Music and Dance College, Heze University, Shandong, Heze, China.
Comput Intell Neurosci. 2022 Jun 16;2022:5499593. doi: 10.1155/2022/5499593. eCollection 2022.
With the rapid development of information technology and mobile Internet, digital image, text, audio, video, and other cross-media data are growing explosively, which has changed people's way of life and work. In view of the issues of negative studying effectivity and challenging attention of college students in the modern-day piano instructing process, this paper puts forward the application of knowledge Atlas technology in piano teaching and constructs a multimodal knowledge Atlas of piano teaching based on deep neural network, so as to make piano teaching more intelligent and improve students' learning efficiency and learning interest. How to realize the semantic association understanding of cross-media data is the core problem of cross-media semantic analysis. First, this paper introduces the basic rules of ontology construction and the basic method of establishing general knowledge graph are introduced. Then, taking the piano teaching content as an example, natural language sentences can be expressed and stored with cross-media data using semantic network. The mathematical understanding is extracted in accordance to the herbal language processing technology, and the entities are fused in accordance to the frequent semantic similarity detection between extraordinary entities, so as to decrease the redundancy and repetition fee of entities and the complexity of the graph. The fused new knowledge is processed according to the quality evaluation rules, the qualified part is added to the knowledge base, and then the above steps are iterated to update the database. The great overall performance of piano instructing understanding graph mannequin primarily based on semantic network is validated through experiments.
随着信息技术和移动互联网的飞速发展，数字图像、文本、音频、视频等跨媒体数据呈爆炸式增长，改变了人们的生活和工作方式。针对现代钢琴教学过程中学生学习效果不佳和注意力难以集中的问题，本文提出了知识图谱技术在钢琴教学中的应用，构建了基于深度神经网络的钢琴教学多模态知识图谱，使钢琴教学更加智能化，提高学生的学习效率和学习兴趣。如何实现跨媒体数据的语义关联理解是跨媒体语义分析的核心问题。首先，本文介绍了本体构建的基本规则和建立通用知识图谱的基本方法。然后，以钢琴教学内容为例，使用语义网络将自然语言句子与跨媒体数据表示和存储。然后，根据自然语言处理技术提取数学理解，根据异常实体之间的频繁语义相似性检测融合实体，以减少实体的冗余和重复费用以及图形的复杂性。根据质量评估规则处理融合的新知识，将合格部分添加到知识库中，然后迭代执行上述步骤以更新数据库。通过实验验证了基于语义网络的钢琴教学理解图模型的优异总体性能。
