005065 mining Methods 0.000 abstract 1
- 241001494479 Pecora Species 0.000 description 2
- 238000011161 development Methods 0.000 description 2
- 238000004458 analytical method Methods 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 238000007596 consolidation process Methods 0.000 description 1
- 230000007547 defect Effects 0.000 description 1
- 230000001419 dependent effect Effects 0.000 description 1
- 238000010586 diagram Methods 0.000 description 1
- 238000004880 explosion Methods 0.000 description 1
- 230000006698 induction Effects 0.000 description 1
- 238000012545 processing Methods 0.000 description 1
- 238000011160 research Methods 0.000 description 1
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
- G06F16/90—Details of database functions independent of the retrieved data types
- G06F16/95—Retrieval from the web
- G06F16/953—Querying, e.g. by the use of web search engines
- G06F16/9535—Search customisation based on user profiles and personalisation
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
- G06F16/30—Information retrieval; Database structures therefor; File system structures therefor of unstructured textual data
- G06F16/36—Creation of semantic tools, e.g. ontology or thesauri
- G06F16/367—Ontology
Landscapes
- Engineering & Computer Science (AREA)
- Databases & Information Systems (AREA)
- Theoretical Computer Science (AREA)
- Data Mining & Analysis (AREA)
- Physics & Mathematics (AREA)
- General Engineering & Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Life Sciences & Earth Sciences (AREA)
- Animal Behavior & Ethology (AREA)
- Computational Linguistics (AREA)
- Information Retrieval, Db Structures And Fs Structures Therefor (AREA)
Abstract
The invention discloses a kind of combination of knowledge based collection of illustrative plates mistake topic recommendation method, the wrong topic related to its weak knowledge point can accurately be recommended to learner by this method.The invention includes building extracts knowledge from extensive unstructured examination question data, builds knowledge mapping；Wrong topic to learner carries out text mining, and participle extracts error topic keyword, and then determine that the mistake inscribes included knowledge point；Analyzed by the semantic acquaintance property to examination question, obtain the semantic neighbour of the knowledge point；It is mapped to by mistake topic knowledge point in knowledge mapping, obtains the examination question entity for meeting its knowledge point.Simultaneously by carrying out acquaintance property weight calculation to test item bank, the phase knowledge and magnanimity matrix of paper is obtained, the recommendation examination question of wrong topic is obtained using collaborative filtering.Finally, two kinds of recommendation results are further combined using modes such as weighting, mixing, stacking and first ranks, provide consequently recommended result.
Description
Technical field
The invention belongs to computer software technical field, and in particular to a kind of combination mistake topic recommendation side of knowledge based collection of illustrative plates
Method.
Background technology
With advancing by leaps and bounds for internet development, people have gradually entered an information from the epoch of an absence of information
The epoch of overload.The problem of growth of information explosion formula causes information overflow in network becomes extremely serious, for a user
Searched out from mass data just becomes difficult to oneself valuable data, and the useful information that some are seldom concerned is past
Toward being submerged in the ocean of information, as isolated island information.Commending system is the effective ways for solving problems, its essence
It is that the resource object for meeting its interest preference is found for user.
In recent years, in view of recommended technology has immense value in every field, the experts and scholars of all trades and professions join in
In the research of recommended technology, the new peak of recommended technology development is brought therefrom.The recommended technology of current main flow, bag
Include rule-based recommended technology, content-based recommendation technology and collaborative filtering.These recommended technologies are pushed away in traditional
Recommend in system and be used widely, and achieve certain success.Processing of the content-based recommendation to complex properties is not friendly enough
It is good, while preferable recommendation can not be produced to new user.Rule-based recommended technology, is overly dependent upon the language of professional domain
Expert defines syntax rule and carrys out extracting rule, it is necessary to take considerable time, cost of labor is too high, while moving costs is huge.Base
In the Collaborative Filtering Recommendation Algorithm of project, in the case where data are extremely sparse, similarity measurement is inaccurate, recommend poor quality not
Foot.Meanwhile, with continuing to develop for proposed algorithm, people also begin to recognize the defect present in existing commending system, such as by
Cold start-up problem, recommendation precision and the Gray sheep brought in Sparse.
Mistake is that institute is inevitable in learning, and mistake topic is the concentrated reflection of the difficult point that learner learns and blind spot, can be most
Show the learning actuality of learner in big degree, and with very strong authenticity.Can these mistake topics of induction and conclusion to learner
Grasping knowledge has tremendous influence, the recommendation of mistake topic knowledge point most important during also just being improved in study as learner
Key link.By using the method with higher recommendation accuracy rate, effectively it can provide accurate to learner using wrong topic collection
Knowledge recommendation, utilize wrong topic collection intensified learning person knowledge understanding and memory.The consolidation of mistake topic is practised as in study
An important part, if its main purpose, which just allows learner to practise its place field, grasps more weak knowledge point, according to
The demand of learner recommends the exercise of certain amount to it, to consolidate knowledge.
By building the knowledge mapping of examination question, entity, relation and path are represented all in the vector space of low-dimensional, so
The semantic acquaintance property of each knowledge point is calculated afterwards, and by mistake topic participle, extracting knowledge point, the semanteme for obtaining the knowledge point is near
Neighbour, the examination question that the semantic neighbour in all knowledge points is then provided in knowledge mapping is recommended.Wrong topic and the phase of test item bank are calculated simultaneously
Like property weight, the acquaintance matrix of examination question is obtained, corresponding recommendation is provided according to collaborative filtering.Recommend knot with reference to two kinds
Really, using modes such as weighting, mixing, stacking and first ranks, final recommendation results are provided.
The content of the invention
The present invention is intended to provide a kind of high wrong topic recommendation method of precision.
The wrong topic recommendation method that the present invention is provided is based on examination question knowledge mapping.By from extensive unstructured examination question
In, knowledge point is extracted, the knowledge mapping of examination question is built.On this basis, by the semantic analysis to mistake topic, the mistake is extracted
The knowledge point of topic, and then the semantic neighbour of the knowledge point is calculated in knowledge mapping, at the same time pass through collaborative filtering meter
The arest neighbors of the examination question is calculated, finally takes the modes such as weighting, mixing and first rank to combine two kinds of recommended technologies, is pushed away so as to improve
The degree of accuracy recommended.Simultaneously as possessing abundant semantic data, the cold start-up that commending system is also solved to a certain extent is asked
Topic.
The present invention proposes a kind of combination mistake topic commending system of knowledge based collection of illustrative plates, its overall flow frame diagram such as Fig. 1 institutes
Show.It includes following five modules.First module is the knowledge mapping that examination question is built according to test item bank；Second module is according to mistake
Examination question in topic input and test item bank to carry out similarity weight calculation to examination question, recommends so as to provide corresponding arest neighbors；The
Three modules are that the vector representation of examination question and knowledge point is calculated according to knowledge mapping；4th module is to utilize semantic similarity, is provided
Semantic neighbour in mistake topic knowledge based collection of illustrative plates；5th module is to combine the recommendation results in the second module and the 4th module, profit
Final recommendation results are provided with modes such as weighting, mixing, stacking and first ranks.
The present invention has the beneficial effect that：
1st, the knowledge mapping that the present invention is built for examination question, the knowledge point vector representation generated using TransE algorithms will
Knowledge mapping quantizes expression, has taken into full account the semantic similarity between knowledge point, can accurately understand between knowledge point
Semantic association, so as to improve the accuracy of knowledge mapping recommendation results.
2nd, the present invention can be provided according to different learners meets the examination question of its knowledge blind spot and recommends, it is to avoid Gray
Sheep problems.
3rd, due to there is abundant test item bank to provide semantic data, it is to avoid the commending system cold start-up that Sparse is brought is asked
Topic.
4th, a variety of recommended technologies are combined, make recommendation results more accurate.
Brief description of the drawings
Fig. 1 is knowledge based collection of illustrative plates combined recommendation overall system architecture；
Fig. 2 is that history knowledge-ID excavates example；
Fig. 3 is TransE algorithm model figures；
Embodiment
In order that the purpose of the present invention, technical scheme and feature are more clearly understood, below in conjunction with specific embodiment, and
1-3, further refinement explanation is carried out to the present invention referring to the drawings.
Step one：
Participle operation is carried out to the per pass examination question in test item bank, the keyword of per pass examination question is obtained, keyword is made into one
The knowledge feature of step is extracted, and obtains the knowledge point corresponding to per pass examination question, thereby determines that the mapping between knowledge point and examination question is closed
System, so as to build using each knowledge point and examination question as node, keyword is the examination question knowledge mapping on side.
Step 2：
The similarity weight of examination question and knowledge point is calculated, so as to obtain paper and the similarity matrix of knowledge point.The phase
The i-th row of knowledge and magnanimity matrix jth row represent examination question i and are knowledge point j proportions.According to the examination question-knowledge dot matrix, calculate wrong
Topic and the similarity of existing examination question in test item bank, are then weighted sequence to examination question, pick out sequence in top-k preceding k roads
Examination question provides the arest neighbors recommendation obtained by collaborative filtering as output.
Step 3：
In knowledge mapping, similar node is also often similar semantically.So, for the knowledge point in examination question
For, different knowledge points, the semantic information that different examination questions may be included is roughly the same.Using TransE algorithms, enhancing association
With the semantic information of examination question in examination question similarity matrix in filter algorithm.As shown in Figure 3：By each examination question triple example
Relation relation in (head, relation, tail) regards the translation from entity head to entity tail as, by constantly adjusting
Whole h, r and t (head, relation and tail vector), make (h+r) as equal with t as possible, i.e. h+r ≈ t.
In the training process of model, TransE uses largest interval method, and its object function is as follows：
Wherein, S is the triple in knowledge base, and S' is the triple of negative sampling, as obtained by replacing h or t.γ is value
Spacing distance parameter more than 0, [x]+represent Positive Function, i.e. x>When 0, [x] +=x；As x≤0, [x] +=0.Gradient is more
It is new only to calculate apart from d (h+r, t) with d (h'+r, t').Using gradient descent method so that loss function is optimal.Work as model
After the completion of training, the vector representation of entity and relation is obtained.
Step 4：
By the TransE algorithms described in step 3, examination question is represented to turn into one group of low-dimensional real-valued vectors.According to TransE
Algorithm, its entity vector S is provided for mistake topic1, then for any examination question in test item bank, provide the entity of any examination question to
Measure S2.Calculated for the similarity between measurement twice examination question by vector space cosine similarity：
The codomain of sim functions is (0,1), when phase knowledge and magnanimity more level off to 1 when, two vectors more represent it is semantic also
It is more identical, that is to say, that examination question S1And S2Just possess more identical knowledge point.Similarly, when similarity gets over convergence 0, two
The semanteme of vector representation is more differed, i.e. examination question S1And S2Just possess the knowledge point more differed.Counted using Semantic Similarity
Calculate, calculate wrong topic and the semantic similarity of existing examination question in test item bank, examination question is ranked up, will sort in top-k preceding k roads
Examination question is exported as the semantic neighbour of knowledge based collection of illustrative plates.
Step 5：
Examination question the recommendation results A and B obtained by step 2 and step 4 is integrated, using weighting, mixing, stacking and
The modes such as first rank are further combined to two kinds of recommendation results, provide consequently recommended result.
<1>Recommendation is combined using weighting technique：
Collect for two kinds of recommendation results A and B as recommendation, then evaluation and system of the comparative learning person to recommendation results
Predict whether be consistent, the model of weighting is generated according to the obtained result of training, weight is dynamically adjusted.Provided according to model
Most rational combining weights are weighted combination, the recommendation examination question collection after final output weighted array.
<2>Married operation：
Set (C)=set (A)+set (B)
for i in set(C):
print(i)
Wherein, A, B are respectively the recommendation collection that two kinds of recommended technologies are provided, and C is two kinds of unions for recommending to collect, and i is pushing away in C
Recommend examination question.
<3>The mode of stacking is combined recommendation.Lamination techniques are as follows：
I and j are respectively to recommend the recommendation examination question in collection A, B.
<4>Recommendation is combined using first level technique：
For the recommendation collection A obtained in step 2 using collaborative filtering, step 4 kind is put into as input, is calculated
Go out the semantic similarity with examination question in test item bank that per pass in A intends recommending examination question, then sort, by preceding k roads knowledge based collection of illustrative plates
Semantic neighbour recommend examination question as consequently recommended.
Claims (1)
1. a kind of combination mistake topic recommendation method of knowledge based collection of illustrative plates, it is characterised in that：Realize that this method includes following five moulds
Block；First module is the knowledge mapping that examination question is built according to test item bank；Second module is according in the input of mistake topic and test item bank
Examination question to carry out similarity weight calculation to examination question, recommends so as to provide corresponding arest neighbors；3rd module is according to knowledge graph
Spectrum calculates the vector representation of examination question and knowledge point；4th module is to utilize semantic similarity, in error topic knowledge based collection of illustrative plates
Semantic neighbour；5th module be combine the second module and the 4th module in recommendation results, using weighting, mixing, stacking and
The modes such as first rank provide final recommendation results；
Step one：
Participle operation is carried out to the per pass examination question in test item bank, the keyword of per pass examination question is obtained, further is made to keyword
Knowledge feature is extracted, and is obtained the knowledge point corresponding to per pass examination question, is thereby determined that the mapping relations between knowledge point and examination question, from
And build using each knowledge point and examination question as node, keyword is the examination question knowledge mapping on side；
Step 2：
The similarity weight of examination question and knowledge point is calculated, so as to obtain paper and the similarity matrix of knowledge point；The phase knowledge and magnanimity
The i-th row of matrix jth row represent examination question i and are knowledge point j proportions；According to the examination question-knowledge dot matrix, calculate wrong topic with
The similarity of existing examination question, is then weighted sequence to examination question in test item bank, picks out preceding k road examination question of the sequence in top-k
As output, the arest neighbors recommendation obtained by collaborative filtering is provided；
Step 3：
In knowledge mapping, similar node is also often similar semantically；So, for the knowledge point in examination question
Speech, different knowledge points, the semantic information that different examination questions may be included is roughly the same；Using TransE algorithms, enhancing collaboration
In filter algorithm in examination question similarity matrix examination question semantic information；As shown in Figure 3：By each examination question triple example (head,
Relation, tail) in relation relation regard translation from entity head to entity tail as, by constantly adjusting h, r
With t (head, relation and tail vector), make (h+r) as equal with t as possible, i.e. h+r ≈ t；
In the training process of model, TransE uses largest interval method, and its object function is as follows：
<mrow>
<mi>L</mi>
<mo>=</mo>
<munder>
<mo>&Sigma;</mo>
<mrow>
<mo>(</mo>
<mi>h</mi>
<mo>,</mo>
<mi>r</mi>
<mo>,</mo>
<mi>t</mi>
<mo>)</mo>
<mo>&Element;</mo>
<mi>S</mi>
</mrow>
</munder>
<munder>
<mo>&Sigma;</mo>
<mrow>
<mo>(</mo>
<msup>
<mi>h</mi>
<mo>&prime;</mo>
</msup>
<mo>,</mo>
<mi>r</mi>
<mo>,</mo>
<msup>
<mi>t</mi>
<mo>&prime;</mo>
</msup>
<mo>)</mo>
<mo>&Element;</mo>
<msubsup>
<mi>S</mi>
<mrow>
<mo>(</mo>
<mi>h</mi>
<mo>,</mo>
<mi>r</mi>
<mo>,</mo>
<mi>t</mi>
<mo>)</mo>
</mrow>
<mo>&prime;</mo>
</msubsup>
</mrow>
</munder>
<msub>
<mrow>
<mo>&lsqb;</mo>
<mi>&gamma;</mi>
<mo>+</mo>
<mi>d</mi>
<mrow>
<mo>(</mo>
<mi>h</mi>
<mo>+</mo>
<mi>r</mi>
<mo>,</mo>
<mi>t</mi>
<mo>)</mo>
</mrow>
<mo>-</mo>
<mi>d</mi>
<mrow>
<mo>(</mo>
<msup>
<mi>h</mi>
<mo>&prime;</mo>
</msup>
<mo>+</mo>
<mi>r</mi>
<mo>,</mo>
<msup>
<mi>t</mi>
<mo>&prime;</mo>
</msup>
<mo>)</mo>
</mrow>
<mo>&rsqb;</mo>
</mrow>
<mo>+</mo>
</msub>
</mrow>
Wherein, S is the triple in knowledge base, and S' is the triple of negative sampling, as obtained by replacing h or t；γ is that value is more than
0 spacing distance parameter, [x]+represent Positive Function, i.e. x>When 0, [x] +=x；As x≤0, [x] +=0；Gradient updating is only
It need to calculate apart from d (h+r, t) with d (h'+r, t')；Using gradient descent method so that loss function is optimal；Work as model training
After the completion of, obtain the vector representation of entity and relation；
Step 4：
By the TransE algorithms described in step 3, examination question is represented to turn into one group of low-dimensional real-valued vectors；According to TransE algorithms,
Its entity vector S is provided for mistake topic1, then for any examination question in test item bank, provide the entity vector S of any examination question2；
Calculated for the similarity between measurement twice examination question by vector space cosine similarity：
<mrow>
<mi>c</mi>
<mi>o</mi>
<mi>s</mi>
<mrow>
<mo>(</mo>
<msub>
<mi>s</mi>
<mn>1</mn>
</msub>
<mo>,</mo>
<msub>
<mi>s</mi>
<mn>2</mn>
</msub>
<mo>)</mo>
</mrow>
<mo>=</mo>
<mfrac>
<mrow>
<msub>
<mi>s</mi>
<mn>1</mn>
</msub>
<mo>&CenterDot;</mo>
<msub>
<mi>s</mi>
<mn>2</mn>
</msub>
</mrow>
<mrow>
<mo>|</mo>
<mo>|</mo>
<msub>
<mi>s</mi>
<mn>1</mn>
</msub>
<mo>|</mo>
<mo>|</mo>
<mo>&CenterDot;</mo>
<mo>|</mo>
<mo>|</mo>
<msub>
<mi>s</mi>
<mn>2</mn>
</msub>
<mo>|</mo>
<mo>|</mo>
</mrow>
</mfrac>
</mrow>
<mrow>
<mi>s</mi>
<mi>i</mi>
<mi>m</mi>
<mrow>
<mo>(</mo>
<msub>
<mi>s</mi>
<mn>1</mn>
</msub>
<mo>,</mo>
<msub>
<mi>s</mi>
<mn>2</mn>
</msub>
<mo>)</mo>
</mrow>
<mo>=</mo>
<mfrac>
<mrow>
<mn>1</mn>
<mo>+</mo>
<mi>c</mi>
<mi>o</mi>
<mi>s</mi>
<mrow>
<mo>(</mo>
<msub>
<mi>s</mi>
<mn>1</mn>
</msub>
<mo>,</mo>
<msub>
<mi>s</mi>
<mn>2</mn>
</msub>
<mo>)</mo>
</mrow>
</mrow>
<mn>2</mn>
</mfrac>
</mrow>
The codomain of sim functions is (0,1), when phase knowledge and magnanimity more level off to 1 when, the semantic also more phase that two vectors are more represented
Together, that is to say, that examination question S1And S2Just possess more identical knowledge point；Similarly, when similarity gets over convergence 0, two vectors
The semanteme of expression is more differed, i.e. examination question S1And S2Just possess the knowledge point more differed；Calculated using Semantic Similarity,
Wrong topic and the semantic similarity of existing examination question in test item bank are calculated, examination question is ranked up, is tried sorting in top-k preceding k roads
Inscribe and exported as the semantic neighbour of knowledge based collection of illustrative plates；
Step 5：
Examination question the recommendation results A and B obtained by step 2 and step 4 is integrated, and utilizes weighting, mixing, stacking and first level
The mode such as not is further combined to two kinds of recommendation results, provides consequently recommended result；
<1>Recommendation is combined using weighting technique：
For two kinds of recommendation results A and B as recommending to collect, then comparative learning person is to the evaluation of recommendation results, pre- with system
Whether survey is consistent, and the result obtained according to training generates the model of weighting, dynamically adjusts weight；The most conjunction provided according to model
The combining weights of reason are weighted combination, the recommendation examination question collection after final output weighted array；
<2>Married operation：
Set (C)=set (A)+set (B)
for i in set(C):
print(i)
Wherein, A, B are respectively the recommendation collection that two kinds of recommended technologies are provided, and C is two kinds of unions for recommending to collect, and i is the recommendation examination in C
Topic；
<3>The mode of stacking is combined recommendation；Lamination techniques are as follows：
I and j are respectively to recommend the recommendation examination question in collection A, B；
<4>Recommendation is combined using first level technique：
For the recommendation collection A obtained in step 2 using collaborative filtering, step 4 kind is put into as input, A is calculated
Middle per pass intends recommending the semantic similarity with examination question in test item bank of examination question, then sorts, by the language of preceding k roads knowledge based collection of illustrative plates
Adopted neighbour recommends examination question as consequently recommended.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201710449002.2A CN107273490B (en) | 2017-06-14 | 2017-06-14 | Combined wrong question recommendation method based on knowledge graph |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201710449002.2A CN107273490B (en) | 2017-06-14 | 2017-06-14 | Combined wrong question recommendation method based on knowledge graph |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN107273490A true CN107273490A (en) | 2017-10-20 |
| CN107273490B CN107273490B (en) | 2020-04-17 |
Family
ID=60067690
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201710449002.2A Expired - Fee Related CN107273490B (en) | 2017-06-14 | 2017-06-14 | Combined wrong question recommendation method based on knowledge graph |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN107273490B (en) |
Cited By (33)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN107909520A (en) * | 2017-11-02 | 2018-04-13 | 浙江工商大学 | The method and apparatus that make the test based on examination question correlation |
| CN107944489A (en) * | 2017-11-17 | 2018-04-20 | 清华大学 | Extensive combination chart feature learning method based on structure semantics fusion |
| CN108491378A (en) * | 2018-03-08 | 2018-09-04 | 国网福建省电力有限公司 | Power information O＆M intelligent response system |
| CN108920556A (en) * | 2018-06-20 | 2018-11-30 | 华东师范大学 | Recommendation expert method based on subject knowledge map |
| CN109033129A (en) * | 2018-06-04 | 2018-12-18 | 桂林电子科技大学 | Multi-source Information Fusion knowledge mapping based on adaptive weighting indicates learning method |
| CN109035088A (en) * | 2018-07-19 | 2018-12-18 | 江苏黄金屋教育发展股份有限公司 | Adaptive learning method based on mistake topic |
| CN109189944A (en) * | 2018-09-27 | 2019-01-11 | 桂林电子科技大学 | Personalized recommending scenery spot method and system based on user's positive and negative feedback portrait coding |
| CN109447713A (en) * | 2018-10-31 | 2019-03-08 | 国家电网公司 | A kind of recommended method and device of knowledge based map |
| CN109766548A (en) * | 2018-12-29 | 2019-05-17 | 北京京师乐学教育科技有限公司 | Examination point recognition methods, device, equipment and computer storage medium |
| CN110134945A (en) * | 2019-04-15 | 2019-08-16 | 平安科技（深圳）有限公司 | The recognition methods of exercise examination point, device, equipment and storage medium |
| CN110362723A (en) * | 2019-05-31 | 2019-10-22 | 平安国际智慧城市科技股份有限公司 | A kind of topic character representation method, apparatus and storage medium |
| CN110362688A (en) * | 2019-06-14 | 2019-10-22 | 北京百度网讯科技有限公司 | Examination question mask method, device, equipment and computer readable storage medium |
| CN110472061A (en) * | 2019-07-08 | 2019-11-19 | 郑州大学 | A kind of knowledge mapping fusion method based on short text similarity calculation |
| CN110472155A (en) * | 2019-07-03 | 2019-11-19 | 五邑大学 | Collaborative recommendation method, device, equipment and the storage medium of knowledge based map |
| CN110866174A (en) * | 2018-08-17 | 2020-03-06 | 阿里巴巴集团控股有限公司 | Pushing method, device and system for court trial problems |
| CN110968669A (en) * | 2019-11-30 | 2020-04-07 | 南京森林警察学院 | Intelligent video analysis police test question classification recommendation method |
| CN111091454A (en) * | 2019-11-05 | 2020-05-01 | 新华智云科技有限公司 | Financial public opinion recommendation method based on knowledge graph |
| CN111125339A (en) * | 2019-11-26 | 2020-05-08 | 华南师范大学 | Test question recommendation method based on formal concept analysis and knowledge graph |
| CN111125350A (en) * | 2019-12-17 | 2020-05-08 | 语联网（武汉）信息技术有限公司 | Method and device for generating LDA topic model based on bilingual parallel corpus |
| CN111339258A (en) * | 2020-02-29 | 2020-06-26 | 西安理工大学 | University computer basic exercise recommendation method based on knowledge graph |
| CN111563166A (en) * | 2020-05-28 | 2020-08-21 | 浙江学海教育科技有限公司 | Pre-training model method for mathematical problem classification |
| CN111813920A (en) * | 2020-07-06 | 2020-10-23 | 龙马智芯（珠海横琴）科技有限公司 | Learning strategy generation method, device, generation equipment and readable storage medium |
| CN112149001A (en) * | 2020-09-17 | 2020-12-29 | 北京师范大学 | Learning companion recommendation system and method |
| CN112800182A (en) * | 2021-02-10 | 2021-05-14 | 联想(北京)有限公司 | Test question generation method and device |
| CN113221547A (en) * | 2021-01-21 | 2021-08-06 | 重庆邮电大学 | Test question recommendation method based on information extraction and knowledge graph |
| CN113254629A (en) * | 2021-06-07 | 2021-08-13 | 重庆第二师范学院 | Learning content recommendation method and system based on artificial intelligence |
| CN113282723A (en) * | 2021-05-21 | 2021-08-20 | 上海伯禹信息科技有限公司 | Deep knowledge tracking pre-training method based on graph neural network |
| CN113742474A (en) * | 2021-11-08 | 2021-12-03 | 北京博瑞彤芸科技股份有限公司 | Intelligent question and answer method and device based on knowledge graph |
| CN113763767A (en) * | 2021-08-25 | 2021-12-07 | 赣州市加薪教育科技有限公司 | Learning test question pushing method and device, computer equipment and storage medium |
| WO2021253480A1 (en) * | 2020-06-19 | 2021-12-23 | 平安科技（深圳）有限公司 | Intelligent exercise recommendation method and apparatus, computer device and storage medium |
| CN113836320A (en) * | 2021-11-26 | 2021-12-24 | 北京世纪好未来教育科技有限公司 | Exercise recommendation method and device, storage medium and electronic equipment |
| CN114495124A (en) * | 2022-01-18 | 2022-05-13 | 上海应用技术大学 | Test question score analysis and exercise improvement system |
| CN117150151A (en) * | 2023-11-01 | 2023-12-01 | 之江实验室 | Wrong question analysis and test question recommendation system and method based on large language model |
Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103593792A (en) * | 2013-11-13 | 2014-02-19 | 复旦大学 | Individual recommendation method and system based on Chinese knowledge mapping |
| US20140351052A1 (en) * | 2013-05-24 | 2014-11-27 | Harbhajan S. Khalsa | Contextual Product Recommendation Engine |
| CN104794664A (en) * | 2015-03-03 | 2015-07-22 | 上海复勤商务咨询有限公司 | Intelligent learning ability evaluation and knowledge recommendation method |
| CN105355111A (en) * | 2015-12-02 | 2016-02-24 | 华中师范大学 | After-class reinforced learning system based on learning situation analysis |
| CN105512214A (en) * | 2015-11-28 | 2016-04-20 | 华中师范大学 | Knowledge database, construction method and learning situation diagnosis system |
-
2017
- 2017-06-14 CN CN201710449002.2A patent/CN107273490B/en not_active Expired - Fee Related
Patent Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20140351052A1 (en) * | 2013-05-24 | 2014-11-27 | Harbhajan S. Khalsa | Contextual Product Recommendation Engine |
| CN103593792A (en) * | 2013-11-13 | 2014-02-19 | 复旦大学 | Individual recommendation method and system based on Chinese knowledge mapping |
| CN104794664A (en) * | 2015-03-03 | 2015-07-22 | 上海复勤商务咨询有限公司 | Intelligent learning ability evaluation and knowledge recommendation method |
| CN105512214A (en) * | 2015-11-28 | 2016-04-20 | 华中师范大学 | Knowledge database, construction method and learning situation diagnosis system |
| CN105355111A (en) * | 2015-12-02 | 2016-02-24 | 华中师范大学 | After-class reinforced learning system based on learning situation analysis |
Non-Patent Citations (1)
| Title |
|---|
| 林海伦 等: "面向网络大数据的知识融合方法综述", 《计算机学报》 * |
Cited By (51)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN107909520A (en) * | 2017-11-02 | 2018-04-13 | 浙江工商大学 | The method and apparatus that make the test based on examination question correlation |
| CN107944489A (en) * | 2017-11-17 | 2018-04-20 | 清华大学 | Extensive combination chart feature learning method based on structure semantics fusion |
| CN107944489B (en) * | 2017-11-17 | 2018-10-16 | 清华大学 | Extensive combination chart feature learning method based on structure semantics fusion |
| CN108491378A (en) * | 2018-03-08 | 2018-09-04 | 国网福建省电力有限公司 | Power information O＆M intelligent response system |
| CN108491378B (en) * | 2018-03-08 | 2021-11-09 | 国网福建省电力有限公司 | Intelligent response system for operation and maintenance of electric power information |
| CN109033129B (en) * | 2018-06-04 | 2021-08-03 | 桂林电子科技大学 | Multi-source information fusion knowledge graph representation learning method based on self-adaptive weight |
| CN109033129A (en) * | 2018-06-04 | 2018-12-18 | 桂林电子科技大学 | Multi-source Information Fusion knowledge mapping based on adaptive weighting indicates learning method |
| CN108920556A (en) * | 2018-06-20 | 2018-11-30 | 华东师范大学 | Recommendation expert method based on subject knowledge map |
| CN108920556B (en) * | 2018-06-20 | 2021-11-19 | 华东师范大学 | Expert recommending method based on discipline knowledge graph |
| CN109035088A (en) * | 2018-07-19 | 2018-12-18 | 江苏黄金屋教育发展股份有限公司 | Adaptive learning method based on mistake topic |
| CN110866174A (en) * | 2018-08-17 | 2020-03-06 | 阿里巴巴集团控股有限公司 | Pushing method, device and system for court trial problems |
| CN110866174B (en) * | 2018-08-17 | 2023-05-02 | 阿里巴巴集团控股有限公司 | Pushing method, device and system for court trial questions |
| CN109189944A (en) * | 2018-09-27 | 2019-01-11 | 桂林电子科技大学 | Personalized recommending scenery spot method and system based on user's positive and negative feedback portrait coding |
| CN109447713A (en) * | 2018-10-31 | 2019-03-08 | 国家电网公司 | A kind of recommended method and device of knowledge based map |
| CN109766548A (en) * | 2018-12-29 | 2019-05-17 | 北京京师乐学教育科技有限公司 | Examination point recognition methods, device, equipment and computer storage medium |
| CN109766548B (en) * | 2018-12-29 | 2023-01-31 | 北京京师乐学教育科技有限公司 | Examination point identification method, examination point identification device, examination point identification equipment and computer storage medium |
| CN110134945A (en) * | 2019-04-15 | 2019-08-16 | 平安科技（深圳）有限公司 | The recognition methods of exercise examination point, device, equipment and storage medium |
| CN110134945B (en) * | 2019-04-15 | 2024-04-23 | 平安科技（深圳）有限公司 | Method, device, equipment and storage medium for identifying examination points of exercise |
| CN110362723B (en) * | 2019-05-31 | 2022-06-21 | 平安国际智慧城市科技股份有限公司 | Topic feature representation method, device and storage medium |
| CN110362723A (en) * | 2019-05-31 | 2019-10-22 | 平安国际智慧城市科技股份有限公司 | A kind of topic character representation method, apparatus and storage medium |
| CN110362688B (en) * | 2019-06-14 | 2022-03-25 | 北京百度网讯科技有限公司 | Test question labeling method, device and equipment and computer readable storage medium |
| CN110362688A (en) * | 2019-06-14 | 2019-10-22 | 北京百度网讯科技有限公司 | Examination question mask method, device, equipment and computer readable storage medium |
| CN110472155A (en) * | 2019-07-03 | 2019-11-19 | 五邑大学 | Collaborative recommendation method, device, equipment and the storage medium of knowledge based map |
| CN110472061A (en) * | 2019-07-08 | 2019-11-19 | 郑州大学 | A kind of knowledge mapping fusion method based on short text similarity calculation |
| CN111091454A (en) * | 2019-11-05 | 2020-05-01 | 新华智云科技有限公司 | Financial public opinion recommendation method based on knowledge graph |
| CN111125339A (en) * | 2019-11-26 | 2020-05-08 | 华南师范大学 | Test question recommendation method based on formal concept analysis and knowledge graph |
| CN111125339B (en) * | 2019-11-26 | 2023-05-09 | 华南师范大学 | Test question recommendation method based on formal concept analysis and knowledge graph |
| CN110968669A (en) * | 2019-11-30 | 2020-04-07 | 南京森林警察学院 | Intelligent video analysis police test question classification recommendation method |
| CN111125350B (en) * | 2019-12-17 | 2023-05-12 | 传神联合（北京）信息技术有限公司 | Method and device for generating LDA topic model based on bilingual parallel corpus |
| CN111125350A (en) * | 2019-12-17 | 2020-05-08 | 语联网（武汉）信息技术有限公司 | Method and device for generating LDA topic model based on bilingual parallel corpus |
| CN111339258A (en) * | 2020-02-29 | 2020-06-26 | 西安理工大学 | University computer basic exercise recommendation method based on knowledge graph |
| CN111339258B (en) * | 2020-02-29 | 2023-03-03 | 西安理工大学 | University computer basic exercise recommendation method based on knowledge graph |
| CN111563166B (en) * | 2020-05-28 | 2024-02-13 | 浙江学海教育科技有限公司 | Pre-training model method for classifying mathematical problems |
| CN111563166A (en) * | 2020-05-28 | 2020-08-21 | 浙江学海教育科技有限公司 | Pre-training model method for mathematical problem classification |
| WO2021253480A1 (en) * | 2020-06-19 | 2021-12-23 | 平安科技（深圳）有限公司 | Intelligent exercise recommendation method and apparatus, computer device and storage medium |
| CN111813920B (en) * | 2020-07-06 | 2021-04-13 | 龙马智芯（珠海横琴）科技有限公司 | Learning strategy generation method, device, generation equipment and readable storage medium |
| CN111813920A (en) * | 2020-07-06 | 2020-10-23 | 龙马智芯（珠海横琴）科技有限公司 | Learning strategy generation method, device, generation equipment and readable storage medium |
| CN112149001A (en) * | 2020-09-17 | 2020-12-29 | 北京师范大学 | Learning companion recommendation system and method |
| CN112149001B (en) * | 2020-09-17 | 2023-04-25 | 北京师范大学 | Learning companion recommendation system and method |
| CN113221547A (en) * | 2021-01-21 | 2021-08-06 | 重庆邮电大学 | Test question recommendation method based on information extraction and knowledge graph |
| CN113221547B (en) * | 2021-01-21 | 2022-05-03 | 重庆邮电大学 | Test question recommendation method based on information extraction and knowledge graph |
| CN112800182A (en) * | 2021-02-10 | 2021-05-14 | 联想(北京)有限公司 | Test question generation method and device |
| CN113282723A (en) * | 2021-05-21 | 2021-08-20 | 上海伯禹信息科技有限公司 | Deep knowledge tracking pre-training method based on graph neural network |
| CN113254629B (en) * | 2021-06-07 | 2022-07-26 | 重庆第二师范学院 | Learning content recommendation method and system based on artificial intelligence |
| CN113254629A (en) * | 2021-06-07 | 2021-08-13 | 重庆第二师范学院 | Learning content recommendation method and system based on artificial intelligence |
| CN113763767A (en) * | 2021-08-25 | 2021-12-07 | 赣州市加薪教育科技有限公司 | Learning test question pushing method and device, computer equipment and storage medium |
| CN113742474A (en) * | 2021-11-08 | 2021-12-03 | 北京博瑞彤芸科技股份有限公司 | Intelligent question and answer method and device based on knowledge graph |
| CN113836320A (en) * | 2021-11-26 | 2021-12-24 | 北京世纪好未来教育科技有限公司 | Exercise recommendation method and device, storage medium and electronic equipment |
| CN114495124A (en) * | 2022-01-18 | 2022-05-13 | 上海应用技术大学 | Test question score analysis and exercise improvement system |
| CN117150151A (en) * | 2023-11-01 | 2023-12-01 | 之江实验室 | Wrong question analysis and test question recommendation system and method based on large language model |
| CN117150151B (en) * | 2023-11-01 | 2024-02-20 | 之江实验室 | Wrong question analysis and test question recommendation system and method based on large language model |
Also Published As
| Publication number | Publication date |
|---|---|
| CN107273490B (en) | 2020-04-17 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN107273490A (en) | A kind of combination mistake topic recommendation method of knowledge based collection of illustrative plates | |
| CN107239446B (en) | A kind of intelligence relationship extracting method based on neural network Yu attention mechanism | |
| CN104391942B (en) | Short essay eigen extended method based on semantic collection of illustrative plates | |
| CN112508334B (en) | Personalized paper grouping method and system integrating cognition characteristics and test question text information | |
| CN107967255A (en) | A kind of method and system for judging text similarity | |
| CN107885853A (en) | A kind of combined type file classification method based on deep learning | |
| CN109543084A (en) | A method of establishing the detection model of the hidden sensitive text of network-oriented social media | |
| CN107945204A (en) | A kind of Pixel-level portrait based on generation confrontation network scratches drawing method | |
| CN105045907A (en) | Method for constructing visual attention-label-user interest tree for personalized social image recommendation | |
| CN104881689B (en) | A kind of multi-tag Active Learning sorting technique and system | |
| CN107291688A (en) | Judgement document's similarity analysis method based on topic model | |
| CN106886543A (en) | The knowledge mapping of binding entity description represents learning method and system | |
| CN103064903B (en) | Picture retrieval method and device | |
| CN108710680A (en) | It is a kind of to carry out the recommendation method of the film based on sentiment analysis using deep learning | |
| CN106991161A (en) | A kind of method for automatically generating open-ended question answer | |
| CN104536881A (en) | Public testing error report priority sorting method based on natural language analysis | |
| CN112990296A (en) | Image-text matching model compression and acceleration method and system based on orthogonal similarity distillation | |
| CN106682696A (en) | Multi-example detection network based on refining of online example classifier and training method thereof | |
| CN102156706A (en) | Mentor recommendation system and method | |
| CN110765254A (en) | Multi-document question-answering system model integrating multi-view answer reordering | |
| CN110059716A (en) | A kind of building of CNN-LSTM-SVM network model and MOOC discontinue one's studies prediction technique | |
| CN111914162B (en) | Method for guiding personalized learning scheme based on knowledge graph | |
| CN107832295A (en) | The title system of selection of reading machine people and system | |
| CN106407482B (en) | A kind of network academic report category method based on multi-feature fusion | |
| CN106097204A (en) | A kind of work commending system towards cold start-up User and recommendation 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 | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant | ||
| CF01 | Termination of patent right due to non-payment of annual fee |
Granted publication date: 20200417 |