Method and device for visualization of Web multimedia resource open annotation data

Abstract

The invention discloses a method for visualization of Web multimedia resource open annotation data. The method comprises the steps acquiring serialized format data treated with Web multimedia resource open annotation from a data source; analyzing the serialized format data into an RDF triple set; converting the RDF triple set into a graph abstract data structure according to the semanteme of an open annotation data model; mapping the graph abstract data structure into specific plot data matched with the specific screen visualization technology; displaying a visualization graph in a screen view area based on the specific plot data, and responding to the mouse manipulation of a user on the visualization graph. The invention further discloses a device for visualization of the Web multimedia resource open annotation data. According to the method and device for visualization of the Web multimedia resource open annotation data, the Web multimedia resource open annotation data meeting the Open Annotation Data Model Specification of the W3C can be represented visually, and effective implementation of the specification is facilitated.

Description

technical field [0001] The present invention relates to the technical field of data visualization, in particular to a method and device for visualizing open label data of Web multimedia resources. Background technique [0002] Although the current World Wide Web (Web for short) contains a large number of multimedia resources in various formats, and its data volume has been growing continuously, compared with traditional content formats such as text and numerical values, Web resources However, multimedia resources such as images, audio and video are still "second-class citizens" of the Web. The reasons for this limitation are: on the one hand, multimedia resources must be embedded in other Web resources (such as HTML pages), and the audio-visual content can only be used by Web users after the interaction and decoding of "plugin" software; , the traditional Web infrastructure is difficult to effectively support users' identification, description, discovery and access to Web m...

AI Technical Summary

Problems solved by technology

[0003] Early Web multimedia annotation was mainly realized by means of multimedia metadata standards (such as MPEG-7, Dublin Core, VRN, etc.), but this multimedia annotation method based on metadata standards (this article is called "the first type of annotation method") exists Many defects: (1) The high-level features of content descriptive metadata need to be manually created, so content producers are reluctant to label multimedia resources due to time-consuming and laborious; (2) There are too many multimedia metadata standards, and their metadata models are not consistent with The vocabulary sets are different, so it is difficult to achieve interoperability between different fields and applications; (3) Most multimedia metadata standards use XML-based (grammatical) metadata models, which cannot provide semantic-based content descriptions in essence. That is, semantic annotation cannot be realized

However, this multimedia annotation method based on Web2.0 social software (this article is called "the second type of annotation method") also has obvious defects: (1) The annotations produced by users are mainly text labels (tags) or comments (comments ), which is a kind of low-quality, uncertain, and formless metadata, which cannot be reused, let alone exchange and share across communities; (2) The annotation models of various social software are different, and they are all Closed systems that cannot effectively interact with each other, where annotation data is "locked" and not directly accessible from the Web

However, this ontology-based multimedia semantic annotation method (referred to as "the third type of annotation method" in this paper) still has many defects: (1) Various multimedia ontologies mostly provide semantic descriptions for specific multimedia metadata standards, so mutual There is still a lack of interoperability between them; (2) Various annotation frameworks use their own dedicated annotation models, which lack scalability and exchangeability, resulting in the inability to share and reuse the generated annotation data; (3) Metadata standards-based The multimedia ontology either lacks an identification mechanism for media fragments, or its media fragment identification uses a non-HTTP URI mechanism (that is, the Web is not accessible)

However, the common shortcomings of the above two annotation frameworks are: (1) They use the OAC model instead of the W3C OA data model. Therefore, they can neither generate open annotation data of Web multimedia resources, nor can open annotation data Visualization; (2) The OAC model data diagram shown in the above two documents is a schematic diagram drawn manually when the paper is written, not a visualization diagram automatically generated by the annotation framework

However, the prototype implementation has the following deficiencies: (1) the annotation client is assumed to be able to create, search and browse annotation data, but the visualization of annotation data does not give specific implementation methods and means; (2) the paper based on The annotation diagram of the OA data model is a schematic diagram drawn manually when the paper is written, not a visual diagram automatically generated by the annotation tool

[0025] After a comprehensive analysis of the above background technologies and existing related technologies, it is not difficult to draw the following final conclusion: At present, there is no open labeling data for Web multimedia resources that follows the W3C "Open Labeling Data Model Specification" in the existing technology. Visualization method and device

Images