Search

CN-115878840-B - Teacher side online teaching resource recommendation method

CN115878840BCN 115878840 BCN115878840 BCN 115878840BCN-115878840-B

Abstract

The application relates to a teacher side online teaching resource recommendation method which comprises the following steps of extracting video, courses and teacher information, storing the extracted information into a search engine, acquiring information of a current teacher or a course, searching whether the current teacher or the course exists or not, if so, acquiring reference videos of the similar teacher or the course, analyzing and assembling a general video reference chain, judging which section is currently in the reference chain according to the current behavior of the teacher, and recommending a plurality of related videos by taking the section as a base point. According to the teacher side online teaching resource recommendation method, when the teacher and the current course information are acquired, the courses or videos of a plurality of teachers with highest course relevance with the current teacher are recommended according to the searched teacher or video information.

Inventors

  • WANG HUI
  • CHEN FANGYI

Assignees

  • 上海卓越睿新数码科技股份有限公司

Dates

Publication Date
20260512
Application Date
20221024

Claims (5)

  1. 1. The teacher-side online teaching resource recommendation method is characterized by comprising the following steps of: extracting key information from the video, the courses and teacher information, and storing the extracted key information into a search engine; Acquiring information of a current teacher or course, and searching whether the current teacher or course exists or not, if so, then Acquiring reference videos of similar teachers or courses, analyzing and summarizing the reference videos into a video reference chain, judging which section of the reference chain is currently positioned according to the current behaviors of the teacher, recommending a plurality of related videos by taking the section as a base point, and if not, judging whether the section is positioned in the reference chain or not according to the current behaviors of the teacher Acquiring current behavior information of a teacher, extracting key knowledge point information from the current behavior information, and recommending a plurality of related videos according to the key knowledge point information; score sorting is carried out on a plurality of videos obtained through recommendation, and a final preset number of videos are selected; the key information extraction of the video, the course and the teacher information comprises the following steps: carrying out vectorization processing and key information extraction on structured and unstructured information of a video; extracting the names of courses, give a course schools, uploaded data, reference videos and historical retrieval information; Extracting history teaching information, uploading data, history watching and quoting video information of a teacher; the vectorization processing and key information extraction of the structured and unstructured information of the video comprise the following steps: selecting partial fields in the video structural information as video key information; performing OCR processing on the video, removing repeated frames and contained frames of the video, and giving time length to each frame according to forward rules; The selecting part of fields in the video structural information as video key information further comprises: embedding the selected key information into the video vector representation and storing the key information into a search engine; the method comprises the steps of performing OCR processing on video, removing repeated frames of the video, assigning time length to each frame according to a forward rule, and further comprising: Extracting keywords from the text after the duplication removal, judging the weight of the keywords according to the duty 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; The method for carrying out vectorization processing and key information extraction on the structured and unstructured information of the video further comprises the following steps: and carrying out 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 weight into video vector representation.
  2. 2. The method for recommending online teaching resources on a teacher side according to claim 1, wherein the steps of acquiring information of a current teacher or course and retrieving whether there are similar teachers or courses include: assigning values to each key information score point of a teacher or a course; judging whether similar teacher or course exists according to the teacher video behavior and the key knowledge point information, if yes, then And obtaining similar teachers or courses and summarizing the video reference chains.
  3. 3. The teacher-side online teaching resource recommendation method according to claim 2, characterized by recommending a plurality of related videos based on videos of the similar teacher or course, comprising: Judging which section of the reference chain the video of the current teacher or course is in; And recommending the previous section and the next two sections of videos by taking the node of the current teacher or course video in the reference chain as a base point.
  4. 4. The teacher side online teaching resource recommendation method of claim 1, wherein, According to the current behavior information of the teacher, key knowledge point information is extracted from the current behavior information to search videos in a video library, and a plurality of videos with highest correlation degree are recommended, wherein the method comprises the following steps: Retrieving the video based on the semantics and keywords; and obtaining three videos with the highest scores according to the same points of the retrieved video and the key information of the current teacher video.
  5. 5. The teacher-side online teaching resource recommendation method of claim 4, further comprising: performing de-duplication on videos of different recommended paths; And (3) different weights are distributed to different paths to obtain weighted scores of one video on all paths, and three videos with highest scores are recommended.

Description

Teacher side online teaching resource recommendation method Technical Field The application relates to the technical field of online education, in particular to a teacher-side online teaching resource recommendation method. Background In a college teaching scene, teachers have not only teaching tasks but also a large number of scientific research tasks, so how to improve the teaching efficiency under the condition of guaranteeing the teaching quality is a classical problem in the college education scene. Although the teaching video can help the teacher save a great part of teaching time, the teaching video needs to be prepared in a great deal, so that the choice of a great deal of teacher is made by multiplexing some disclosed high-quality teaching videos in own courses. On some teaching video platforms, the number of teaching videos is excessive, and the time for a teacher to search and view the selected behaviors is long. Therefore, finding the teaching video required by the teacher from massive video resources becomes one of the bottlenecks for reducing the lesson preparation cost of the teacher. In the current video recommendation scene of online education, the aimed user object is generally a learner, and video recommendation is performed according to the learning interest of the learner. The learning interests of users may be divergent and disordered in general, the teaching targets and teaching systems of teachers are specific, and the processes are generally strictly ordered, for example, students are difficult to learn differentiation without learning derivatives, so that the existing recommendation tasks and schemes cannot be suitable for video recommendation in the teacher's lesson preparation scene. Disclosure of Invention In view of the foregoing, it is desirable to provide a teacher-side online teaching resource recommendation method that can quickly recommend videos required by teachers in a complex video library. A teacher-side online teaching resource recommendation method, the method comprising: extracting key information from the video, the courses and teacher information, and storing the extracted key information into a search engine; Acquiring information of a current teacher or course, and searching whether the current teacher or course exists or not, if so, then Acquiring reference videos of similar teachers or courses, analyzing and summarizing the reference videos into a video reference chain, judging which section of the reference chain is currently positioned according to the current behaviors of the teacher, recommending a plurality of related videos by taking the section as a base point, and if not, judging whether the section is positioned in the reference chain or not according to the current behaviors of the teacher Acquiring current behavior information of a teacher, extracting information such as key knowledge points and the like from the current behavior information, and recommending a plurality of related videos according to the information; and sorting the scores of the plurality of videos obtained by recommendation, and selecting the final preset number of videos. In one embodiment, the extracting key information of the video, the course and the teacher information includes: carrying out vectorization processing and key information extraction on structured and unstructured information of a video; extracting the names of courses, give a course schools, uploaded data, reference videos and historical retrieval information; And extracting the history teaching information, the uploaded data, the history watching and quoting video information of the teacher. In one embodiment, the vectorizing and key information extracting of the structured and unstructured information of the video includes: selecting partial fields in the video structural information as video key information; OCR processing is performed on the video, repeated frames and contained frames of the video are removed, and time duration is given to each frame according to a forward rule. In one embodiment, the selecting a part of fields in the video structural information as video key information further includes: the selected key information is embedded into the video vector representation and stored in a search engine. In one embodiment, the OCR processing is performed on the video, removing repeated frames of the video, and assigning a duration to each frame according to a forward rule, and then further includes: extracting keywords from the text after the duplication removal, judging the weight of the keywords according to the duty 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. In one embodiment, the vectorizing processing and the key information extracting are performed on the structured and unstructured information of the video, and the method further includes: and carrying out ASR processing on the