WO2022156223A1 - 一种基于增强现实技术知识森林导航学习系统实现方法 - Google Patents

一种基于增强现实技术知识森林导航学习系统实现方法 Download PDF

Info

Publication number
WO2022156223A1
WO2022156223A1 PCT/CN2021/116052 CN2021116052W WO2022156223A1 WO 2022156223 A1 WO2022156223 A1 WO 2022156223A1 CN 2021116052 W CN2021116052 W CN 2021116052W WO 2022156223 A1 WO2022156223 A1 WO 2022156223A1
Authority
WO
WIPO (PCT)
Prior art keywords
knowledge
cluster
topic
topics
forest
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.)
Ceased
Application number
PCT/CN2021/116052
Other languages
English (en)
French (fr)
Inventor
郑庆华
武云封
刘均
魏笔凡
张玲玲
罗强
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Xian Jiaotong University
Original Assignee
Xian Jiaotong University
Priority date (The priority date 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 date listed.)
Filing date
Publication date
Application filed by Xian Jiaotong University filed Critical Xian Jiaotong University
Publication of WO2022156223A1 publication Critical patent/WO2022156223A1/zh
Anticipated expiration legal-status Critical
Ceased legal-status Critical Current

Links

Images

Classifications

    • G—PHYSICS
    • G06—COMPUTING OR CALCULATING; COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
    • G06F16/90—Details of database functions independent of the retrieved data types
    • G06F16/95—Retrieval from the web
    • G06F16/954—Navigation, e.g. using categorised browsing
    • G—PHYSICS
    • G06—COMPUTING OR CALCULATING; COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
    • G06F16/90—Details of database functions independent of the retrieved data types
    • G06F16/901—Indexing; Data structures therefor; Storage structures
    • G06F16/9027—Trees
    • G—PHYSICS
    • G06—COMPUTING OR CALCULATING; COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
    • G06F16/90—Details of database functions independent of the retrieved data types
    • G06F16/95—Retrieval from the web
    • G06F16/951—Indexing; Web crawling techniques
    • 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

Definitions

  • the invention belongs to the field of augmentation, and in particular relates to an implementation method of a knowledge forest navigation learning system based on augmented reality technology.
  • the rapid development of Internet technology has provided great convenience for people to acquire knowledge, but also brought about the problem of knowledge fragmentation.
  • the phenomenon of knowledge fragmentation refers to the fact that knowledge of different aspects of a specific topic is distributed in different data sources, which is manifested in three aspects: scattered location, one-sided content, and disordered structure. In the process of user learning, it will cause cognitive overload and learning. Trek problem.
  • knowledge forest is proposed as a brand-new fragmented knowledge aggregation model.
  • fragmented knowledge is aggregated into a form that conforms to human perception. Know the mode of learning multi-dimensional, related, integrated and other characteristics.
  • Visualization technology is a technology that uses the principles and methods of computer graphics to express large-scale data in an intuitive form, involving computer graphics, image processing, computer vision and other fields.
  • Data visualization is limited to two-dimensional image display. Compared with directly reading the data in the database, two-dimensional visualization can be displayed relatively intuitively.
  • Knowledge Forest realizes the knowledge forest web version system through js visualization technology such as d3. Although the knowledge forest of the web version has the characteristics of showing the aggregated results of the knowledge forest, it cannot get rid of the limitations of the two-dimensional visualization technology, such as the limitations on the visualization scale, layout and use of the knowledge forest.
  • the two-dimensional hierarchical data visualization method either cannot fully display the hierarchical structure, resulting in one-sided cognition; or the high distortion rate causes the cognitive burden of learners; or the content is obscured and the interactive experience is reduced, all of which make knowledge forest visualization difficult.
  • the purpose of the present invention is to provide an implementation method of a knowledge forest navigation learning system based on augmented reality technology, so as to overcome the deficiencies of the prior art.
  • the present invention adopts the following technical solutions:
  • a method for implementing a knowledge forest navigation learning system based on augmented reality technology comprising the following steps:
  • S1 obtain data, the data includes several clusters into which all knowledge topics of a course are divided, the learning sequence relationship between each cluster, the knowledge topics included in each cluster, and the recognition between knowledge topics in each cluster.
  • step S2 using the knowledge topic obtained in step S1 and its corresponding faceted data to construct a knowledge topic faceted tree;
  • step S3 according to the several clusters divided in step S1 and the learning sequence relationship between each cluster, the coordinate data of each knowledge cluster is generated by the algorithm of the fan-shaped layout, and the course data sand table scene knowledge cluster scene is obtained;
  • step S4 the knowledge subject faceted tree model obtained in step S2 is laid out in the knowledge cluster of the corresponding course data sand table scene through the node classification and layout method, and then a knowledge forest navigation learning system containing the course data sand table scene is generated;
  • step S5 customize the interaction of the augmented reality system according to the knowledge forest navigation learning system obtained in step S4.
  • the walktrap algorithm to perform community detection on the curriculum information extracted from the knowledge forest database, it is divided into clusters, and the data of different clusters corresponding to different courses are obtained.
  • each cluster also includes the number of knowledge topics included in the cluster; the cognitive relationship between the knowledge topics in the cluster includes the starting topic name and the ending topic name of each cognitive relationship; the faceted data includes The facet under the knowledge topic and the knowledge fragment data corresponding to the facet.
  • the facets under the knowledge topic include the name, level, and parent facet of the facet; knowledge fragments include the type, content, URL, and facet of the knowledge fragment; cognitive relationships include the upper-lower relationship, and the learning dependency relationship. and causality.
  • step S2 the knowledge topic faceted tree is constructed by the method of model splicing.
  • step S2 the specific process of step S2 is as follows: first, construct models of trunks, branches and leaves, and add M fixed points at appropriate positions on the trunk model according to the maximum number M of first-level facets contained in a knowledge topic;
  • the maximum number of second-level facets contained in the level-level facet is N, and N fixed points are added at appropriate positions on the first-level branch;
  • the number of fractional planes is N2, M1 points are randomly selected on the M fixed points of the trunk model, M1 first-level branch models are added, and the branches are removed from the M fixed points of each fixed first-level branch model.
  • N2 points from the low end to the top to add N2 secondary branch models, and finally add leaf models to the top of each primary branch and secondary branch model, and adjust the balance of M1 branch models on the trunk model
  • the spatial coordinates, spatial steering and size scaling data of the leaf model are saved from top to bottom according to the branch model and the trunk model, so as to obtain the knowledge topic faceted tree model of N2 branches.
  • step S3 is: according to the number of knowledge topics w of each knowledge cluster as the weight to determine its proportion in the sand table, and use a power function to balance the weights of different clusters:
  • the resulting w 0 is used as the new weight for each cluster; the resulting new weight is normalized:
  • w i is the new weight for each cluster, is the normalized weight occupied by each cluster; then the angle is determined according to the weight, and each knowledge cluster is divided into a fan-shaped area.
  • the circle center coordinates x i and y i of every two clusters are connected to obtain the learning sequence relationship between clusters; thus completing the course data sand table scene knowledge cluster inter-scenario construction.
  • step S4 specifically includes the following steps: classifying the knowledge topics in a cluster, classifying the knowledge topics with only one connection point as the first type of knowledge topics, and placing the first type of knowledge topics on the periphery of the sand table in the knowledge cluster. ; Put the knowledge topics directly connected to the first type of knowledge topics in the middle layer of the sand table, and put the remaining knowledge topics in the innermost part of the sand table.
  • the augmented reality system interaction in step S5 includes gesture control for zooming in, zooming out, and rotation, and a return button for controlling the return to the previous scene.
  • the present invention has the following beneficial technical effects:
  • the present invention is an implementation method of a knowledge forest navigation and learning system based on augmented reality technology.
  • the node classification and layout method are arranged in order to generate a knowledge forest navigation and learning system containing a sand table scene of course data, which breaks through the two-dimensional visualization technology in forest scale, layout and interaction.
  • the use of knowledge topic faceted tree combined with sand table scene knowledge clusters to form a multi-dimensional display can realize the cognitive relationship between multiple topics and the association between topics, concise and clear, easy to interact; adaptive according to needs Faceted trees with different styles can be generated easily, so that the generated tree is simple and beautiful, and the storage space is greatly saved.
  • the invention realizes the intelligent scene discrimination and course discrimination methods, and reduces the learning cost of users.
  • the cluster structure is introduced, and the community detection algorithm is used to divide all topics into different clusters.
  • the connection between clusters represents the learning sequence relationship between clusters and provides navigation for users. After entering a cluster, display By analyzing all the knowledge topics in this cluster and the cognitive relationship between the knowledge topics, the number of knowledge topics in the sand table is greatly reduced, and users may see a course sand table with a more reasonable, clear, and easy-to-interact layout.
  • FIG. 1 is a flowchart of a knowledge topic faceted tree in an embodiment of the present invention.
  • FIG. 2 is a schematic diagram of scene construction between knowledge clusters in a course data sand table scene in an embodiment of the present invention.
  • FIG. 3 is a schematic diagram of scene construction in a sandbox scene knowledge cluster according to an embodiment of the present invention.
  • FIG. 4 is a schematic diagram of interaction of a knowledge forest navigation learning system in an embodiment of the present invention.
  • An implementation method of a knowledge forest navigation learning system based on augmented reality technology which breaks through the limitations of two-dimensional visualization technology in forest scale, layout and interaction, brings users a strong novelty experience and a sense of reality, and makes users more focused , long-lasting, making the learning effect more efficient and the learning interest more intense.
  • a mobile terminal system for knowledge forest navigation and learning based on augmented reality technology including a data layer, a service layer and a user layer, the data layer includes a knowledge forest database and user account information and learning logs; Discrimination, course discrimination, faceted tree and sand table automatic layout; the user layer mainly includes camera shooting, AR display and other functions.
  • a method for implementing a knowledge forest navigation learning system based on augmented reality technology comprising the following steps:
  • S1 obtain data, the data includes several clusters into which all knowledge topics of a course are divided, the learning sequence relationship between each cluster, the knowledge topics included in each cluster, and the recognition between knowledge topics in each cluster.
  • Each of the clusters also includes the number of knowledge topics contained in the cluster; the cognitive relationship between the knowledge topics in the cluster includes the starting topic name and the ending topic name of each cognitive relationship; the faceted data includes the knowledge topic under the knowledge topic.
  • the facets under the knowledge topic include the facet name, level, and the parent facet to which it belongs;
  • knowledge fragments include the type, content, URL, and facet of the knowledge fragment;
  • cognitive relationships include the upper and lower relationships, learning dependencies, and causality relation;
  • the construction of the knowledge topic faceted tree scene the knowledge topic and the corresponding faceted data obtained in step S1 are used to construct the knowledge topic faceted tree, which realizes the construction of the knowledge topic faceted tree by the method of model splicing; , first build the model of the trunk, branches and leaves, the model is a 3D model, according to the maximum number M of first-level facets contained in a knowledge topic, add M fixed points at appropriate positions on the trunk model; The maximum number of second-level facets contained in the facet is N, and N fixed points are added at appropriate positions on the first-level branch; then, according to the number of first-level facets M1 owned by the knowledge topic and the second-level facets owned by each first-level facet The number of facets is N2, and M1 points are randomly selected on the M fixed points of the trunk model, and M1 first-level branch models are added.
  • N2 points from the top to the top to add N2 secondary branch models, and finally add leaf models to the top of each primary branch and secondary branch model, and adjust the balance of M1 branch models on the trunk model.
  • the spatial coordinates, spatial steering and size scaling data of the leaf model are saved according to the branch model and the trunk model from top to bottom, so as to obtain the knowledge topic faceted tree model of N2 branches;
  • the obtained w 0 is used as the new weight of each cluster, which ensures that the number of knowledge topics contained in the knowledge cluster can be distinguished from the weight of the knowledge cluster, and also avoids the problem of excessive weight gap;
  • w i is the new weight for each cluster, is the normalized weight occupied by each cluster; then the angle is determined according to the weight, and each knowledge cluster is divided into a fan-shaped area, and the central angle corresponding to the fan-shaped area is:
  • ⁇ is the length of the polar axis of a point on the ellipse.
  • step S4 the construction of the scene in the knowledge cluster of the course data sand table scene: as shown in Figure 3, the knowledge subject faceted tree model obtained in step S2 is laid out in the corresponding course data sand table scene knowledge cluster through the node classification and sequential layout method, and then Generate a knowledge forest navigation and learning system containing a sand table scene of course data; specifically, classify the knowledge topics in a cluster, and the knowledge topics with only one connection point are classified as the first type of knowledge topics, and this kind of knowledge topics has only one connection lead.
  • the knowledge topic directly connected with the first type of knowledge topic is classified as the second type of knowledge topic, this category is to connect a type of knowledge topic It is placed in the middle layer of the sand table to ensure that it is perfectly connected to the peripheral one type of topics and also to the remaining knowledge topics; the remaining knowledge topics are divided into the third type of knowledge topics, because They are not directly connected to the outermost category of knowledge topics, but are often connected to each other or serve as the intermediate points for the interconnection of the second category of knowledge topics, placing such points in the innermost part of the sandbox.
  • the nodes of the first type of knowledge topic, the second type of knowledge topic and the third type of knowledge topic are placed on the sand table in the knowledge cluster in turn, and then the corresponding nodes that have a cognitive relationship between the knowledge topics are connected. line to realize the layout of the scenes in the sand table scene cluster;
  • step S5 customize the interaction of the augmented reality system according to the knowledge forest navigation learning system obtained in step S4.
  • the overall interaction process of the system is realized.
  • the user scans the book, if the user enters the sand table scene knowledge In the inter-cluster scene, click one of the discs to enter the sand table scene knowledge cluster scene, click one of the small trees, you can enter the faceted tree scene of the knowledge topic, click the facet button to pop up the display under the facet knowledge fragment content, and switch to the knowledge fragment corresponding to this facet.
  • the return to sand table button on the screen can return to the sand table scene and the knowledge cluster scene of the course to which the knowledge topic belongs.
  • the user uses a third-party OCR to identify all the text contained in the scanned image; first The obtained text information is matched with the keywords in the course keyword dictionary. If the matching is successful, the scene flag is set as the course learning, and the matched course keywords are regarded as the identification course of the course learning scene. If the keywords in the course keyword dictionary are not matched, continue to match the keywords in the topic keyword dictionary. If the match is successful, set the scene flag to topic learning, and count the number of topics appearing in the text and the corresponding courses. , and then count the number of topics included in each course on this page, select the course with the most topics as the recommended course for the topic learning scenario, and the system displays all knowledge topics belonging to the course. If all matches fail, set the scene flag to rescan.
  • the screen size of mobile devices is limited, and a course generally has dozens of topics, and individual courses even have hundreds of topics. If all the topic faceted trees are displayed in the sandbox, it will be very crowded, which will seriously affect the visualization experience. Therefore, the concept of clusters is introduced, and all topics in a course are divided into several clusters, which greatly alleviates the problem of small screens, so that the knowledge forest data can be displayed to users in a hierarchical manner. At the same time, the knowledge forest model generated by the layout method is concise. Clear and easy to interact with. First, the information of each cluster and the relationship between each cluster are displayed in the sand table. Then, after entering a cluster, all knowledge topics in the cluster and the cognitive relationship between knowledge topics are displayed.
  • the knowledge forest realization method of the present invention includes two parts: the layout method between knowledge clusters and the layout method within the knowledge cluster.
  • the present invention realizes the method of intelligent scene discrimination and course discrimination. In the scanning interface of the system, the user only needs to scan the cover of a book or For any page, the system will automatically determine the user's learning intention.
  • the scene discriminator can determine whether the user wants to learn a topic or the content of a course.
  • the course discriminator can automatically determine which course the user scans. The user makes further choices, which improves the user experience.
  • the present invention provides a faceted tree automatic layout method based on statistical analysis and dot positioning. There is no need to store hundreds or thousands of 3D models, and faceted trees with different styles can be adaptively generated according to needs, which makes the generated tree simple and beautiful, easy to interact, greatly saves storage space, and improves the aesthetics of the model.
  • Knowledge forest It is a brand-new mode of fragmented knowledge aggregation.
  • the process of fragmented knowledge aggregation is regarded as the three-stage knowledge forest generation process of "knowledge topic faceted tree generation - fragmented knowledge assembly - cognitive relationship mining". It aggregates multi-source, one-sided, and disorderly fragmented knowledge, such as Chinese and English texts, pictures, videos and other multimedia resources into a model that conforms to the characteristics of human cognitive learning, such as multidimensionality, relevance, and integration.
  • Knowledge topic faceted tree refers to a hierarchical tree that describes the organizational structure of a knowledge topic.
  • the data of the knowledge topic faceted tree mainly includes branch data.
  • Cognitive relationship refers to the dependency relationship between knowledge topics due to cognitive needs in the learning process.
  • the system uses the community detection algorithm to divide all topics into different clusters, and the lines between clusters represent the learning sequence relationship between clusters. Users can have an overview of all the topics of the course, and the connections between clusters also provide guidance for the user's learning. After clicking on a knowledge cluster on the course sand table interface, all themes and cognitive relationships between themes in the cluster will be displayed. Cognitive relationships between topics provide users with navigation, which can be very effective in alleviating the problem of learning trek.
  • the user can choose the topic they are interested in, and then jump to the faceted tree scene to learn a topic in detail, click the return button, the system returns to the knowledge cluster interface, and the user can choose another topic to learn. Click the return button again, the system returns to the sand table scene, and the user can select another cluster for learning. Click the return button again, the system returns to the scanning interface, and the user can scan the cover of another book to study.

Landscapes

  • Engineering & Computer Science (AREA)
  • Databases & Information Systems (AREA)
  • Theoretical Computer Science (AREA)
  • Data Mining & Analysis (AREA)
  • Physics & Mathematics (AREA)
  • General Engineering & Computer Science (AREA)
  • General Physics & Mathematics (AREA)
  • Radar, Positioning & Navigation (AREA)
  • Remote Sensing (AREA)
  • Software Systems (AREA)
  • Information Retrieval, Db Structures And Fs Structures Therefor (AREA)
  • User Interface Of Digital Computer (AREA)

Abstract

一种基于增强现实技术知识森林导航学习系统实现方法，通过节点分类依次布局方法布局，生成含有课程数据沙盘场景的知识森林导航学习系统，突破了二维可视化技术在森林规模、布局和交互上存在的局限，采用知识主题分面树结合沙盘场景知识簇间场景构成多维度展示，实现多主题之间的认知关系以及主题之间的关联，易于交互；根据需要自适应地生成样式各异的分面树，使得生成的树简洁美观，极大地节省了存储空间，该方法实现了智能场景判别和课程判别方法，采用簇结构，在进入一个簇后，展示出这个簇内的所有知识主题以及知识主题间的认知关系，沙盘中知识主题数量大大减少，用户可能看到布局更加合理、清晰、易于交互的课程沙盘。

Description

一种基于增强现实技术知识森林导航学习系统实现方法 技术领域
本发明属于增强领域，具体涉及一种基于增强现实技术知识森林导航学习系统实现方法。
背景技术
互联网技术的飞速发展为人们获取知识提供了极大的便利，同时也带来了知识碎片化问题。知识碎片化现象指的是特定主题不同分面的知识分布在不同的数据源中，表现在位置分散、内容片面、结构无序三个方面，在用户学习的过程中会引起认知过载和学习迷航问题。针对这一问题，知识森林作为一种全新的碎片化知识聚合模式被提出，通过“主题分面树生成—碎片化知识装配—认知关系挖掘”三个阶段将碎片化知识聚合成为符合人类认知学习多维性、关联性、整合性等特点的模式。
知识森林中存在海量的层次数据，用户难以直接通过数据库学习知识森林数据里的知识，此外，利用人工进行知识主题分面树和认知关系的布局耗时耗力并且难以维护。
可视化技术是利用计算机图形学的原理和方法，将大规模数据以直观地形式表现出来的技术，涉及到计算机图形学、图像处理、计算机视觉等多个领域。数据可视化局限于二维图像展现，相对于直接阅读数据库中数据，二维可视化已经可以相对直观地展现，知识森林通过d3等js可视化技术，实现了知识森林网页版系统。虽然网页版的知识森林具有展现知识森林聚合结果的特点，但不能摆脱二维可视化技术的局限，比如对知识森林可视化规模、布局以及使用的限制。二维的层次数据可视化方法或者无法完全显示层次结构，造成认知的片面；或者畸变率高，造成学习者的认知负担；或者内容遮 蔽，降低交互体验，均造成知识森林可视化困难重重。
现有的增强现实应用多使用固定模型，即每次使用时用户看到的和可操作的都是一样的，没有什么变化，固定模型无法满足所有课程和所有主题的可视化。
发明内容
本发明的目的在于提供一种基于增强现实技术知识森林导航学习系统实现方法，以克服现有技术的不足。
为达到上述目的，本发明采用如下技术方案：
一种基于增强现实技术知识森林导航学习系统实现方法，包括以下步骤：
S1，获取数据，所述数据包括一个课程的所有知识主题划分成的若干个簇、各个簇之间的学习先后序关系、每个簇包含的知识主题、每个簇内知识主题之间的认知关系数据、知识主题对应分面以及分面对应知识分面碎片数据；
S2，利用步骤S1中获得的知识主题及其对应的分面数据构建知识主题分面树；
S3，根据步骤S1中划分成的若干个簇及各个簇之间的学习先后序关系，通过扇形布局的算法生成每个知识簇的坐标数据，得到课程数据沙盘场景知识簇间场景；
S4，将步骤S2得到的知识主题分面树模型通过节点分类依次布局方法布局在与其对应的课程数据沙盘场景知识簇内，然后生成含有课程数据沙盘场景的知识森林导航学习系统；
S5，根据步骤S4得到的知识森林导航学习系统定制增强现实系统交互。
进一步的，利用walktrap算法将从知识森林数据库中提取到的课程信息进行社团检测划分为簇，得到不同课程对应的不同簇的数据。
进一步的，所述每个簇还包括这个簇包含的知识主题的个数；簇内知识 主题之间的认知关系包括每条认知关系的起始主题名以及终止主题名；分面数据包括知识主题下的分面以及该分面对应的知识碎片数据。
进一步的，知识主题下的分面包括分面的名称、层级、所属的上级分面，知识碎片包括知识碎片的类型、内容、URL和所属分面；认知关系包括上下位关系，学习依赖关系和因果关系。
进一步的，步骤S2中通过模型拼接的方法实现知识主题分面树的构建。
进一步的，步骤S2具体过程为：首先构建树干、树枝以及树叶的模型，根据一个知识主题包含的最多一级分面数量M，在树干模型上合适的位置添加M个固定点位；根据一个一级分面包含的最多二级分面数量N，在一级树枝上合适的位置添加N个固定点位；然后根据知识主题拥有的一级分面数量M1和每个一级分面拥有的二级分面数量N2，在树干模型的M个固定点位上随机选择M1个点位，添加M1个一级树枝模型，在每个固定好的一级树枝模型的M个固定点位中从树枝低端向顶端方向选择N2个点位添加N2个二级树枝模型，最后将树叶模型添加到每个一级树枝以及二级树枝模型的顶端，并且调整M1个树枝模型在树干模型上的平衡性，同时将树叶模型的空间坐标、空间转向和大小缩放数据按照树枝模型和树干模型从上至下进行保存，从而得到N2分枝的知识主题分面树模型。
进一步的，步骤S3具体过程为：根据每个知识簇的知识主题数量w作为权重确定其在沙盘中所占的比例，使用一个幂函数对不同簇的权重进行均衡：
Figure PCTCN2021116052-appb-000001
得到的w 0作为每个簇的新权重；对得到的新权重进行归一化处理：
Figure PCTCN2021116052-appb-000002
w i是每个簇的新权重，
Figure PCTCN2021116052-appb-000003
是每个簇所占的归一化权重；然后根据权重来确定角度，给每个知识簇划分一个扇形区域。
进一步的，将从第一簇到最后一个簇之间，每2个簇的圆心坐标x i、y i相连接，得到簇与簇之间的学习先后序关系；从而完成课程数据沙盘场景知识簇间场景构建。
进一步的，步骤S4具体包括以下步骤：对一个簇内的知识主题进行分类，将只有一个连接点的知识主题归为第一类知识主题，将第一类知识主题放在知识簇内沙盘的外围；将与第一类知识主题直接相连的知识主题放在沙盘的中间层，将剩余的知识主题放在沙盘的最内部。
进一步的，步骤S5中所述增强现实系统交互包括进行放大、缩小、旋转的手势控制，以及控制返回上一场景的返回按钮。
与现有技术相比，本发明具有以下有益的技术效果：
本发明一种基于增强现实技术知识森林导航学习系统实现方法，通过节点分类依次布局方法布局，生成含有课程数据沙盘场景的知识森林导航学习系统，突破了二维可视化技术在森林规模、布局和交互上存在的局限，采用知识主题分面树结合沙盘场景知识簇间场景构成多维度展示，能够实现多主题之间的认知关系以及主题之间的关联，简明清晰，易于交互；根据需要自适应地生成样式各异的分面树，使得生成的树简洁美观，极大地节省了存储空间，本发明实现了智能场景判别和课程判别方法，降低了用户的学习成本。引入了簇结构，利用社团检测算法把所有主题分为不同的簇，簇与簇间的连线表示簇与簇之间的学习先后序关系，为用户提供了导航，在进入一个簇后，展示出这个簇内的所有知识主题以及知识主题间的认知关系，沙盘中知识主题数量大大减少，用户可能看到布局更加合理、清晰、易于交互的课程沙盘。
附图说明
图1为本发明实施例中知识主题分面树流程图。
图2为本发明实施例中课程数据沙盘场景知识簇间场景构建原理图。
图3为本发明实施例中沙盘场景知识簇内场景构建原理图。
图4为本发明实施例中知识森林导航学习系统交互示意图。
具体实施方式
下面结合附图对本发明做进一步详细描述：
一种基于增强现实技术知识森林导航学习系统实现方法，突破了二维可视化技术在森林规模、布局和交互上存在的局限，给用户带来强烈的新奇体验以及真实感，使得用户注意力更加集中、持久，使得学习效果更加高效，学习兴趣更加浓厚。
一种基于增强现实技术的知识森林导航学习移动端系统，包括数据层、服务层和用户层，所述数据层包括知识森林数据库以及用户的账户信息和学习日志；服务层包括实现文字识别、场景判别、课程判别，分面树和沙盘自动布局；用户层主要包括相机拍摄、AR展示等功能。
一种基于增强现实技术知识森林导航学习系统实现方法，包括以下步骤：
S1，获取数据，所述数据包括一个课程的所有知识主题划分成的若干个簇、各个簇之间的学习先后序关系、每个簇包含的知识主题、每个簇内知识主题之间的认知关系数据、知识主题对应分面以及分面对应知识分面碎片数据；
具体的，利用walktrap算法将从知识森林数据库中提取到的课程信息进行社团检测划分为簇，得到不同课程对应的不同簇的数据；
所述每个簇还包括这个簇包含的知识主题的个数；簇内知识主题之间的认知关系包括每条认知关系的起始主题名以及终止主题名；分面数据包括知识主题下的分面以及该分面对应的知识碎片数据。其中知识主题下的分面包括分面的名称、层级、所属的上级分面，知识碎片包括知识碎片的类型、内容、URL和所属分面；认知关系包括上下位关系，学习依赖关系和因果关系；
S2，知识主题分面树场景的构建：利用步骤S1中获得的知识主题及其对应的分面数据构建知识主题分面树，其通过模型拼接的方法实现知识主题分 面树的构建；具体的，首先构建树干、树枝以及树叶的模型，所述模型为3D模型，根据一个知识主题包含的最多一级分面数量M，在树干模型上合适的位置添加M个固定点位；根据一个一级分面包含的最多二级分面数量N，在一级树枝上合适的位置添加N个固定点位；然后根据知识主题拥有的一级分面数量M1和每个一级分面拥有的二级分面数量N2，在树干模型的M个固定点位上随机选择M1个点位，添加M1个一级树枝模型，在每个固定好的一级树枝模型的M个固定点位中从树枝低端向顶端方向选择N2个点位添加N2个二级树枝模型，最后将树叶模型添加到每个一级树枝以及二级树枝模型的顶端，并且调整M1个树枝模型在树干模型上的平衡性，同时将树叶模型的空间坐标、空间转向和大小缩放数据按照树枝模型和树干模型从上至下进行保存，从而得到N2分枝的知识主题分面树模型；
在树干上合适的位置打点来安放下级树枝，规定所有的子树枝只能在所打的点处以固定的角度来放置，保证了树枝之间没有交叉，实现了自适应地生成不同主题的分面树，使得生成的树简洁美观，易于交互。
S3，课程数据沙盘场景知识簇间场景构建：如图2所示，根据步骤S1中划分成的若干个簇及各个簇之间的学习先后序关系，通过扇形布局的算法生成每个知识簇的坐标数据；具体的：根据每个知识簇的知识主题数量w作为权重确定其在沙盘中所占的比例，为了使知识簇的分布能够体现簇间的学习先后序关系，并使分布区域的大小和自身簇内的知识主题数量一致，在保证美观性的情况下，使用一个幂函数对不同簇的权重进行均衡：
Figure PCTCN2021116052-appb-000004
得到的w 0作为每个簇的新权重，这样保证了从知识簇的权重大小可以区分其含有的知识主题数量，同时也避免了权重差距过大的问题；
对得到的新权重进行归一化处理：
Figure PCTCN2021116052-appb-000005
w i是每个簇的新权重，
Figure PCTCN2021116052-appb-000006
是每个簇所占的归一化权重；然后根据权重来确定角度，给每个知识簇划分一个扇形区域，该扇形区域对应的圆心角即：
Figure PCTCN2021116052-appb-000007
设所有的知识簇所占的区域圆心均分布在一个椭圆上，这样每个知识簇的区域圆心坐标使用如下的公式即可计算得到：
x i＝ρ 1cosθ i,y i＝ρ 2sinθ i
其中ρ即所在椭圆上一点的极轴长度。
再计算每个簇的半径：
Figure PCTCN2021116052-appb-000008
根据得到的簇的圆心坐标x i、y i以及半径r i，画出每个簇的圆盘；
在每个簇上圆盘位置上再画上w颗小树；
将从第一簇到最后一个簇之间，每2个簇的圆心坐标x i、y i相连接，得到簇与簇之间的学习先后序关系；从而完成课程数据沙盘场景知识簇间场景构建。
S4，课程数据沙盘场景知识簇内场景的构建：如图3所示，将步骤S2得到的知识主题分面树模型通过节点分类依次布局方法布局在与其对应的课程数据沙盘场景知识簇内，然后生成含有课程数据沙盘场景的知识森林导航学习系统；具体的，对一个簇内的知识主题进行分类，只有一个连接点的知识主题归为第一类知识主题，这种知识主题只有一个连线引出或进入该知识主题，将第一类知识主题放在知识簇内沙盘的外围；将与第一类知识主题直接相连的知识主题归为第二类知识主题，这一类是连接一类知识主题的枢纽， 将其放在沙盘的中间层，以保证和外围的一类主题完美连接的同时也能很好地连接剩余的知识主题；将剩余的知识主题被划分为第三类知识主题，由于他们不与最外围的一类知识主题直接相连，而往往互相连接或者作为第二类知识主题相互连接的中间点，将这类点放在沙盘的最内部。知识主题分类完毕后，依次将第一类知识主题、第二类知识主题以及第三类知识主题的节点安放在知识簇内沙盘上，然后将知识主题之间存在认知关系的对应节点进行连线，实现沙盘场景簇内场景的布局；
S5，根据步骤S4得到的知识森林导航学习系统定制增强现实系统交互。
根据步骤S2得到的分面树场景以及步骤S3得到的沙盘场景知识簇间场景以及步骤S4得到的沙盘场景知识簇内场景，实现系统的整体交互流程，当用户扫描书本后，如果进入沙盘场景知识簇间场景，点击其中的一个圆盘可以进入到沙盘场景知识簇内场景，点击其中的一棵小树，可以进入到该知识主题的分面树场景，点击分面按钮会弹出显示分面下知识碎片内容，并且切换至该分面对应的知识碎片。当用户扫描书本后，如果进入分面树场景，屏幕上的返回沙盘按钮可以返回到该知识主题所属课程的沙盘场景知识簇间场景。
如图1、图4所示，在知识森林导航学习系统的扫描界面，用户在对一本书的封面或者内页进行扫描后，利用第三方OCR识别出扫描的图片中包含的所有文字；首先将得到的文字信息与课程关键词字典中的关键词进行匹配，匹配成功，将场景标志置为课程学习，将匹配到的课程关键词当作课程学习场景的识别课程。如果未匹配到课程关键词字典中的关键词，继续与主题关键词字典中的关键词进行匹配，若匹配成功，将场景标志置为主题学习，同时统计文字中出现的主题数以及对应的课程，进而统计出此页内容包含在每门课程中的主题数，选择包含主题数最多的课程当作主题学习场景的推荐课程，同时系统展示出属于该课程的所有知识主题。若都匹配失败，将场景标 志置为重新扫描。
移动设备屏幕大小有限，而一门课程一般有数十个主题，个别课程甚至有上百个主题，如果将所有主题分面树全部展示在沙盘中会十分拥挤，严重影响可视化体验。因此，引入簇的概念，将一门课程中的所有主题划分成若干个簇，极大地缓解了小屏幕问题，使得知识森林数据可以游刃有余地分级展示给用户，同时布局方法生成的知识森林模型简明清晰，易于交互。首先，在沙盘中展示各个簇的信息以及各个簇之间的关系。然后，在进入一个簇后，展示出这个簇内的所有知识主题以及知识主题间的认知关系，这时沙盘中知识主题数量已经大大减少，用户可能看到布局更加合理、清晰、易于交互的课程沙盘。本发明的知识森林实现方法包括知识簇间布局方法以及知识簇内布局方法两部分，本发明实现了智能场景判别和课程判别方法，在系统的扫描界面，用户只需要扫描一本书的封面或者其中的任意一页，系统会自动判断出用户的学习意图，场景判别器可以判断用户想学习一个主题还是一门课程的内容，课程判别器可以自动判断出用户扫描的书属于哪门课程，无需用户进行进一步的选择，提高了用户体验。
本发明针对知识森林不同主题拥有的分面数及碎片数不同的特征而导致的不同主题AR树形状也不同的问题，基于统计分析以及打点定位的分面树自动布局方法。无需存储成百上千种的三维模型，根据需要自适应地生成样式各异的分面树，使得生成的树简洁美观，易于交互，极大地节省了存储空间，提高了模型美观度。
知识森林：是一种全新的碎片化知识聚合模式，将碎片化知识聚合过程看作是“知识主题分面树生成—碎片化知识装配—认知关系挖掘”三阶段的知识森林生成过程，能够将多源、片面、无序的碎片化知识如中英文的文本以及图片、视频等多媒体资源聚合成符合人类认知学习多维性、关联性、整合性等特点的模式。
知识主题分面树：指描述某一知识主题组织结构的层次结构树，知识主题分面树的数据主要包括分枝数据。
认知关系：指在学习过程中知识主题间由于认知需要而形成的依赖关系。
在沙盘场景中，系统利用社团检测算法，把所有主题分为不同的簇，簇与簇间的连线表示簇与簇之间的学习先后序关系。用户可以对课程的所有主题有一个概览，簇与簇之间的连线也为用户的学习提供了先后关系的指引。在课程沙盘界面点击某个知识簇后，簇内的所有主题及主题间认知关系将被展示。主题与主题之间的认知关系为用户提供了导航，可以很有效地缓解学习迷航的问题。用户可以选择自己感兴趣的主题，进而可以跳转到分面树场景，对一个主题进行具体的学习，点击返回按钮，系统返回知识簇界面，用户可以选择另外一个主题进行学习。再次点击返回按钮，系统回到沙盘场景，用户可以选择另外一个簇进行学习。再次点击返回按钮，系统返回扫描界面，用户可以扫描另一本书的封面进行学习。

Claims (10)

  1. 一种基于增强现实技术知识森林导航学习系统实现方法，其特征在于，包括以下步骤：
    S1，获取数据，所述数据包括一个课程的所有知识主题划分成的若干个簇、各个簇之间的学习先后序关系、每个簇包含的知识主题、每个簇内知识主题之间的认知关系数据、知识主题对应分面以及分面对应知识分面碎片数据；
    S2，利用步骤S1中获得的知识主题及其对应的分面数据构建知识主题分面树；
    S3，根据步骤S1中划分成的若干个簇及各个簇之间的学习先后序关系，通过扇形布局的算法生成每个知识簇的坐标数据，得到课程数据沙盘场景知识簇间场景；
    S4，将步骤S2得到的知识主题分面树模型通过节点分类依次布局方法布局在与其对应的课程数据沙盘场景知识簇内，然后生成含有课程数据沙盘场景的知识森林导航学习系统；
    S5，根据步骤S4得到的知识森林导航学习系统定制增强现实系统交互。
  2. 根据权利要求1所述的一种基于增强现实技术知识森林导航学习系统实现方法，其特征在于，利用walktrap算法将从知识森林数据库中提取到的课程信息进行社团检测划分为簇，得到不同课程对应的不同簇的数据。
  3. 根据权利要求1所述的一种基于增强现实技术知识森林导航学习系统实现方法，其特征在于，所述每个簇还包括这个簇包含的知识主题的个数；簇内知识主题之间的认知关系包括每条认知关系的起始主题名以及终止主题名；分面数据包括知识主题下的分面以及该分面对应的知识碎片数据。
  4. 根据权利要求3所述的一种基于增强现实技术知识森林导航学习系统实现方法，其特征在于，知识主题下的分面包括分面的名称、层级、所属的上级分面，知识碎片包括知识碎片的类型、内容、URL和所属分面；认知 关系包括上下位关系，学习依赖关系和因果关系。
  5. 根据权利要求1所述的一种基于增强现实技术知识森林导航学习系统实现方法，其特征在于，步骤S2中通过模型拼接的方法实现知识主题分面树的构建。
  6. 根据权利要求5所述的一种基于增强现实技术知识森林导航学习系统实现方法，其特征在于，步骤S2具体过程为：首先构建树干、树枝以及树叶的模型，根据一个知识主题包含的最多一级分面数量M，在树干模型上合适的位置添加M个固定点位；根据一个一级分面包含的最多二级分面数量N，在一级树枝上合适的位置添加N个固定点位；然后根据知识主题拥有的一级分面数量M1和每个一级分面拥有的二级分面数量N2，在树干模型的M个固定点位上随机选择M1个点位，添加M1个一级树枝模型，在每个固定好的一级树枝模型的M个固定点位中从树枝低端向顶端方向选择N2个点位添加N2个二级树枝模型，最后将树叶模型添加到每个一级树枝以及二级树枝模型的顶端，并且调整M1个树枝模型在树干模型上的平衡性，同时将树叶模型的空间坐标、空间转向和大小缩放数据按照树枝模型和树干模型从上至下进行保存，从而得到N2分枝的知识主题分面树模型。
  7. 根据权利要求1所述的一种基于增强现实技术知识森林导航学习系统实现方法，其特征在于，步骤S3具体过程为：根据每个知识簇的知识主题数量w作为权重确定其在沙盘中所占的比例，使用一个幂函数对不同簇的权重进行均衡：
    Figure PCTCN2021116052-appb-100001
    得到的w 0作为每个簇的新权重；对得到的新权重进行归一化处理：
    Figure PCTCN2021116052-appb-100002
    w i是每个簇的新权重，
    Figure PCTCN2021116052-appb-100003
    是每个簇所占的归一化权重；然后根据权重来 确定角度，给每个知识簇划分一个扇形区域。
  8. 根据权利要求7所述的一种基于增强现实技术知识森林导航学习系统实现方法，其特征在于，将从第一簇到最后一个簇之间，每2个簇的圆心坐标x i、y i相连接，得到簇与簇之间的学习先后序关系；从而完成课程数据沙盘场景知识簇间场景构建。
  9. 根据权利要求1所述的一种基于增强现实技术知识森林导航学习系统实现方法，其特征在于，步骤S4具体包括以下步骤：对一个簇内的知识主题进行分类，将只有一个连接点的知识主题归为第一类知识主题，将第一类知识主题放在知识簇内沙盘的外围；将与第一类知识主题直接相连的知识主题放在沙盘的中间层，将剩余的知识主题放在沙盘的最内部。
  10. 根据权利要求1所述的一种基于增强现实技术知识森林导航学习系统实现方法，其特征在于，步骤S5中所述增强现实系统交互包括进行放大、缩小、旋转的手势控制，以及控制返回上一场景的返回按钮。
PCT/CN2021/116052 2021-01-22 2021-09-01 一种基于增强现实技术知识森林导航学习系统实现方法 Ceased WO2022156223A1 (zh)

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
CN202110089688.5 2021-01-22
CN202110089688.5A CN112818263B (zh) 2021-01-22 2021-01-22 一种基于增强现实技术知识森林导航学习系统实现方法

Publications (1)

Publication Number Publication Date
WO2022156223A1 true WO2022156223A1 (zh) 2022-07-28

Family

ID=75858876

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/CN2021/116052 Ceased WO2022156223A1 (zh) 2021-01-22 2021-09-01 一种基于增强现实技术知识森林导航学习系统实现方法

Country Status (2)

Country Link
CN (1) CN112818263B (zh)
WO (1) WO2022156223A1 (zh)

Families Citing this family (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN112818263B (zh) * 2021-01-22 2023-08-18 西安交通大学 一种基于增强现实技术知识森林导航学习系统实现方法

Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN108108430A (zh) * 2017-12-18 2018-06-01 西安交通大学 一种基于Unity3D知识森林虚拟现实系统实现的方法
US20190087448A1 (en) * 2017-09-15 2019-03-21 Baranitharan Venkatakrishnan Method of Generating a Knowledge-Tree for Intuitive Continual Learning
CN305595640S (zh) * 2019-06-19 2020-02-07
CN111125524A (zh) * 2019-12-20 2020-05-08 四川文轩教育科技有限公司 基于知识图谱的数据分析与数据推送方法
CN112818263A (zh) * 2021-01-22 2021-05-18 西安交通大学 一种基于增强现实技术知识森林导航学习系统实现方法

Family Cites Families (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US9632654B1 (en) * 2013-03-15 2017-04-25 Shauki Elassaad System and method for augmented knowledge discovery
US9275148B1 (en) * 2011-09-23 2016-03-01 Shauki Elassaad System and method for augmented browsing and knowledge discovery
CN103049580B (zh) * 2013-01-17 2016-02-10 北京工商大学 一种层次数据的可视化方法和设备
CN106484754B (zh) * 2016-07-28 2019-08-23 西安交通大学 基于层次数据与图数据可视化技术的知识森林布局方法
CN108304519B (zh) * 2018-01-24 2020-08-18 西安交通大学 一种基于图数据库的知识森林构建方法
CN110930274B (zh) * 2019-12-02 2023-04-28 中山大学 一种基于认知诊断的实践效果评估及学习路径推荐系统和方法

Patent Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20190087448A1 (en) * 2017-09-15 2019-03-21 Baranitharan Venkatakrishnan Method of Generating a Knowledge-Tree for Intuitive Continual Learning
CN108108430A (zh) * 2017-12-18 2018-06-01 西安交通大学 一种基于Unity3D知识森林虚拟现实系统实现的方法
CN305595640S (zh) * 2019-06-19 2020-02-07
CN111125524A (zh) * 2019-12-20 2020-05-08 四川文轩教育科技有限公司 基于知识图谱的数据分析与数据推送方法
CN112818263A (zh) * 2021-01-22 2021-05-18 西安交通大学 一种基于增强现实技术知识森林导航学习系统实现方法

Also Published As

Publication number Publication date
CN112818263B (zh) 2023-08-18
CN112818263A (zh) 2021-05-18

Similar Documents

Publication Publication Date Title
US12314317B2 (en) Video generation
US10768421B1 (en) Virtual monocle interface for information visualization
US9152695B2 (en) Generation of metadata and computational model for visual exploration system
US11030445B2 (en) Sorting and displaying digital notes on a digital whiteboard
WO2018072071A1 (zh) 知识图谱构建系统及方法
US8863034B2 (en) 3D tag clouds for visualizing federated cross-system tags
JP2020024698A (ja) ナレッジグラフの生成方法、装置、機器およびコンピュータ読み取り可能な記憶媒体
US20150113018A1 (en) Interactive visual analytics for situational awareness of social media
US20140365481A1 (en) Semantic grouping in search
US20170131872A1 (en) Mobile User Interface
US12561380B2 (en) Computer-implemented system and method for analyzing clusters of coded documents
US10341272B2 (en) Personality reply for digital content
CN112818263B (zh) 一种基于增强现实技术知识森林导航学习系统实现方法
Stober et al. Musicgalaxy: A multi-focus zoomable interface for multi-facet exploration of music collections
Chen et al. Amplifying the music listening experience through song comments on music streaming platforms
CN118734120B (zh) 基于知识管理人工智能系统中的识别模型
US20240221311A1 (en) Four-dimensional spatio-temporal structure aggregation platform for metaverse information
Xu et al. Visualization digital system of digital museum based on big data technology
KR102263317B1 (ko) 속성 언어 관리 방법 및 장치
CN115480721A (zh) 媒体内容的展示方法、装置、设备及计算机可读存储介质
Hu Interest-Guided Exploration of Large Information Space
Gumtang et al. Kandinsky Meets Social Conversation: Towards Abstract Art-Inspired Visualization Abstraction for Mobile Devices
Farahbakhsh D2. 2 State of the Art Analysis Report
Speiser WorldLens: exploring world events through media
EP3465480A1 (en) Analyzing clusters of coded documents

Legal Events

Date Code Title Description
121 Ep: the epo has been informed by wipo that ep was designated in this application

Ref document number: 21920602

Country of ref document: EP

Kind code of ref document: A1

NENP Non-entry into the national phase

Ref country code: DE

122 Ep: pct application non-entry in european phase

Ref document number: 21920602

Country of ref document: EP

Kind code of ref document: A1

122 Ep: pct application non-entry in european phase

Ref document number: 21920602

Country of ref document: EP

Kind code of ref document: A1

32PN Ep: public notification in the ep bulletin as address of the adressee cannot be established

Free format text: NOTING OF LOSS OF RIGHTS PURSUANT TO RULE 112(1) EPC (EPO FORM 1205A DATED 14.02.2024)

122 Ep: pct application non-entry in european phase

Ref document number: 21920602

Country of ref document: EP

Kind code of ref document: A1