CN109359215B - Video intelligent pushing method and system - Google Patents
Video intelligent pushing method and system Download PDFInfo
- Publication number
- CN109359215B CN109359215B CN201811465037.6A CN201811465037A CN109359215B CN 109359215 B CN109359215 B CN 109359215B CN 201811465037 A CN201811465037 A CN 201811465037A CN 109359215 B CN109359215 B CN 109359215B
- Authority
- CN
- China
- Prior art keywords
- video
- knowledge
- teaching
- test question
- student
- 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.)
- Active
Links
Classifications
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
- Y02D—CLIMATE CHANGE MITIGATION TECHNOLOGIES IN INFORMATION AND COMMUNICATION TECHNOLOGIES [ICT], I.E. INFORMATION AND COMMUNICATION TECHNOLOGIES AIMING AT THE REDUCTION OF THEIR OWN ENERGY USE
- Y02D10/00—Energy efficient computing, e.g. low power processors, power management or thermal management
Landscapes
- Electrically Operated Instructional Devices (AREA)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
The invention discloses an intelligent video pushing method and system, wherein the method comprises the following steps: collecting input test questions, and adding test question attribute labels to the test questions; establishing a test question portrait for a test question, and establishing a text semantic network based on a knowledge graph for a test question text on the basis of the test question attribute label; analyzing the learning result assessment condition of the students and the big data of weak items to obtain the learning emotion images of the students; collecting teaching videos, respectively analyzing images and voices in the teaching videos, and marking video portrait labels on the teaching videos; constructing a knowledge graph; and constructing a pushing model and pushing the video. According to the intelligent video pushing method and system, the video is pushed while the test questions are pushed, the video is the recorded knowledge point explanation video, the processing analysis of pictures and audios is automatically carried out, the corresponding test questions are obtained, the association of the test questions and the video is realized, and students obtain the video explanation of the corresponding knowledge points while the students obtain the pushed test questions.
Description
Technical Field
The invention relates to the technical field of computer and data processing, in particular to an intelligent video pushing method and system.
Background
Along with the development of the Internet, the education industry promotes remote education in ten years ago, and remote video teaching and electronic document sharing are realized through an Internet virtual classroom, so that teachers and students form interaction of teaching and learning on the Internet; the coming of the current 4G era is more convenient to learn, the user can learn on line through palm tools such as a mobile phone and the like more conveniently and directly only through a heavy computer and a mobile phone with a large flow, and the wireless network enables daily interaction of people to be more effective through the rapid network propulsion of the 4G.
In the prior art, a method and a system for pushing test questions of a related video are disclosed, wherein the method comprises the steps of marking video attribute information on video data according to attribute information, marking test question attribute information on the test questions, and the attribute information comprises the following steps: grade, subject, version, and chapter; forming a mapping relation between video data and test questions according to the video attribute information and the test question attribute information; extracting video attribute information of video data currently played by students, and matching corresponding test questions according to the mapping relation; and pushing the corresponding test questions to the students according to a preset pushing rule.
In the prior art, video data and test questions are associated, and matched test questions can be pushed according to video attribute information of the video data so as to enable students to conduct targeted consolidation exercises.
However, when students receive test question pushing in other situations, for example, test question pushing based on wrong questions and the like, no targeted video explanation exists.
Disclosure of Invention
In view of the above, the present invention provides a method and a system for intelligent pushing of video, comprising the steps of:
collecting input test questions, adding test question attribute labels to the test questions, wherein the test question attribute labels comprise: the method comprises the steps of determining core knowledge points and related knowledge points of a test question according to a test question attribute label to form a test question library, wherein the core knowledge points are knowledge points which are mainly inspected by the test question, and the related knowledge points are knowledge points which are inspected in the test question and are related to the core knowledge points;
establishing a test question portrait for a test question, and establishing a text semantic network based on a knowledge graph for a test question text on the basis of the test question attribute label;
analyzing the learning result assessment condition of the student and the big data of the weak item to obtain the learning emotion image of the student: examination and homework exercises are carried out on the test questions in the test question library by students, and the learning progress, the learning completion degree and the knowledge point grasping degree of the students are analyzed from single departments and comprehensive ranking fluctuation; analyzing the learning level of a certain subject of the student at a class thereof, the teaching level of the class of the student at the class thereof, the regional teaching level condition of the school of the student and the regional teaching level condition of the student, combining the knowledge point examination surface and the answering condition of the examination, the homework and the online exercise of the student, and accurately analyzing the learning breakthrough points suitable for the current progress of the student by combining the knowledge point difficulty and the provincial past examination condition of the school;
collecting teaching videos, respectively analyzing images and voices in the teaching videos, and marking video portrait labels on the teaching videos;
the method for constructing the knowledge graph comprises the following steps:
entity link, namely carrying out named entity identification based on BiLSTM+CRF algorithm on voice and test question text in video to carry out entity extraction, linking the extracted entities with the same entity information on different sub-knowledge maps, and realizing entity link by using a CoLink unsupervised learning framework;
extracting knowledge graph characteristics: carrying out knowledge graph feature extraction on the basis of a tranD algorithm of knowledge graph feature learning, and accurately describing the entity by virtue of contextual entity features of the entity;
the method for constructing the push model and the video push comprises the following steps:
establishing consistency or relevance among students, test questions and teaching videos based on the knowledge graph, finding out relevance paths among different types of entities by using the knowledge graph, and calculating the relevance of entity nodes based on an iterative weight propagation algorithm; performing characteristic semantic association on students, test questions and teaching videos; potential association of students, questions and teaching videos is found out based on a Manifold algorithm, and semantic association among different labels is found out by combining DN-DBpedia corpus with an ESA model;
building a test question and video pushing algorithm based on a convolutional neural network and an attention mechanism; the method comprises the steps of dividing areas, teaching material versions, schools and grade schedules, combining teaching quality, student learning situation images, knowledge point difficulty and knowledge point examination frequency, combining knowledge point map comprehensive matching, matching a classification strategy, and formulating a video with optimal classification rate and classification difficulty for pushing;
through learning data of the classmates, accurately follow-up the teaching rhythm, and carry out video pushing by combining student learning situation portraits and matched classification strategies;
after video learning is completed, high-quality name school precision questions are precisely matched and pushed, the learning effect is consolidated, the mastering degree is verified, and corresponding data flow back to student learning situation images.
Preferably, the collecting teaching video analyzes the image and the voice respectively, and marks the teaching video with video portrait labels, further,
performing voice recognition on the audio part of the teaching video, recording a voice recognition result, performing knowledge extraction on the voice recognition result based on a knowledge graph to obtain a teaching video knowledge keyword, and recording a corresponding video playing position;
image recognition is carried out on the teaching video image part, face recognition, optical character recognition and formula recognition are respectively carried out, and teacher information in the video is recognized through the face recognition; knowledge extraction based on knowledge graph is carried out through optical character recognition, and knowledge points and knowledge keywords in the video image are extracted; for the test question explanation type video, identifying text information comprising test question stems, answers and analysis by using optical characters; the formula identifies a LaTex formula and a formula structure type in the video image;
carrying out semantic analysis based on a knowledge graph on the voice recognition result and text information in the picture, and analyzing a teaching scene, wherein the teaching scene comprises a teaching knowledge point range, a teaching video knowledge keyword and a video type, and the video type comprises a test question explanation type and a knowledge point explanation type; establishing semantic relation based on knowledge graph for test question attribute and test question text;
the method comprises the steps of marking a video tag, wherein the tag comprises a knowledge point range of teaching, a video type, teacher information, knowledge keywords and video positions thereof, a LaTex formula, a formula structure type and a video semantic relation based on a knowledge map, and when the video type is a test question explanation type, the tag further comprises a test question stem, an answer and analysis.
Preferably, the learning data of the same class students is used for accurately following the teaching rhythm, and the video pushing is performed by combining the student learning situation portrait and the matched classification strategy, and further, the link address of the video is associated with the test questions, and is pushed to the students.
Preferably, the method further comprises: and after receiving the link address of the video and the test questions, the students call a player to play the teaching video stored in the link address of the video.
The invention also discloses a video intelligent pushing system, which comprises test question collecting equipment, teaching video collecting equipment, storage equipment, student status statistics module, a processor, a knowledge graph constructing module, a pushing model constructing and a video pushing module constructing, wherein,
the test question acquisition equipment is connected with the processor, and is used for acquiring and inputting test questions and sending the test questions to the processor;
the teaching video acquisition equipment is connected with the processor and used for acquiring teaching videos and sending the teaching videos to the processor;
the student learning condition statistics module is connected with the storage device, and analyzes the learning progress, the learning completion degree and the knowledge point grasping degree of the students from single departments and comprehensive ranking fluctuation through examination and homework practice of the students on the test questions in the test question library; analyzing the learning level of a certain subject of the student at a class thereof, the teaching level of the class of the student at the class thereof, the regional teaching level condition of the school of the student and the regional teaching level condition of the student, combining the knowledge point examination surface and the answering condition of the examination, the homework and the online exercise of the student, combining the knowledge point difficulty and the past examination condition accurate analysis of the province of the school to the learning breakthrough point suitable for the current progress of the student, obtaining the learning condition image of the student and feeding back to the storage device for storage;
the processor is respectively connected with the test question acquisition equipment, the teaching video acquisition equipment and the storage equipment, adds test question attribute labels to the test questions, and the test question attribute labels comprise: the method comprises the steps of determining core knowledge points and related knowledge points of a test question according to a test question attribute label to form a test question library, wherein the core knowledge points are knowledge points which are mainly inspected by the test question, the related knowledge points are knowledge points which are inspected in the test question and are related to the core knowledge points, establishing a test question image for the test question, establishing a text semantic network based on the knowledge map for a test question text on the basis of the test question attribute label, and sending the test question library and the test question image to a storage device for storage; respectively analyzing images and voices in the teaching video, marking video portrait labels on the teaching video, and sending the video portrait labels to a storage device for storage;
the knowledge graph constructing module is respectively connected with the storage equipment, the pushing model constructing module and the video pushing module and is used for entity linking, named entity identification based on BiLSTM+CRF algorithm is carried out on voice and test question text in video stored in the storage equipment, entity extraction is carried out, the extracted entities link the same entity information on different sub-knowledge graphs, and entity linking is realized by using a CoLink unsupervised learning framework; extracting knowledge graph characteristics: carrying out knowledge graph feature extraction on the basis of a tranD algorithm of knowledge graph feature learning, and accurately describing the entity by virtue of contextual entity features of the entity;
the building pushing model and the video pushing module are connected with the building knowledge graph module, consistency or relevance among students, test questions and teaching videos is built based on the knowledge graph, the knowledge graph is utilized to find the relevance paths among different types of entities, and the relevance of the entity nodes is calculated based on an iterative weight propagation algorithm; performing characteristic semantic association on students, test questions and teaching videos; potential association of students, questions and teaching videos is found out based on a Manifold algorithm, and semantic association among different labels is found out by combining DN-DBpedia corpus with an ESA model; building a test question and video pushing algorithm based on a convolutional neural network and an attention mechanism; the method comprises the steps of dividing areas, teaching material versions, schools and grade schedules, combining teaching quality, student learning situation images, knowledge point difficulty and knowledge point examination frequency, combining knowledge point map comprehensive matching, matching a classification strategy, and formulating a video with optimal classification rate and classification difficulty for pushing; through the study data of the classmates, the teaching rhythm is accurately followed, the student learning situation portrait and the matched classification strategy are combined, video pushing is carried out, and the link address and the test question of the video are pushed to the student at the same time.
Preferably, the test question collecting device is a scanner or a camera, and the teaching video collecting device is a camera.
Preferably, the storage device is cloud storage or physical storage.
Preferably, the processor analyzes the image and the voice in the teaching video, marks the teaching video with a video portrait tag, further,
performing voice recognition on the audio part of the teaching video, recording a voice recognition result, performing knowledge extraction on the voice recognition result based on a knowledge graph to obtain a teaching video knowledge keyword, and recording a corresponding video playing position;
image recognition is carried out on the teaching video image part, face recognition, optical character recognition and formula recognition are respectively carried out, and teacher information in the video is recognized through the face recognition; knowledge extraction based on knowledge graph is carried out through optical character recognition, and knowledge points and knowledge keywords in the video image are extracted; for the test question explanation type video, identifying text information comprising test question stems, answers and analysis by using optical characters; the formula identifies a LaTex formula and a formula structure type in the video image;
carrying out semantic analysis based on a knowledge graph on the voice recognition result and text information in the picture, and analyzing a teaching scene, wherein the teaching scene comprises a teaching knowledge point range, a teaching video knowledge keyword and a video type, and the video type comprises a test question explanation type and a knowledge point explanation type; establishing semantic relation based on knowledge graph for test question attribute and test question text;
the method comprises the steps of marking a video tag, wherein the tag comprises a knowledge point range of teaching, a video type, teacher information, knowledge keywords and video positions thereof, a LaTex formula, a formula structure type and a video semantic relation based on a knowledge map, and when the video type is a test question explanation type, the tag further comprises a test question stem, an answer and analysis.
Preferably, the system further comprises a player, and after receiving the link address and the test questions of the video, the student calls the player to play the teaching video stored in the link address of the video.
The invention associates the video with the test questions, and pushes the video simultaneously when pushing the test questions.
Compared with the prior art, the intelligent video pushing method and system provided by the invention have the advantages that at least the following effects are realized:
according to the intelligent video pushing method and system, the video is pushed while the test questions are pushed, the video is the recorded knowledge point explanation video, the processing analysis of pictures and audios is automatically carried out, the corresponding test questions are obtained, the association of the test questions and the video is realized, and students obtain the video explanation of the corresponding knowledge points while the students obtain the pushed test questions.
Of course, it is not necessary for any one product embodying the invention to achieve all of the technical effects described above at the same time.
Other features of the present invention and its advantages will become apparent from the following detailed description of exemplary embodiments of the invention, which proceeds with reference to the accompanying drawings.
Drawings
The accompanying drawings, which are incorporated in and constitute a part of this specification, illustrate embodiments of the invention and together with the description, serve to explain the principles of the invention.
FIG. 1 is a flow chart of an intelligent video pushing method in the embodiment 1;
fig. 2 is a schematic structural diagram of the video intelligent push system in embodiment 2;
fig. 3 is a schematic structural diagram of an intelligent video push system in embodiment 3.
Detailed Description
Various exemplary embodiments of the present invention will now be described in detail with reference to the accompanying drawings. It should be noted that: the relative arrangement of the components and steps, numerical expressions and numerical values set forth in these embodiments do not limit the scope of the present invention unless it is specifically stated otherwise.
The following description of at least one exemplary embodiment is merely exemplary in nature and is in no way intended to limit the invention, its application, or uses.
Techniques, methods, and apparatus known to one of ordinary skill in the relevant art may not be discussed in detail, but are intended to be part of the specification where appropriate.
In all examples shown and discussed herein, any specific values should be construed as merely illustrative, and not a limitation. Thus, other examples of exemplary embodiments may have different values.
It should be noted that: like reference numerals and letters denote like items in the following figures, and thus once an item is defined in one figure, no further discussion thereof is necessary in subsequent figures.
Example 1:
with reference to fig. 1, this embodiment provides an intelligent video pushing method, which specifically includes the following steps:
step 101: collecting input test questions, adding test question attribute labels to the test questions, wherein the test question attribute labels comprise: the method comprises the steps of determining core knowledge points and related knowledge points of a test question according to a test question attribute label to form a test question library, wherein the core knowledge points are knowledge points which are mainly inspected by the test question, and the related knowledge points are knowledge points which are inspected in the test question and are related to the core knowledge points;
the method for collecting and inputting the test questions can be carried out by adopting a network uploading mode after editing, and also can be carried out by adopting a scanning mode after scanning by a scanner.
Step 102: establishing a test question portrait for a test question, and establishing a text semantic network based on a knowledge graph for a test question text on the basis of the test question attribute label;
step 103: analyzing the learning result assessment condition of the student and the big data of the weak item to obtain the learning emotion image of the student: examination and homework exercises are carried out on the test questions in the test question library by students, and the learning progress, the learning completion degree and the knowledge point grasping degree of the students are analyzed from single departments and comprehensive ranking fluctuation; analyzing the learning level of a certain subject of the student at a class thereof, the teaching level of the class of the student at the class thereof, the regional teaching level condition of the school of the student and the regional teaching level condition of the student, combining the knowledge point examination surface and the answering condition of the examination, the homework and the online exercise of the student, and accurately analyzing the learning breakthrough points suitable for the current progress of the student by combining the knowledge point difficulty and the provincial past examination condition of the school;
step 104: collecting teaching videos, respectively analyzing images and voices in the teaching videos, and marking video portrait labels on the teaching videos;
the teaching video is recorded by the camera and then uploaded, the teaching video can be specially aimed at a certain knowledge point, and the teaching video is specifically aimed at the core knowledge point and can be taught for the test questions.
Specifically, the method of step 104 is as follows:
performing voice recognition on the audio part of the teaching video, recording a voice recognition result, performing knowledge extraction on the voice recognition result based on a knowledge graph to obtain a teaching video knowledge keyword, and recording a corresponding video playing position;
image recognition is carried out on the teaching video image part, face recognition, optical character recognition and formula recognition are respectively carried out, and teacher information in the video is recognized through the face recognition; knowledge extraction based on knowledge graph is carried out through optical character recognition, and knowledge points and knowledge keywords in the video image are extracted; for the test question explanation type video, identifying text information comprising test question stems, answers and analysis by using optical characters; the formula identifies a LaTex formula and a formula structure type in the video image;
carrying out semantic analysis based on a knowledge graph on the voice recognition result and text information in the picture, and analyzing a teaching scene, wherein the teaching scene comprises a teaching knowledge point range, a teaching video knowledge keyword and a video type, and the video type comprises a test question explanation type and a knowledge point explanation type; establishing semantic relation based on knowledge graph for test question attribute and test question text;
the method comprises the steps of marking a video tag, wherein the tag comprises a knowledge point range of teaching, a video type, teacher information, knowledge keywords and video positions thereof, a LaTex formula, a formula structure type and a video semantic relation based on a knowledge map, and when the video type is a test question explanation type, the tag further comprises a test question stem, an answer and analysis.
Step 105: the knowledge graph is constructed as follows:
entity link, namely carrying out named entity identification based on BiLSTM+CRF algorithm on voice and test question text in video to carry out entity extraction, linking the extracted entities with the same entity information on different sub-knowledge maps, and realizing entity link by using a CoLink unsupervised learning framework;
extracting knowledge graph characteristics: carrying out knowledge graph feature extraction on the basis of a tranD algorithm of knowledge graph feature learning, and accurately describing the entity by virtue of contextual entity features of the entity;
step 106: the method for constructing the push model and the video push comprises the following steps:
a push model building step: establishing consistency or relevance among students, test questions and teaching videos based on the knowledge graph, finding out relevance paths among different types of entities by using the knowledge graph, and calculating the relevance of entity nodes based on an iterative weight propagation algorithm; performing characteristic semantic association on students, test questions and teaching videos; potential association of students, questions and teaching videos is found out based on a Manifold algorithm, and semantic association among different labels is found out by combining DN-DBpedia corpus with an ESA model;
building a test question and video pushing algorithm based on a convolutional neural network and an attention mechanism;
and (3) video pushing: the method comprises the steps of dividing areas, teaching material versions, schools and grade schedules, combining teaching quality, student learning situation images, knowledge point difficulty and knowledge point examination frequency, combining knowledge point map comprehensive matching, matching a classification strategy, and formulating a video with optimal classification rate and classification difficulty for pushing;
through learning data of the classmates, accurately follow-up the teaching rhythm, and carry out video pushing by combining student learning situation portraits and matched classification strategies;
after video learning is completed, high-quality name school precision questions are precisely matched and pushed, the learning effect is consolidated, the mastering degree is verified, and corresponding data flow back to student learning situation images.
The learning data of the same class classmates is used for accurately following the teaching rhythm, and the video pushing is carried out by combining the student learning situation portrait and the matched classification strategy.
Of course, the method also comprises the steps of: and after receiving the link address of the video and the test questions, the students call a player to play the teaching video stored in the link address of the video.
The invention can have the following application scenarios:
1. video pushing under student wrong question scene:
based on student learning condition images and test question images, matching video images, retrieving the best matched test question explanation video and pushing the best matched test question explanation video to students.
2. Pushing test questions under weak knowledge points of students:
and searching out the best matched knowledge point explanation video based on student portrait and learning weak knowledge point analysis results, and pushing the best matched knowledge point explanation video to students.
3. Study exercise test question pushing after teaching video study:
based on the student portrait and the video portrait, the best matching test questions are searched out and pushed to students. When knowledge points or knowledge keywords appear in the video, the best matched test questions or knowledge point explanation video is searched out based on the student portrait, the knowledge points or the knowledge keywords and is pushed to students.
Example 2:
with reference to fig. 2, this embodiment provides a video intelligent pushing system, which includes a test question collecting device 201, a teaching video collecting device 202, a storage device 205, a student status statistics module 203, a processor 204, a knowledge graph construction module 206, and a pushing model and video pushing module 207, wherein,
the test question collection device 201 is connected with the processor 204, and is used for collecting and inputting test questions and sending the test questions to the processor 204;
the teaching video acquisition device 202 is connected with the processor 204, and is used for acquiring teaching videos and sending the teaching videos to the processor 204;
the student learning condition statistics module 203 is connected with the storage device 205, and analyzes the learning progress, learning completion degree and knowledge point grasping degree of the students from the single department and comprehensive ranking fluctuation through examination and homework practice of the students on the test questions in the test question library; analyzing the learning level of a certain subject of the student at a class thereof, the teaching level of the class of the student at the class thereof, the regional teaching level condition of the school of the student and the regional teaching level condition of the student, combining the knowledge point examination surface and the answering condition of the examination, the homework and the online exercise of the student, combining the knowledge point difficulty and the past examination condition of the school to accurately analyze the learning breakthrough point suitable for the current progress of the student, obtaining the learning condition image of the student and feeding back to the storage device 205 for storage;
the processor 204 is respectively connected with the test question collecting device 201 and the teaching video collecting device 202, and adds a test question attribute tag to the test questions, wherein the test question attribute tag comprises: the method comprises the steps of determining core knowledge points and related knowledge points of a test question according to a test question attribute label to form a test question library, wherein the core knowledge points are knowledge points which are mainly inspected by the test question, the related knowledge points are knowledge points which are inspected in the test question and are related to the core knowledge points, establishing a test question image for the test question, establishing a text semantic network based on the knowledge map for a test question text on the basis of the test question attribute label, and sending the test question library and the test question image to a storage device 205 for storage; respectively analyzing images and voices in the teaching video, marking video portrait labels on the teaching video, and sending the video portrait labels to a storage device 205 for storage;
the knowledge graph constructing module 206 is respectively connected with the storage device 205, the push model constructing module and the video push module 207, and is used for entity linking, performing named entity recognition based on a BiLSTM+CRF algorithm on the voice and test question text in the video stored in the storage device 205, performing entity extraction, linking the extracted entities with the same entity information on different sub-knowledge graphs, and implementing entity linking by using a CoLink unsupervised learning framework; extracting knowledge graph characteristics: carrying out knowledge graph feature extraction on the basis of a tranD algorithm of knowledge graph feature learning, and accurately describing the entity by virtue of contextual entity features of the entity;
the push model building and video pushing module 207 is connected to the knowledge graph building module 206 to build a push model: establishing consistency or relevance among students, test questions and teaching videos based on the knowledge graph, finding out relevance paths among different types of entities by using the knowledge graph, and calculating the relevance of entity nodes based on an iterative weight propagation algorithm; performing characteristic semantic association on students, test questions and teaching videos; potential association of students, questions and teaching videos is found out based on a Manifold algorithm, and semantic association among different labels is found out by combining DN-DBpedia corpus with an ESA model; building a test question and video pushing algorithm based on a convolutional neural network and an attention mechanism; the method comprises the steps of dividing areas, teaching material versions, schools and grade schedules, combining teaching quality, student learning situation images, knowledge point difficulty and knowledge point examination frequency, combining knowledge point map comprehensive matching, matching a classification strategy, and formulating a video with optimal classification rate and classification difficulty for pushing; through the study data of the classmates, the teaching rhythm is accurately followed, the student learning situation portrait and the matched classification strategy are combined, video pushing is carried out, and the link address and the test question of the video are pushed to the student at the same time.
The test question collecting device 201 is a scanner or a camera, and the teaching video collecting device 202 is a camera.
The storage device 205 is cloud storage or physical storage.
The processor 204 analyzes the images and voices in the teaching video and marks the teaching video with video portrait labels, further,
performing voice recognition on the audio part of the teaching video, recording a voice recognition result, performing knowledge extraction on the voice recognition result based on a knowledge graph to obtain a teaching video knowledge keyword, and recording a corresponding video playing position;
image recognition is carried out on the teaching video image part, face recognition, optical character recognition and formula recognition are respectively carried out, and teacher information in the video is recognized through the face recognition; knowledge extraction based on knowledge graph is carried out through optical character recognition, and knowledge points and knowledge keywords in the video image are extracted; for the test question explanation type video, identifying text information comprising test question stems, answers and analysis by using optical characters; the formula identifies a LaTex formula and a formula structure type in the video image;
carrying out semantic analysis based on a knowledge graph on the voice recognition result and text information in the picture, and analyzing a teaching scene, wherein the teaching scene comprises a teaching knowledge point range, a teaching video knowledge keyword and a video type, and the video type comprises a test question explanation type and a knowledge point explanation type; establishing semantic relation based on knowledge graph for test question attribute and test question text;
the method comprises the steps of marking a video tag, wherein the tag comprises a knowledge point range of teaching, a video type, teacher information, knowledge keywords and video positions thereof, a LaTex formula, a formula structure type and a video semantic relation based on a knowledge map, and when the video type is a test question explanation type, the tag further comprises a test question stem, an answer and analysis.
In the case of example 3,
on the basis of embodiment 2, the video intelligent pushing system in this embodiment further includes a player 208, and after receiving the link address and the test question of the video, the student calls the player 208 to play the teaching video stored in the link address of the video.
According to the embodiment, the intelligent video pushing method and system provided by the invention have the advantages that at least the following advantages are realized:
according to the intelligent video pushing method and system, the video is pushed while the test questions are pushed, the video is the recorded knowledge point explanation video, the processing analysis of pictures and audios is automatically carried out, the corresponding test questions are obtained, the association of the test questions and the video is realized, and students obtain the video explanation of the corresponding knowledge points while the students obtain the pushed test questions.
While certain specific embodiments of the invention have been described in detail by way of example, it will be appreciated by those skilled in the art that the above examples are for illustration only and are not intended to limit the scope of the invention. It will be appreciated by those skilled in the art that modifications may be made to the above embodiments without departing from the scope and spirit of the invention. The scope of the invention is defined by the appended claims.
Claims (9)
1. The intelligent video pushing method is characterized by comprising the following steps:
collecting input test questions, adding test question attribute labels to the test questions, wherein the test question attribute labels comprise: the method comprises the steps of determining core knowledge points and related knowledge points of a test question according to a test question attribute label to form a test question library, wherein the core knowledge points are knowledge points which are mainly inspected by the test question, and the related knowledge points are knowledge points which are inspected in the test question and are related to the core knowledge points;
establishing a test question portrait for a test question, and establishing a text semantic network based on a knowledge graph for a test question text on the basis of the test question attribute label;
analyzing the learning result assessment condition of the student and the big data of the weak item to obtain the learning emotion image of the student: examination and homework exercises are carried out on the test questions in the test question library by students, and the learning progress, the learning completion degree and the knowledge point grasping degree of the students are analyzed from single departments and comprehensive ranking fluctuation; analyzing the learning level of a certain subject of the student at a class thereof, the teaching level of the class of the student at the class thereof, the regional teaching level condition of the school of the student and the regional teaching level condition of the student, combining the knowledge point examination surface and the answering condition of the examination, the homework and the online exercise of the student, and accurately analyzing the learning breakthrough points suitable for the current progress of the student by combining the knowledge point difficulty and the provincial past examination condition of the school;
collecting teaching videos, respectively analyzing images and voices in the teaching videos, and marking video portrait labels on the teaching videos;
the method for constructing the knowledge graph comprises the following steps:
entity link, namely carrying out named entity identification based on BiLSTM+CRF algorithm on voice and test question text in video to carry out entity extraction, linking the extracted entities with the same entity information on different sub-knowledge maps, and realizing entity link by using a CoLink unsupervised learning framework;
extracting knowledge graph characteristics: carrying out knowledge graph feature extraction on the basis of a tranD algorithm of knowledge graph feature learning, and accurately describing the entity by virtue of contextual entity features of the entity;
the method for constructing the push model and the video push comprises the following steps:
establishing consistency or relevance among students, test questions and teaching videos based on the knowledge graph, finding out relevance paths among different types of entities by using the knowledge graph, and calculating the relevance of entity nodes based on an iterative weight propagation algorithm; performing characteristic semantic association on students, test questions and teaching videos; potential association of students, questions and teaching videos is found out based on a Manifold algorithm, and semantic association among different labels is found out by combining DN-DBpedia corpus with an ESA model;
building a test question and video pushing algorithm based on a convolutional neural network and an attention mechanism; the method comprises the steps of dividing areas, teaching material versions, schools and grade schedules, combining teaching quality, student learning situation images, knowledge point difficulty and knowledge point examination frequency, combining knowledge point map comprehensive matching, matching a classification strategy, and formulating a video with optimal classification rate and classification difficulty for pushing;
through learning data of the classmates, accurately follow-up the teaching rhythm, and carry out video pushing by combining student learning situation portraits and matched classification strategies;
after video learning is completed, high-quality name school precision questions are precisely matched and pushed, the learning effect is consolidated, the mastering degree is verified, and corresponding data flow back to student learning situation images.
2. The intelligent video pushing method according to claim 1, wherein the collecting teaching video, analyzing the image and the voice respectively, and marking the teaching video with video portrait tags, further,
performing voice recognition on the audio part of the teaching video, recording a voice recognition result, performing knowledge extraction on the voice recognition result based on a knowledge graph to obtain a teaching video knowledge keyword, and recording a corresponding video playing position;
image recognition is carried out on the teaching video image part, face recognition, optical character recognition and formula recognition are respectively carried out, and teacher information in the video is recognized through the face recognition; knowledge extraction based on knowledge graph is carried out through optical character recognition, and knowledge points and knowledge keywords in the video image are extracted; for the test question explanation type video, identifying text information comprising test question stems, answers and analysis by using optical characters; the formula identifies a LaTex formula and a formula structure type in the video image;
carrying out semantic analysis based on a knowledge graph on the voice recognition result and text information in the picture, and analyzing a teaching scene, wherein the teaching scene comprises a teaching knowledge point range, a teaching video knowledge keyword and a video type, and the video type comprises a test question explanation type and a knowledge point explanation type; establishing semantic relation based on knowledge graph for test question attribute and test question text;
the method comprises the steps of marking a video tag, wherein the tag comprises a knowledge point range of teaching, a video type, teacher information, knowledge keywords and video positions thereof, a LaTex formula, a formula structure type and a video semantic relation based on a knowledge map, and when the video type is a test question explanation type, the tag further comprises a test question stem, an answer and analysis.
3. The intelligent video pushing method according to claim 1, wherein the learning data of the same class students is used for accurately following the teaching rhythm, and the video pushing is performed by combining student learning situation portraits and matched classification strategies, and further, the link address of the video is associated with the test questions, and is pushed to students.
4. The intelligent video pushing method according to claim 2, further comprising: and after receiving the link address of the video and the test questions, the students call a player to play the teaching video stored in the link address of the video.
5. The intelligent video pushing system is characterized by comprising test question collecting equipment, teaching video collecting equipment, storage equipment, student learning condition statistics module, a processor, a knowledge graph constructing module, a pushing model constructing module and a video pushing module constructing module,
the test question acquisition equipment is connected with the processor, and is used for acquiring and inputting test questions and sending the test questions to the processor;
the teaching video acquisition equipment is connected with the processor and used for acquiring teaching videos and sending the teaching videos to the processor;
the student learning condition statistics module is connected with the storage device, and analyzes the learning progress, the learning completion degree and the knowledge point grasping degree of the students from single departments and comprehensive ranking fluctuation through examination and homework practice of the students on the test questions in the test question library; analyzing the learning level of a certain subject of the student at a class thereof, the teaching level of the class of the student at the class thereof, the regional teaching level condition of the school of the student and the regional teaching level condition of the student, combining the knowledge point examination surface and the answering condition of the examination, the homework and the online exercise of the student, combining the knowledge point difficulty and the past examination condition accurate analysis of the province of the school to the learning breakthrough point suitable for the current progress of the student, obtaining the learning condition image of the student and feeding back to the storage device for storage;
the processor is respectively connected with the test question acquisition equipment, the teaching video acquisition equipment and the storage equipment, adds test question attribute labels to the test questions, and the test question attribute labels comprise: the method comprises the steps of determining core knowledge points and related knowledge points of a test question according to a test question attribute label to form a test question library, wherein the core knowledge points are knowledge points which are mainly inspected by the test question, the related knowledge points are knowledge points which are inspected in the test question and are related to the core knowledge points, establishing a test question image for the test question, establishing a text semantic network based on the knowledge map for a test question text on the basis of the test question attribute label, and sending the test question library and the test question image to a storage device for storage; respectively analyzing images and voices in the teaching video, marking video portrait labels on the teaching video, and sending the video portrait labels to a storage device for storage;
the knowledge graph constructing module is respectively connected with the storage equipment, the pushing model constructing module and the video pushing module and is used for entity linking, named entity identification based on BiLSTM+CRF algorithm is carried out on voice and test question text in video stored in the storage equipment, entity extraction is carried out, the extracted entities link the same entity information on different sub-knowledge graphs, and entity linking is realized by using a CoLink unsupervised learning framework; extracting knowledge graph characteristics: carrying out knowledge graph feature extraction on the basis of a tranD algorithm of knowledge graph feature learning, and accurately describing the entity by virtue of contextual entity features of the entity;
the building pushing model and the video pushing module are connected with the building knowledge graph module, consistency or relevance among students, test questions and teaching videos is built based on the knowledge graph, the knowledge graph is utilized to find the relevance paths among different types of entities, and the relevance of the entity nodes is calculated based on an iterative weight propagation algorithm; performing characteristic semantic association on students, test questions and teaching videos; potential association of students, questions and teaching videos is found out based on a Manifold algorithm, and semantic association among different labels is found out by combining DN-DBpedia corpus with an ESA model; building a test question and video pushing algorithm based on a convolutional neural network and an attention mechanism; the method comprises the steps of dividing areas, teaching material versions, schools and grade schedules, combining teaching quality, student learning situation images, knowledge point difficulty and knowledge point examination frequency, combining knowledge point map comprehensive matching, matching a classification strategy, and formulating a video with optimal classification rate and classification difficulty for pushing; through the study data of the classmates, the teaching rhythm is accurately followed, the student learning situation portrait and the matched classification strategy are combined, video pushing is carried out, and the link address and the test question of the video are pushed to the student at the same time.
6. The intelligent video pushing system according to claim 5, wherein the test question collection device is a scanner or a camera, and the teaching video collection device is a camera.
7. The video intelligent push system of claim 5, wherein the storage device is cloud storage or physical storage.
8. The intelligent video pushing system of claim 5, wherein the processor analyzes the images and voices in the teaching video and marks the teaching video with video portrait labels, further,
performing voice recognition on the audio part of the teaching video, recording a voice recognition result, performing knowledge extraction on the voice recognition result based on a knowledge graph to obtain a teaching video knowledge keyword, and recording a corresponding video playing position;
image recognition is carried out on the teaching video image part, face recognition, optical character recognition and formula recognition are respectively carried out, and teacher information in the video is recognized through the face recognition; knowledge extraction based on knowledge graph is carried out through optical character recognition, and knowledge points and knowledge keywords in the video image are extracted; for the test question explanation type video, identifying text information comprising test question stems, answers and analysis by using optical characters; the formula identifies a LaTex formula and a formula structure type in the video image;
carrying out semantic analysis based on a knowledge graph on the voice recognition result and text information in the picture, and analyzing a teaching scene, wherein the teaching scene comprises a teaching knowledge point range, a teaching video knowledge keyword and a video type, and the video type comprises a test question explanation type and a knowledge point explanation type; establishing semantic relation based on knowledge graph for test question attribute and test question text;
the method comprises the steps of marking a video tag, wherein the tag comprises a knowledge point range of teaching, a video type, teacher information, knowledge keywords and video positions thereof, a LaTex formula, a formula structure type and a video semantic relation based on a knowledge map, and when the video type is a test question explanation type, the tag further comprises a test question stem, an answer and analysis.
9. The intelligent video pushing system according to claim 5, further comprising a player, wherein after receiving the link address and the test question of the video, the student calls the player to play the teaching video stored in the link address of the video.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201811465037.6A CN109359215B (en) | 2018-12-03 | 2018-12-03 | Video intelligent pushing method and system |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201811465037.6A CN109359215B (en) | 2018-12-03 | 2018-12-03 | Video intelligent pushing method and system |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN109359215A CN109359215A (en) | 2019-02-19 |
| CN109359215B true CN109359215B (en) | 2023-08-22 |
Family
ID=65330830
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201811465037.6A Active CN109359215B (en) | 2018-12-03 | 2018-12-03 | Video intelligent pushing method and system |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN109359215B (en) |
Families Citing this family (26)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110196939A (en) * | 2019-04-26 | 2019-09-03 | 上海理工大学 | Text emphasis tagging system and method based on examination question examination point |
| CN110275987B (en) * | 2019-05-09 | 2021-10-19 | 上海平安智慧教育科技有限公司 | Intelligent teaching consultant generation method, system, equipment and storage medium |
| CN110021213A (en) * | 2019-05-14 | 2019-07-16 | 上海乂学教育科技有限公司 | Mathematics preamble learning method in artificial intelligence study |
| CN111061856B (en) * | 2019-06-06 | 2022-05-27 | 北京理工大学 | Knowledge perception-based news recommendation method |
| CN110489648A (en) * | 2019-08-15 | 2019-11-22 | 上海乂学教育科技有限公司 | Education resource dynamic pushing method and system |
| CN111738041A (en) * | 2019-09-30 | 2020-10-02 | 北京沃东天骏信息技术有限公司 | Video segmentation method, device, equipment and medium |
| 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 |
| CN111225235B (en) * | 2020-01-16 | 2020-12-04 | 北京合众美华教育投资有限公司 | Method for playing network teaching video |
| CN111597305B (en) * | 2020-05-15 | 2020-12-22 | 法政国际教育投资有限公司 | Entity marking method, entity marking device, computer equipment and storage medium |
| CN111611434B (en) * | 2020-05-19 | 2023-07-25 | 深圳康佳电子科技有限公司 | Online course interaction method and interaction platform |
| CN112950421A (en) * | 2020-05-30 | 2021-06-11 | 苏州才豪电子科技有限公司 | Online education management method based on block chain |
| CN111815274A (en) * | 2020-07-03 | 2020-10-23 | 北京字节跳动网络技术有限公司 | Information processing method and device and electronic equipment |
| CN111798138A (en) * | 2020-07-07 | 2020-10-20 | 广州视源电子科技股份有限公司 | Data processing method, computer storage medium and related equipment |
| CN111783709B (en) * | 2020-07-09 | 2022-09-06 | 中国科学技术大学 | Information prediction method and device for education video |
| CN112231522B (en) * | 2020-09-24 | 2021-09-14 | 北京奥鹏远程教育中心有限公司 | Online course knowledge tree generation association method |
| CN112163118A (en) * | 2020-09-28 | 2021-01-01 | 北京爱论答科技有限公司 | Recommended video acquisition method and device for teaching |
| CN112163027B (en) * | 2020-09-30 | 2021-10-22 | 华院计算技术(上海)股份有限公司 | Django and LaTeX-based mathematical test question and paper combining system and method |
| CN112883226B (en) * | 2021-03-29 | 2023-05-26 | 读书郎教育科技有限公司 | Intelligent classroom learning video recommendation method |
| CN113434036B (en) * | 2021-05-26 | 2024-02-20 | 上海慕知文化传媒有限公司 | AR augmented reality teaching material auxiliary system |
| CN113505238A (en) * | 2021-07-09 | 2021-10-15 | 北京小谛机器人科技有限公司 | Semantic map construction method and device |
| CN113761030A (en) * | 2021-09-07 | 2021-12-07 | 辽宁铂阳科技集团有限公司 | Wrong question statistical analysis system |
| CN113742496B (en) * | 2021-09-10 | 2024-05-28 | 国网江苏省电力有限公司电力科学研究院 | Electric power knowledge learning system and method based on heterogeneous resource fusion |
| CN114241835A (en) * | 2021-11-17 | 2022-03-25 | 北京执象科技发展有限公司 | Student spoken language quality evaluation method and device |
| CN116501892B (en) * | 2023-05-06 | 2024-03-29 | 广州番禺职业技术学院 | Training knowledge graph construction method based on automatic following system of Internet of things |
| CN116823557B (en) * | 2023-08-30 | 2023-11-17 | 南通四建集团有限公司 | Interactive training management method and system for safety education of construction workers |
| CN116955589B (en) * | 2023-09-19 | 2024-01-30 | 山东山大鸥玛软件股份有限公司 | Intelligent proposition method, system, proposition terminal and storage medium based on teaching material knowledge graph |
Citations (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2013091434A1 (en) * | 2011-12-23 | 2013-06-27 | Liu Zhuochen | Intelligent interaction system for video teaching |
| GB201501437D0 (en) * | 2015-01-28 | 2015-03-11 | Canon Kk | Client-driven push of resources by a server device |
| WO2016112679A1 (en) * | 2015-01-14 | 2016-07-21 | 百度在线网络技术（北京）有限公司 | Method, system and storage medium for realizing intelligent answering of questions |
| CN106951439A (en) * | 2017-02-13 | 2017-07-14 | 广东小天才科技有限公司 | The examination question method for pushing and system of a kind of associated video |
| CN108596472A (en) * | 2018-04-20 | 2018-09-28 | 贵州金符育才教育科技有限公司 | A kind of the artificial intelligence tutoring system and method for natural sciences study |
| CN108664649A (en) * | 2018-05-17 | 2018-10-16 | 深圳习习网络科技有限公司 | Knowledge content method for pushing, device and push server |
| CN108900366A (en) * | 2018-09-26 | 2018-11-27 | 江苏曲速教育科技有限公司 | Uniform data central management system and its management method |
Family Cites Families (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| TWI534768B (en) * | 2013-12-17 | 2016-05-21 | Jian-Cheng Liu | Wisdom teaching counseling test method |
-
2018
- 2018-12-03 CN CN201811465037.6A patent/CN109359215B/en active Active
Patent Citations (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2013091434A1 (en) * | 2011-12-23 | 2013-06-27 | Liu Zhuochen | Intelligent interaction system for video teaching |
| WO2016112679A1 (en) * | 2015-01-14 | 2016-07-21 | 百度在线网络技术（北京）有限公司 | Method, system and storage medium for realizing intelligent answering of questions |
| GB201501437D0 (en) * | 2015-01-28 | 2015-03-11 | Canon Kk | Client-driven push of resources by a server device |
| CN106951439A (en) * | 2017-02-13 | 2017-07-14 | 广东小天才科技有限公司 | The examination question method for pushing and system of a kind of associated video |
| CN108596472A (en) * | 2018-04-20 | 2018-09-28 | 贵州金符育才教育科技有限公司 | A kind of the artificial intelligence tutoring system and method for natural sciences study |
| CN108664649A (en) * | 2018-05-17 | 2018-10-16 | 深圳习习网络科技有限公司 | Knowledge content method for pushing, device and push server |
| CN108900366A (en) * | 2018-09-26 | 2018-11-27 | 江苏曲速教育科技有限公司 | Uniform data central management system and its management method |
Non-Patent Citations (1)
| Title |
|---|
| 基于知识图谱的学科单选题考点提取研究;刘一然;骆力明;;计算机应用研究(06);全文 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN109359215A (en) | 2019-02-19 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN109359215B (en) | Video intelligent pushing method and system | |
| CN111833672B (en) | Teaching video display method, device and system | |
| US20120107789A1 (en) | Learning environment with user defined content | |
| CN112163162B (en) | Method for recommending selected courses based on portrait identification, storage medium and electronic equipment | |
| CN109710931A (en) | A kind of class teaching content evaluation system and its method based on Intellectual Analysis Technology | |
| CN113609103A (en) | AI-based interactive reading support database construction method and system | |
| CN111046194A (en) | Method for constructing multi-mode teaching knowledge graph | |
| CN110753256A (en) | Video playback method and device, storage medium and computer equipment | |
| CN113254708A (en) | Video searching method and device, computer equipment and storage medium | |
| CN112132536A (en) | Post recommendation method, system, computer equipment and storage medium | |
| CN112800177B (en) | FAQ knowledge base automatic generation method and device based on complex data types | |
| CN116226410B (en) | Teaching evaluation and feedback method and system for knowledge element connection learner state | |
| CN116228361A (en) | Course recommendation method, device, equipment and storage medium based on feature matching | |
| CN116595188A (en) | Educational knowledge graph system based on artificial intelligence and big data | |
| CN110648262A (en) | Teaching system for mathematical thinking training and data processing method | |
| CN111813919B (en) | MOOC course evaluation method based on syntactic analysis and keyword detection | |
| CN114691904A (en) | Course resource marking method, device, system, equipment and storage medium | |
| KR102126834B1 (en) | Automatic scoring system using qr code | |
| CN112115275A (en) | Knowledge graph construction method and system for math tutoring question-answering system | |
| CN112347354B (en) | Courseware management method for intelligent education | |
| CN117455126B (en) | Ubiquitous practical training teaching and evaluation management system and method | |
| CN112395403B (en) | Knowledge graph-based question and answer method, system, electronic equipment and medium | |
| CN113239234B (en) | Method for providing video book and method for establishing video book | |
| JP5538511B2 (en) | Learning support system, learning support server, learning support method, and learning support program | |
| CN117033665B (en) | Method and device for aligning map knowledge points with video |
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 |