Wikxhibit: Using HTML and Wikidata to Author Applications that Link Data Across the Web

Title of the Paper

Wikxhibit: Using HTML and Wikidata to Author Applications that Link Data Across the Web

Paper Information

• Research Field: Data Integration and Semantic Web, Data-Driven Application Development for Non-Programmers
• Keywords: Wikidata, Web APIs, Mashup, Data Integration, Semantic Web, HTML, Data Visualization, Data Query, Data Linking, Database

Research Background and Problem Statement

• Identified Issues or Challenges:

  • Data is distributed across numerous websites, making it difficult to unify and integrate due to diverse formats, structures, and protocols.
  • Existing tools and methods (e.g., Semantic Web, Linked Open Data) have not become mainstream, as programmers need to write custom code to achieve cross-website data integration.
  • For non-programmers, creating data-driven web pages based on Wikidata is nearly impossible.

• Significance:

  • Data integration can drive applications in data-driven fields such as art galleries and political data visualization. However, existing methods are highly unfriendly to non-technical users, leading to high failure rates despite strong demand.

• Research Motivation and Related Work:

  • Previous tools (e.g., MashMaker, Mavo) have reduced some of the complexity of data integration but still require advanced programming skills and lack a unified data model.
  • Wikidata, as a structured database, can serve as a "data catalog" to reference related data across multiple websites. However, there are currently no efficient tools enabling ordinary users to leverage this data.
  • This study aims to bridge the gap between technical capabilities and practical needs by simplifying tools for cross-website data integration and visualization, empowering users without programming experience.

Proposed Solution

• Proposed Solution:

  • Wikxhibit: A JavaScript-based HTML extension library that allows users to create data-driven web pages using HTML without additional programming.
  • Combines four components: Wikidata as a "unified index" for cross-website data, Shapir for standardized API data access, Mavo for interactive visualization, and HTML for rapid page generation.

• Innovative Features:

  • Utilizes Wikidata as a "Universal Join Table" to link data across multiple third-party APIs, achieving unified entity representation.
  • Integrates Shapir to support standardized API data access and employs Schema.org for a unified data model.
  • Non-programmers can display content from multiple data sources by simply adding HTML attributes based on Wikxhibit.

• Implementation Steps and Key Technologies:

  1. Use the Wikxhibit library to parse special attributes in HTML, querying Wikidata's SPARQL endpoint to retrieve structured data.
  2. Connect to third-party APIs using external identifiers and retrieve relevant data through the Shapir library.
  3. Generate dynamic web pages using Mavo for data visualization, supporting multilingual queries and cross-website entity integration.
  4. Conduct user testing and provide code generators to further simplify the creation process.

Research Outcomes

• Specific Outcomes:

  • Developed the Wikxhibit tool, enabling data integration and visualization based on Wikidata and its cross-referenced APIs.
  • Supports non-programming users in quickly creating applications using only HTML, with an average completion time of 5-15 minutes.
  • Automatically links entities from different website data sources, such as artists, artworks, and related content, into a single page for display.

• Advantages Compared to Existing Solutions:

  • Compared to Semantic Web and other Mashup tools, Wikxhibit lowers the technical barrier while achieving cross-source data integration.
  • Users do not need to write complex SPARQL queries or code; simple HTML is sufficient for data visualization.
  • Provides real-time display and efficient integration of various data types.

• Experimental and Evaluation Results:

  • In user studies, 12 participants successfully completed creation and modification tasks, with non-programmers achieving satisfactory results.
  • During free tasks, participants easily displayed data from multiple websites, such as movie information and political party data.
  • Suggestions for improvement include stronger error notification mechanisms and enhanced tools for searching Wiki attributes.

• Limitations and Future Directions:

  • Currently supports only data reading functionality, with no data writing capabilities yet implemented.
  • Conflicts between different data sources remain unresolved, such as inconsistent descriptions of the same entity across APIs.
  • Future research could explore extending support for Schema.org wrappers and further optimizing the user experience for non-technical users.
  • Plans to collaborate with educational institutions and individual projects to promote the practical application of Wikxhibit.