The CUbRIK histoGraph Factsheet
0 likes245 views
histoGraph was developed as part of the EU-funded CUbRIK project to create an interface for accessing historical sources and discovering links between entities. It builds a social graph of people in photos of European integration history by having humans and AI work together to identify faces, which are then linked based on co-occurrence. Users can interact with the graph to explore connections between individuals and supporting documents. The system represents the complexity of truth in the humanities by allowing multiple answers to identity questions and facilitating discussion between experts.
The CUbRIK histoGraph Factsheet
- 1. histoGraph and the CUbRIK project histoGraph was developed by the CVCE and partners as part of the European research project CUbRIK, funded by the European Commission’s FP7 programme. CUbRIK incorporates computer science research, the design of human-computation tasks, data visualisation, social engineering and the humanities. histoGraph offers an effective interface that enables users to access collections of historical sources and discover links among and entities within them. In its current form, histoGraph creates a social graph of people who appear in photos related to the history of European integration. This graph is automatically enriched with relevant sources. To build it, the faces in the photos first need to be identified. This work is shared by humans and machines, and the input from one improves the performance of the other. histoGraph demonstrates new approaches that encourage the public to become involved in humanities research and facilitate exchanges between users. histoGraph is based on a curated collection of more than 3 000 images which represent the main events and players in the history of European integration. This image collection is managed and hosted by the CVCE. To prepare the photos for the graph, an image indexation pipeline is used to detect individual faces in the photographs. A crowd of ‘clickworkers’ with no specific training double-checks whether the algorithms have correctly detected all the faces and whether they have missed any. The next step is to identify the faces. An automatic face recognition process is triggered to associate each of the faces with a list of ten possible identities. This list of candidates is then disseminated via Twitter or email to a crowd of experts who vote and comment on their preferred identity. The image metadata, for example people’s names and the time and place an image was taken as well as contextual information about associated historical events, can be reviewed by expert users and delegated to a crowd of specialists on the history of European integration for review. Based on the co-occurrence of people in images, a social graph is created to link individuals with each other. The more times people appear together in an image, the stronger the connections become. Users can interact with the History of Europe social graph in different ways, for example filtering the graph using a timeline, selecting a network of individual players and exploring which documents connect two people. Crowd discussion and a new representation of truth Another challenge for histoGraph and digital humanities in general is the notion of truth. In the humanities the concept of truth is highly complex: people might change careers, their home countries might be renamed or they might choose to go by Web interface of the histoGraph application
- 2. Centre Virtuel de la Connaissance sur l’Europe Château de Sanem | L - 4992 Sanem G.-D. Luxembourg Tél. +352 59 59 20-1 | Fax +352 59 59 20-555 | cvce@cvce.eu The CUbRIK project has received research funding from the European Union, Grant FP7-287704 different names. This means that at different points in time there might be more than one ‘true’ answer to the simple question ‘Who is this?’. In order to represent the discursive nature of truth in the humanities, histoGraph makes use of a community-driven tool to answer questions. Users are given the opportunity to answer questions themselves and thus benefit from the knowledge within the crowd of experts. The system allows for more than one answer and offers users the possibility to vote, thereby stimulating discussion by allowing them to suggest and defend different opinions. To learn more about CUbRIK and the modules which have been developed during the project, visit : http://www.cvce.eu/digital-humanities http://www.cubrikproject.eu/ A Luxembourg public undertaking with an international outlook The CVCE is a public undertaking under the aegis of the Luxembourg Ministry of Higher Education and Research. It is based at the Château de Sanem. The Centre carries out its activities with the help of a multinational team and a number of visiting researchers, and though partnerships with external organisations. It is a member of various international networks (DARIAH, DHBenelux, UACES, the RICHIE network, AFHé, NeDiMAH, DHd, etc.). It works in close cooperation with specialists in European studies and ICT, other research centres and universities, particularly for competitive projects.