CN-121996806-A - Agricultural courseware knowledge point segmentation and combination method, device, equipment and storage medium

CN121996806ACN 121996806 ACN121996806 ACN 121996806ACN-121996806-A

Abstract

The invention provides a method, a device, equipment and a storage medium for segmenting and combining knowledge points of agricultural courseware, which comprise the steps of obtaining an agricultural courseware video comprising a plurality of key frame images and explanation voices corresponding to the key frame images, performing voice recognition on the explanation voices to obtain an agricultural knowledge text corresponding to the key frame images, extracting key entities from the agricultural knowledge text corresponding to the key frame images, performing semantic association on the key entities and entity nodes in an agricultural domain knowledge graph to obtain semantic association results of the key frame images, dividing the agricultural courseware video into a plurality of agricultural knowledge point fragments based on the semantic association results of the key frame images, labeling knowledge point labels for the agricultural knowledge point fragments, and extracting knowledge point fragments to be used from the plurality of agricultural knowledge point fragments based on user use requirements to generate a new agricultural knowledge learning video. The intelligent segmentation, labeling and individuation on-demand combination of the agricultural courseware video knowledge points are realized.

Inventors

CHEN HUINA
MA XIAORUI
Bu Haiqiao
WANG MIN
WANG HONGBIAO

Assignees

北京市农林科学院

Dates

Publication Date: 20260508
Application Date: 20251226

Claims (10)

1. The method for dividing and combining the knowledge points of the agricultural courseware is characterized by comprising the following steps of: The method comprises the steps of acquiring an agricultural courseware video, wherein the agricultural courseware video comprises a plurality of key frame images and explanation voices corresponding to the key frame images; aiming at the explanation voice corresponding to each key frame image, carrying out voice recognition on the explanation voice to obtain an agricultural knowledge text corresponding to the key frame image; Extracting a key entity from the agricultural knowledge text corresponding to the key frame image, and carrying out semantic association on the key entity and entity nodes in the agricultural domain knowledge graph to obtain a semantic association result of the key frame image; Dividing the agricultural courseware video into a plurality of agricultural knowledge point segments based on semantic association results of each key frame image, and labeling knowledge point labels for the agricultural knowledge point segments; and extracting knowledge point fragments to be used from the plurality of agricultural knowledge point fragments based on the use requirement of a user, and generating a new agricultural knowledge learning video based on the combination of the knowledge point fragments to be used.
2. The method for segmenting and combining knowledge points of agricultural courseware according to claim 1, wherein the step of performing speech recognition on the explanation speech to obtain an agricultural knowledge text corresponding to the key frame image comprises the following steps: inputting the explanation voice into a pre-trained voice recognition model to obtain an initial knowledge text output by the voice recognition model; And carrying out text error correction on the initial knowledge text to obtain the agricultural knowledge text corresponding to the key frame image.
3. The method for segmenting and combining knowledge points of agricultural courseware according to claim 1, wherein extracting key entities from the agricultural knowledge text corresponding to the key frame image, and performing semantic association on the key entities and entity nodes in the agricultural domain knowledge graph to obtain a semantic association result of the key frame image comprises: Extracting key entities from the agricultural knowledge texts corresponding to the key frame images by adopting a natural language processing technology; calculating the semantic similarity between the key entity and each entity node in the agricultural domain knowledge graph, and determining the entity node with the semantic similarity larger than a first similarity threshold value as the semantic association result of the key frame image.
4. The method for dividing and combining knowledge points of agricultural courseware according to claim 1, wherein the knowledge graph of the agricultural field is constructed by the following steps: extracting a knowledge triplet from knowledge data in the multi-source agricultural field, wherein the knowledge triplet comprises an entity, an attribute and a relation; and constructing a knowledge graph of the agricultural field based on the knowledge triples.
5. The method for dividing and combining knowledge points of agricultural courseware according to claim 1, wherein dividing the agricultural courseware video into a plurality of agricultural knowledge point segments based on semantic association results of each key frame image comprises: Respectively carrying out visual analysis on each key frame image to obtain visual picture information of each key frame image; determining a knowledge point boundary based on semantic association results and visual picture information of each key frame image; and dividing the agricultural courseware video into a plurality of agricultural knowledge point segments based on the knowledge point boundaries.
6. The method for splitting and combining knowledge points of agricultural courseware according to claim 5, wherein determining knowledge point boundaries based on semantic association results and visual picture information of the respective key frame images comprises: under the condition that semantic association results of adjacent key frame images respectively belong to two node clusters with different semanteme, determining a first candidate boundary set based on video time points between the adjacent key frame images; Determining a second candidate boundary set based on video time points between adjacent key frame images in the case that the visual picture information of the adjacent key frame images has jumps; a knowledge point boundary is determined based on the first candidate boundary set and the second candidate boundary set.
7. The utility model provides an agricultural courseware knowledge point segmentation and composite set which characterized in that includes: The system comprises a courseware video acquisition module, a courseware video processing module and a storage module, wherein the courseware video acquisition module is used for acquiring an agricultural courseware video, and the agricultural courseware video comprises a plurality of key frame images and explanation voices corresponding to the key frame images; The voice recognition module is used for carrying out voice recognition on the explanation voices corresponding to the key frame images to obtain agricultural knowledge texts corresponding to the key frame images; The semantic association module is used for extracting key entities from the agricultural knowledge texts corresponding to the key frame images, and carrying out semantic association on the key entities and entity nodes in the agricultural domain knowledge graph to obtain semantic association results of the key frame images; The knowledge point dividing module is used for dividing the agricultural courseware video into a plurality of agricultural knowledge point fragments based on semantic association results of the key frame images, and labeling knowledge point labels for the agricultural knowledge point fragments; And the personalized combination module is used for extracting knowledge point fragments to be used from the plurality of agricultural knowledge point fragments based on the use requirement of a user, and generating a new agricultural knowledge learning video based on the combination of the knowledge point fragments to be used.
8. An electronic device comprising a memory, a processor and a computer program stored on the memory and running on the processor, wherein the processor implements the method of agricultural courseware knowledge point segmentation and combination of any one of claims 1 to 6 when the computer program is executed by the processor.
9. A non-transitory computer readable storage medium having stored thereon a computer program, wherein the computer program when executed by a processor implements the method of agricultural courseware knowledge point segmentation and combination of any of claims 1 to 6.
10. A computer program product comprising a computer program which, when executed by a processor, implements the method of agricultural courseware knowledge point segmentation and combination of any one of claims 1 to 6.

Description

Agricultural courseware knowledge point segmentation and combination method, device, equipment and storage medium Technical Field The invention relates to the technical field of intelligent education and agricultural information, in particular to a method, a device, equipment and a storage medium for dividing and combining knowledge points of agricultural courseware. Background With popularization of agricultural technology, streaming media courseware represented by short video and live recording and broadcasting becomes an important training form. However, the current agricultural streaming media courseware has the following problems that (1) the content organization structure is solidified, the courseware is usually continuous video for a long time, and a fine knowledge point structure is lacked, so that a learner is difficult to quickly locate, review or skip specific content, and the learning efficiency is low. (2) The knowledge points are isolated and difficult to be organized, and related knowledge in different coursewares are isolated from each other and cannot be dynamically associated and combined into a personalized learning path according to the knowledge base, interest targets or weak links of learners. (3) And the retrieval and the multiplexing are difficult, and a learner is difficult to accurately extract and reorganize the required knowledge segments from a massive video library to make new lessons. Therefore, how to realize the knowledge point segmentation and personalized combination of agricultural courseware is a problem to be solved at present. Disclosure of Invention The invention provides a method, a device, equipment and a storage medium for cutting and combining knowledge points of agricultural courseware, which are used for realizing intelligent cutting, labeling and personalized on-demand combination of video knowledge points of agricultural courseware. The invention provides a method for cutting and combining knowledge points of agricultural courseware, which comprises the following steps: The method comprises the steps of acquiring an agricultural courseware video, wherein the agricultural courseware video comprises a plurality of key frame images and explanation voices corresponding to the key frame images; aiming at the explanation voice corresponding to each key frame image, carrying out voice recognition on the explanation voice to obtain an agricultural knowledge text corresponding to the key frame image; Extracting a key entity from the agricultural knowledge text corresponding to the key frame image, and carrying out semantic association on the key entity and entity nodes in the agricultural domain knowledge graph to obtain a semantic association result of the key frame image; Dividing the agricultural courseware video into a plurality of agricultural knowledge point segments based on semantic association results of each key frame image, and labeling knowledge point labels for the agricultural knowledge point segments; and extracting knowledge point fragments to be used from the plurality of agricultural knowledge point fragments based on the use requirement of a user, and generating a new agricultural knowledge learning video based on the combination of the knowledge point fragments to be used. According to the method for segmenting and combining the knowledge points of the agricultural courseware provided by the invention, the speech recognition is carried out on the explanation speech to obtain the agricultural knowledge text corresponding to the key frame image, and the method comprises the following steps: inputting the explanation voice into a pre-trained voice recognition model to obtain an initial knowledge text output by the voice recognition model; And carrying out text error correction on the initial knowledge text to obtain the agricultural knowledge text corresponding to the key frame image. According to the method for segmenting and combining the knowledge points of the agricultural courseware provided by the invention, the key entity is extracted from the agricultural knowledge text corresponding to the key frame image, the key entity and the entity node in the knowledge graph of the agricultural field are subjected to semantic association, and the semantic association result of the key frame image is obtained, and the method comprises the following steps: Extracting key entities from the agricultural knowledge texts corresponding to the key frame images by adopting a natural language processing technology; calculating the semantic similarity between the key entity and each entity node in the agricultural domain knowledge graph, and determining the entity node with the semantic similarity larger than a first similarity threshold value as the semantic association result of the key frame image. According to the method for dividing and combining knowledge points of agricultural courseware, which is provided by the invention, the knowledge graph in the agricult