SPICE logo

Meet SPICE Work Packages: Modelling and Recommendation

Today it's time to introduce you to SPICE Work Package 3: Modelling and recommendation , whose main objective is analysis of citizen curation activities , development of tools for semi-automated semantic annotation of curatorial products and development of tools to support the exploration of users' interpretations on museum objects.

WP3 is the collaborative, joint effort of three partners : Universidad Complutense de Madrid (Spain), CELI SRL (Italy) and The University of Haifa (Israel). In WP3 we are facing an interesting challenge of developing a flexible and generic user and community modeling framework that will support the diverse requirements of multiple stakeholders (SPICE case studies) and will allow easy and quick bootstrapping as well as ongoing, long term support and continuous updates based on reasoning on user generated content provided by individual visitors throughout their (recurrent) visits. It further aims to develop tools to support the exploration of interpretations and museum objects and their associated community models and plans to develop a recommender system to suggest alternative perspectives within and across communities to support social cohesion and understanding.

In the first year of the project, an initial user model and community model data structures were defined and implemented , as well as APIs for accessing the data. In addition, a semantic multilingual annotator has been implemented, supporting English, Finnish, Hebrew, Italian and Spanish. It consists of a natural language processing pipeline that performs sentiment analysis, emotion detection and entity linking. An initial version for the set of tools oriented to configure, detect, visualize, understand, explain, and navigate through citizen communities has also been developed. The tools help citizens to build representations of themselves within groups (social bonding) and facilitate understanding and perspective taking across groups (social bridging). We also started designing the social recommender.

We coordinated with WP2 the info to be provided to the User Model and the use of the recommendations. The coordination with WP4 was around the linked data hub and how to store the user and community models in it. With WP5, discussion was made as to how user and community models can be reflected in the user interface. We coordinated with WP6 the use of the same vocabularies and that the developed ontology reflects the structure of the user and community models. Finally, we met with the case studies (WP7) to understand their needs and use of the user and community models and implications for the planned recommender.

A little reflection on what SPICE brings

In Israel; in recent radio interview Mrs Reem Younis, who founded the first Israeli-Arab Hi-tech company Alpha Omega Engineering 30 years ago, had this to say about going forward in light of the recent conflicts, which resounds well with the SPICE methodology: "There are certain narratives, you need to hear mine, and I need to hear yours. We don't need to agree. However, we need to empathize with your pains, and you need to empathize with ours. This is already occurring in small groups of 20-30 people." "This is a project we hope will in the end have a real impact."

"Tolerance and being open to diversity of opinions, even those that are opposite to your own and willing to listen, and accept the fact that people are different and may have different views is important in our multicultural societies where everywhere we have a mix of religions, nationalities and ethnicities. SPICE aims at encouraging tolerance, as a way to promote social cohesion. WP3 aims at providing the means for representation and reasoning on this diversity and enabling to share and present this diversity to visitors to cultural heritage sites". Tsvi Kuflik, The University of Haifa

Subscribe to our newsletter

All the latest content, recommendations and news in your inbox!

Accept Reject
This website collects only cookies that are essential for analytics purposes. More details