Enriching Cultural Heritage Metadata through Gamified Crowdsourcing (Bachelor Thesis, Ongoing)

Author

Samuel Gasser

Description

Numerous images from historical related datasets are poorly annotated, for instance, images may be missing tags, categories, and spatio-temporal information. This is mainly due to lack of accurate information. Incomplete cultural data lose their relevance in data searching and consequently their value. Such an issue is present in the datasets used in the City-Stories project.
The goal of this thesis is to develop game-based annotation interfaces that will utilize user contribution to enrich the image metadata. This work will address the annotation of image metadata terms (tags, categories, location, and year) in four scenarios:
(i)    “AnnotationCompetition” – UI that allows users to tag images in an online and offline mode.
(ii)    “AnnotationValidator” – UI for validating automatically generated image labels and categories.
(iii)    “LocationFinder” – UI with a map where users can test their knowledge about places and try to find the location for which the image is about.
(iv)    “YearFinder” – UI with a timeline where users can guess the year of the image.

The development of user interfaces will be based on gamification which is a popular incentive to increase motivation and engagement in crowdsourcing. For each annotation, users earn points which are calculated differently for the four use cases.

Start / End Dates

2020/05/18 - 2020/08/17

Supervisors

Research Topics