Using Wiki Data to Augment Historical Data Visualizations
Historical data is complex and dense due to its dependence on context. It must have breadth, as all events are set in a world stage, and depth, since local circumstances provide necessary detail. In-depth historical research requires the analysis of material evidence and bibliography which might not be properly contextualized. The construction of historical datasets provides for local context and can guide research within its scope but adding global context can easily become burdensome. One approach to this problem is using public wiki data from Wikipedia or other important wiki sources such as geschichtewiki to augment historical datasets. By using common elements such as dates, people, and events as connecting points it is possible to fill information gaps in historical data and provide a richer contextual fabric for exploration.
The aim of this project is exploring how public wiki data can be used to improve the visual analytics of historical data. Some of the proposed tasks and research challenges include:
- Consolidating common elements between different wiki sources and historical datasets
- What common historical elements can be used? Dates, people, places, etc.
- Can different wiki implementations be used without resorting to scraping/mining?
- Augmenting visualizations with information from wiki sources. E.g,:
- Displaying relevant global events on timelines for context
- Displaying extra dimensions not available in the dataset
- Fill missing data from dataset when possible
- Guiding the exploration of historical datasets. E.g,:
- Using metrics such as the connectedness or accesses of the wiki article to discover trends and patterns
- Suggest relations between objects based on wiki links
There are many projects who already use wiki data as main source of historical data for visualizations. They can serve both as inspiration and starting point for this topic. Some interesting links include:
- The wikidata website has a page with interesting vis projects using it
- Seealso has a curated collection of vis using wikidata
- The Pudding publishes interactive journalistic essays using data from wikidata
- This paper has an interesting perspective in how to use wikipedia as a repository of collective memory
Furthermore, this topic will certainly involve using SPARQL, but previous experience is not required.