An Exploratory Study of Automated Timeline Generation within the Open Government Act (Woo)

Abstract

Gaining oversight from a large dataset is a challenging task. Especially within the Woo process where documents have to be selected without looking into them. This thesis investigates the use of timeline visualisation software within the Dutch Woo request process, aiming to increase its effectiveness by organising the data and providing context for documents. The study involved first the mapping of the distinct parts of the Woo, process and its stakeholders. Using this process map, two methods are established that could enhance the process of request specification and document selection. The process specification process focuses on finding out what information the requester is looking for through the use of topic extraction. Document selection focuses on selecting documents by placing them in context within timelines. Semi-structured interviews were conducted with 10 participants. The participants were presented with the proposed methods and insights were gathered through the use of a qualitative analysis. The analysis showed positive results on how the Woo process could benefit from the implementation of the proposed methods. Participants stated that the solutions provide an overview and structure, and allow a conversation topic by providing context. However, they emphasised that these solutions should only be used as tools and not as leading. Therefore, the structured methods could be implemented in software to effectively, efficiently and transparently improve the request specification, and facilitate the process of selecting relevant documents.