CN115757937A - Student side online teaching resource recommendation method - Google Patents
Student side online teaching resource recommendation method Download PDFInfo
- Publication number
- CN115757937A CN115757937A CN202211301610.6A CN202211301610A CN115757937A CN 115757937 A CN115757937 A CN 115757937A CN 202211301610 A CN202211301610 A CN 202211301610A CN 115757937 A CN115757937 A CN 115757937A
- Authority
- CN
- China
- Prior art keywords
- video
- information
- videos
- student
- keywords
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
Images
Landscapes
- Information Retrieval, Db Structures And Fs Structures Therefor (AREA)
Abstract
The application relates to a student side online teaching resource recommendation method, which comprises the following steps: extracting key information of the video and the student information, and respectively storing the key information into a retrieval engine and serving as input of a neural network recommendation model; judging whether a recommendation mechanism is triggered in the online learning process, if so, acquiring keywords and relevant information of the current trigger scene, and acquiring videos in a video library according to the keywords and the relevant information to obtain a plurality of videos with the highest relevance. According to the method for recommending the student-side online teaching resources, when a recommendation mechanism is triggered by a student in the learning process, the search engine can search according to the keywords and the related information of the learning scene and recommend videos according to the relevancy of the searched videos and the keywords and the related information of the learning scene, and the neural network recommendation model can recommend the videos according to the input student information.
Description
Technical Field
The application relates to the technical field of online education, in particular to a student-side online teaching resource recommendation method.
Background
In an online teaching scene, the course teaching video published by the teacher is generally oriented to all students, but the learning abilities of different students may be different, which results in different conditions of harvest and knowledge mastering of each student in the same course.
Meanwhile, massive video resources exist on the online teaching platform, and each course can be similar to the same type of course. In a class of courses, the teaching systems of different teachers are different, but generally different emphasis points exist, for example, some teachers explain a certain knowledge point more coarsely, and another teacher explains the knowledge point more finely.
When the students search the relevant learning videos, the students are very troublesome, not only because the searching action takes a long time, but also a certain time is needed to watch the searching result to judge whether the videos are useful.
Some existing recommendation methods mostly recommend according to video names and video playing amounts, but the recommendation accuracy is low, the matching degree with the self condition of the student is low, and the teaching recommendation which best meets the self condition of the student cannot be provided. For example, some teachers explain the knowledge point, but the explanation is still more concise than the teaching video issued by the teachers, so that the students cannot learn much help, and the condition is equal to recommendation failure.
Disclosure of Invention
Therefore, a student-side online teaching resource recommendation method with high accuracy needs to be provided for solving the technical problems.
The student side online teaching resource recommendation method comprises the following steps:
extracting key information of the video and the student information, and respectively storing the key information into a retrieval engine and serving as input of a neural network recommendation model;
judging whether a recommendation mechanism is triggered in the online learning process, if so, determining whether the recommendation mechanism is triggered in the online learning process
And acquiring keywords and related information of the current trigger scene, and acquiring videos in the video library according to the keywords and the related information to obtain a plurality of videos with highest correlation.
In one embodiment, the extracting key information from the video and the student information includes:
vectorizing processing and key information extraction are carried out on the structured and unstructured information of the video;
and extracting the information of the school where the student is and the historical behavior information, and inputting the information into a neural network recommendation model.
In one embodiment, the vectorizing processing and key information extraction on the structured and unstructured information of the video includes:
selecting partial fields in the video structured information as video key information;
and performing OCR processing on the video, removing repeated frames and contained frames of the video, and endowing time length for each frame according to a forward rule.
In one embodiment, the selecting a part of fields in the video structured information as video key information further includes:
and embedding the selected key information into the video vector representation and storing the key information into a retrieval engine.
In one embodiment, the OCR processing on the video, removing the video duplicate frames, and assigning a time length to each frame according to the forward rule, then further includes:
and extracting keywords from the text after the duplication removal, judging the weight of the keywords according to the ratio of the duration of the frame in which the keywords are located to the duration of the video, and embedding the weight into the video vector representation.
In one embodiment, the vectorizing processing and key information extraction on the structured and unstructured information of the video further includes:
and performing ASR processing on the video, extracting keywords of the text, judging the weight of the keywords according to the occurrence frequency of the keywords, and embedding the keywords into video vector representation.
In one embodiment, the determining whether to trigger a recommendation mechanism in the online learning process includes:
whether pause or repeated playing occurs in the video playing process;
whether wrong questions or low-degree questions appear in the online answering process.
In one embodiment, the acquiring the keywords and the related information of the current trigger scenario includes:
acquiring keywords and keyword duration information of a current video frame;
and acquiring keywords contained in the error questions and the bound knowledge point information.
In one embodiment, the obtaining the videos in the video library to obtain a plurality of videos with the highest correlation includes:
assigning scores of all key words of the video or the wrong questions;
acquiring ten videos with the highest scores according to the video obtained by retrieval and the current video or the same point of the keywords of wrong questions;
removing the duplicate of the videos of different recommended paths;
and distributing different weights to different paths to obtain the weighted score of one video on all paths, and recommending the three videos with the highest scores.
In one embodiment, the method further comprises:
whether the set trigger time is reached, if so, then
And inputting the current learning behaviors and school information of the students into the neural network recommendation model to obtain three result videos.
According to the method for recommending the student-side online teaching resources, the videos and the student key information are extracted and stored in the retrieval engine and the input neural network recommendation model respectively, when a student triggers a recommendation mechanism in the learning process, the retrieval engine can search according to the keywords and the related information of a learning scene and recommend the videos according to the relevancy of the searched videos and the keywords and the related information of the learning scene, and the neural network recommendation model can recommend the videos according to the input student information.
Drawings
FIG. 1 is a diagram of steps of a student-side online teaching resource recommendation method according to an embodiment of the present application;
FIG. 2 is a diagram of the steps of a student-side online teaching resource recommendation method according to another embodiment of the present application;
fig. 3 is a schematic diagram of an architecture of a student-side online teaching resource recommendation method according to an embodiment of the application.
Detailed Description
In order to make the objects, technical solutions and advantages of the embodiments of the present application clearer, the technical solutions in the embodiments of the present application will be clearly and completely described below with reference to the drawings in the embodiments of the present application, and it is obvious that the described embodiments are some embodiments of the present application, but not all embodiments. All other embodiments, which can be derived by a person skilled in the art from the embodiments given herein without making any creative effort, shall fall within the protection scope of the present application.
It will be understood that when an element is referred to as being "secured to" or "disposed" on another element, it can be directly on the other element or intervening elements may also be present. When a component is referred to as being "connected" to another component, it can be directly connected to the other component or intervening components may also be present. The terms "vertical," "horizontal," "upper," "lower," "left," "right," and the like as used in the description of the present application are for illustrative purposes only and do not represent the only embodiments.
Furthermore, the terms "first", "second" and "first" are used for descriptive purposes only and are not to be construed as indicating or implying relative importance or implicitly indicating the number of technical features indicated. Thus, a feature defined as "first" or "second" may explicitly or implicitly include at least one such feature. In the description of the present application, "plurality" means at least two, e.g., two, three, etc., unless specifically limited otherwise.
In this application, unless expressly stated or limited otherwise, the first feature "on" or "under" the second feature may mean that the first feature is in direct contact with the second feature, or that the first feature and the second feature are in indirect contact via an intermediate. Also, a first feature "on," "over," and "above" a second feature may be directly or diagonally above the second feature, or may simply indicate that the first feature is at a higher level than the second feature. A first feature being "under," "below," and "beneath" a second feature may be directly under or obliquely under the first feature, or may simply mean that the first feature is at a lesser elevation than the second feature.
Unless defined otherwise, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this application belongs. The terminology used in the description of the present application is for the purpose of describing particular embodiments only and is not intended to be limiting of the application. As used in the description of the present application, the term "and/or" includes any and all combinations of one or more of the associated listed items.
As shown in fig. 1, in one embodiment, a method for recommending student-side online education resources comprises the following steps:
and step S110, extracting key information of the video and the student information, and respectively storing the key information into a retrieval engine and an input neural network recommendation model.
Specifically, extracting relevant information of videos and students, wherein the relevant information comprises structured information and unstructured information of the videos; information of school where students are located and historical behavior information; the historical behavior information of the students comprises video watching records and historical searching records of the students; the information is respectively stored in a retrieval engine and an input neural network recommendation model, so that a recommendation video can be conveniently obtained subsequently.
Step S120, judging whether a recommendation mechanism is triggered in the online learning process, if so, executing
Specifically, the recommendation mechanism can be set according to actual conditions, such as when an online test generates a wrong question.
Step S130, keywords and relevant information of the current trigger scene are obtained, and accordingly videos in the video library are retrieved, and a plurality of videos with the highest relevance are recommended.
Specifically, videos having the same points with the current scene keywords and the related information are searched, the degree of correlation between the videos and the current scene keywords is judged according to the searched videos, and finally a plurality of videos with the highest degree of correlation are selected.
According to the method for recommending the student-side online teaching resources, the videos and the student key information are extracted and stored in the retrieval engine and the input neural network recommendation model respectively, when a student triggers a recommendation mechanism in the learning process, the retrieval engine can search according to the keywords and the related information of the learning scene and recommend the videos according to the relevancy of the searched videos and the keywords and the related information of the learning scene, and the neural network recommendation model can recommend the videos according to the input student information.
As shown in fig. 2, in one embodiment, a method for recommending student-side online education resources comprises the following steps:
step S210, selecting a part of fields in the video structured information as video key information.
Specifically, the video structured information is a relatively precise field which is manually filled when a video is published to an online teaching platform, and includes a video name and original course information of the video initial publication, where the original course information of the video includes teacher information and school information of a beginning course, and chapter directory information of the video in the original course.
Step S220, the selected key information is embedded into the video vector representation and stored in a retrieval engine.
Specifically, after certain analysis is performed on the structured information, part of the fields are selected as key features of the video, and after normalization and mapping operations are assisted, the fields are embedded into corresponding positions of the video.
Step S230, performing OCR (Optical Character Recognition) processing on the video, removing repeated frames and included frames of the video, and assigning a duration to each frame according to a forward rule.
In particular, in video information processing, a large part of work is to extract key information from text in the video information processing. For example, for OCR texts, research has found that if a teacher explains a page of PPT in a teaching video, the time for the page of PPT to be completely displayed generally lasts for about 5s (by completely displaying, the content of the page of PPT is completely displayed on the page, and there is no new content added by a subsequent animation effect). Therefore, we choose to frame the video every 5 s.
In the process of acquiring keywords of the key frames, it may happen that a certain page of PPT exists in multiple frames or a certain page of PPT exists in multiple frames when being gradually and completely displayed, and therefore, when obtaining an OCR text, certain processing needs to be performed on a text result.
Judging two frames of pictures, and removing the previous frame of picture if the two frames of pictures are completely the same; if the two frames are the same, loading different frame parts into the next frame and removing the previous frame. After the key frames after deduplication are obtained, each key frame is given a duration in a forward manner according to the key frame interval duration. For example, if the previous frame is removed, the time of the next frame minus the time of the previous frame is the time duration of the next frame.
Step S240, extracting keywords from the text after the de-weighting, judging the weight of the keywords according to the ratio of the duration of the frame where the keywords are located to the duration of the video, and embedding the weight into the video vector representation.
Specifically, the weight of each keyword of the key frame in the video is judged based on the effective duration of the video and the duration of each key frame. The calculation formula of the weight of each keyword in the key frame in the video is as follows:
the effective duration of the video is the time of the last key frame minus the duration of the head and the end of the film.
Step S250, ASR (Automatic Speech Recognition) processing is carried out on the video, keywords of the text are extracted, the weight of the keywords is judged according to the frequency of the keywords, and the keywords are embedded into video vector representation.
Specifically, for the ASR result text, keyword extraction is directly performed on all extracted sentences. And carrying out weight assignment according to the extracted information such as the occurrence frequency of the keywords in the sentences.
It should be noted that after the processed information is obtained, the processed information is embedded into a corresponding position of the video, and the vector identification of the video is more reasonable.
And step S260, extracting the information of the school where the student is and the historical behavior information, and inputting the information into a neural network recommendation model.
Specifically, the historical behavior information of the student includes video watching records and historical search records of the student, and is used for inputting a neural network recommendation model so that the neural network recommendation model recommends related videos.
Step S270, whether pause or repeated playing occurs in the video playing process or not, if yes, execution is carried out
Specifically, when a student learns a certain video, the stop duration in a certain key frame exceeds a threshold value or certain key frame segments are repeatedly played, the student is judged to have difficulty in understanding or not understand the pause or repeated playing part, and a recommendation mechanism is triggered.
Step S280, keywords and keyword duration information of the current video frame are obtained.
Step S290, whether wrong or low-grade question appears in the on-line answering process, if yes, executing
Specifically, after answering questions online, if the questions with low scores or wrong questions exist, the students are judged not to understand the wrong questions and the low scores, and then a recommendation mechanism is triggered.
Step S2100, obtaining keywords and bound knowledge point information contained in the error question.
Specifically, the keywords, the keyword duration information, the keywords included in the error question, and the bound knowledge point information of the video frame are all words capable of representing the video or the error question, and are specifically set according to actual conditions.
Step S2110, the scores of the keywords of the video or the wrong questions are assigned.
Specifically, the scores of the keywords are set according to the specific relevance.
And S2120, acquiring ten videos with the largest score according to the same point of the searched videos and the current video or the keywords of the wrong questions.
Specifically, when the retrieved video is the same as some keywords of the current video or the wrong topic, the score of the retrieved video is the sum of the scores of the keywords, and meanwhile, whether the sum score is larger than a set threshold value is judged, so that whether the retrieved video can be used as the related video is judged.
Step S2130, whether the set triggering time is reached is judged, if yes, execution is carried out
Specifically, the neural network recommendation model recommends the relevant videos according to the current learning behaviors and learning contents of the students.
Step S2140, the current learning behavior and school information of the students are input into a neural network recommendation model, and three result videos are obtained.
Specifically, after the neural network recommendation model obtains the input information, relevant recommendation is performed on the video in the video library according to the input information, so that a recommended video is obtained.
In step S2150, the video of different recommended paths is deduplicated.
Specifically, if the recommended video of each path has duplicates, the duplicated video is removed.
Step S2160, different weights are assigned to different paths to obtain the weighted scores of one video on all paths, and then three videos with the highest scores are recommended.
The architecture schematic diagram of the student-side online teaching resource recommendation method is shown in fig. 3, videos and student key information are extracted and stored in a retrieval engine and an input neural network recommendation model respectively, when a student triggers a recommendation mechanism in the learning process, namely when playing videos are stopped or repeatedly played, when answers are low in score or wrong, the retrieval engine can retrieve related videos in a video library and recommend a specified number of videos, when set triggering time is reached, the neural network recommendation model can recommend the videos according to received student information, then weighting and assigning values to all recommended videos, and finally three videos with the highest scores are selected for recommendation.
It should be noted that the acquiring of the current behavior sequence of the student in fig. 3 includes a video viewing record of the student and a historical search record.
The technical features of the embodiments described above may be arbitrarily combined, and for the sake of brevity, all possible combinations of the technical features in the embodiments described above are not described, but should be considered as being within the scope of the present specification as long as there is no contradiction between the combinations of the technical features.
The above-mentioned embodiments only express several embodiments of the present application, and the description thereof is more specific and detailed, but not construed as limiting the scope of the present application. It should be noted that, for a person skilled in the art, several variations and modifications can be made without departing from the concept of the present application, and these are all within the scope of protection of the present application. Therefore, the protection scope of the present patent shall be subject to the appended claims.
Claims (10)
1. The student side online teaching resource recommendation method is characterized by comprising the following steps:
extracting key information of the video and the student information, and respectively storing the key information into a retrieval engine and serving as input of a neural network recommendation model;
judging whether a recommendation mechanism is triggered in the online learning process, if so, determining whether the recommendation mechanism is triggered in the online learning process
And acquiring keywords and related information of the current trigger scene, and acquiring videos in the video library according to the keywords and the related information to obtain a plurality of videos with highest correlation.
2. The method for recommending student-side online teaching videos as claimed in claim 1, wherein said extracting key information from videos and student information comprises:
vectorization processing and key information extraction are carried out on the structured and unstructured information of the video;
and extracting the information of the school where the student is and the historical behavior information to be used as the input of the neural network recommendation model.
3. The method for recommending videos for student-side online teaching according to claim 2, wherein said vectorizing process and key information extraction of structured and unstructured information of videos comprises:
selecting partial fields in the video structured information as video key information;
and performing OCR processing on the video, removing repeated frames and contained frames of the video, and endowing time length for each frame according to a forward rule.
4. The method for recommending student-side online teaching videos as claimed in claim 3, wherein said selecting partial fields in the video structured information as video key information further comprises:
and embedding the selected key information into the video vector representation and storing the key information into a retrieval engine.
5. The method as claimed in claim 4, wherein the video recommendation method for student side online teaching comprises OCR processing the video, removing repeated frames of the video, and assigning a time length to each frame according to a forward rule, and then further comprising:
and extracting keywords from the text after the duplication is removed, judging the weight of the keywords according to the ratio of the duration of the frame where the keywords are located to the duration of the video, and embedding the weight into the video vector representation.
6. The video recommendation method for student-side online teaching according to claim 5, wherein said vectorizing and key information extraction are performed on structured and unstructured information of a video, further comprising:
and performing ASR processing on the video, extracting keywords of the text, judging the weight of the keywords according to the occurrence frequency of the keywords, and embedding the keywords into video vector representation.
7. The method for recommending student-side online teaching videos according to claim 6, wherein said determining whether to trigger a recommendation mechanism in the online learning process comprises:
whether pause or repeated playing occurs in the video playing process;
whether wrong questions or low-degree questions appear in the online answering process.
8. The method for recommending student-side online teaching videos according to claim 7, wherein the acquiring keywords of the current trigger scenario and related information thereof comprises:
acquiring keywords and keyword duration information of a current video frame;
and acquiring keywords contained in the error questions and the bound knowledge point information.
9. The method for recommending student-side online teaching videos according to claim 8, wherein said obtaining videos in a video library to obtain a plurality of videos with highest correlation comprises:
assigning scores of all key words of the video or the wrong questions;
acquiring ten videos with the highest scores according to the video obtained by retrieval and the current video or the same point of the keywords of wrong questions;
removing the duplicate of the videos of different recommended paths;
and distributing different weights to different paths to obtain the weighted score of one video on all paths, and recommending the three videos with the highest scores.
10. The student-side online teaching video recommendation method according to claim 9, further comprising:
whether the set trigger time is reached, if so, then
And inputting the current learning behaviors and school information of the students into the neural network recommendation model to obtain three result videos.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202211301610.6A CN115757937A (en) | 2022-10-24 | 2022-10-24 | Student side online teaching resource recommendation method |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202211301610.6A CN115757937A (en) | 2022-10-24 | 2022-10-24 | Student side online teaching resource recommendation method |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN115757937A true CN115757937A (en) | 2023-03-07 |
Family
ID=85352934
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202211301610.6A Pending CN115757937A (en) | 2022-10-24 | 2022-10-24 | Student side online teaching resource recommendation method |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN115757937A (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN116701707A (en) * | 2023-08-08 | 2023-09-05 | 成都市青羊大数据有限责任公司 | Educational big data management system |
-
2022
- 2022-10-24 CN CN202211301610.6A patent/CN115757937A/en active Pending
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN116701707A (en) * | 2023-08-08 | 2023-09-05 | 成都市青羊大数据有限责任公司 | Educational big data management system |
| CN116701707B (en) * | 2023-08-08 | 2023-11-10 | 成都市青羊大数据有限责任公司 | Educational big data management system |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN109359215B (en) | Video intelligent pushing method and system | |
| CN103425635B (en) | Method and apparatus are recommended in a kind of answer | |
| CN107463698B (en) | Method and device for pushing information based on artificial intelligence | |
| CN106126524B (en) | Information pushing method and device | |
| CN110362671B (en) | Topic recommendation method, device and storage medium | |
| CN108920450B (en) | Knowledge point reviewing method based on electronic equipment and electronic equipment | |
| CN108121715B (en) | Character labeling method and character labeling device | |
| US20150026184A1 (en) | Methods and systems for content management | |
| CN114332679A (en) | Video processing method, device, equipment, storage medium and computer program product | |
| CN109615009B (en) | Learning content recommendation method and electronic equipment | |
| CN115757937A (en) | Student side online teaching resource recommendation method | |
| CN113254708A (en) | Video searching method and device, computer equipment and storage medium | |
| CN109582780B (en) | Intelligent question and answer method and device based on user emotion | |
| CN111026786B (en) | Dictation list generation method and home education equipment | |
| CN113255843A (en) | Speech manuscript evaluation method and device | |
| CN112214579A (en) | Machine intelligent evaluation method and system for short answer questions | |
| CN111198957A (en) | Push method and device, electronic equipment and storage medium | |
| CN116228361A (en) | Course recommendation method, device, equipment and storage medium based on feature matching | |
| CN112464665B (en) | Subject term extraction and application method | |
| CN111813919B (en) | MOOC course evaluation method based on syntactic analysis and keyword detection | |
| CN113569112A (en) | Tutoring strategy providing method, system, device and medium based on question | |
| CN114139053A (en) | Lifelong learning resource intelligent pushing system based on credit bank and big data analysis | |
| CN115878840A (en) | Teacher-side online teaching resource recommendation method | |
| CN111666437A (en) | Image-text retrieval method and device based on local matching | |
| CN115878849B (en) | Video tag association method and device and electronic equipment |
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 |