Open Access

Maik Fröbe, Christopher Akiki, Timo Breuer, Thomas Eckart, Annemarie Friedrich, Lukas Gienapp, Jan Heinrich Merker, Martin Potthast, Harrisen Scells, Philipp Schaer, Benno Stein

• Many universities offer information retrieval (IR) courses with different specializations as part of their computer science or information science programs. Student involvement and collaboration in these courses can increase engagement in the course and improve learning outcomes. We report on our first steps towards creating synergies between information retrieval courses at four German universities by conducting a shared task assignment with a document collection combining the IR and ACL anthologies, which we enrich with relevance judgments. We prepared two versions of this collection. First, a minimal test collection with 100 documents for which students can manually obtain results from intermediate results. Second, a more extensive test collection of 126 958 documents is used in a shared task setup, where students create topics, relevance judgments, and retrieval systems. The shared task setup therefore covers a broad spectrum of applied IR research. Our collaborative teachingMany universities offer information retrieval (IR) courses with different specializations as part of their computer science or information science programs. Student involvement and collaboration in these courses can increase engagement in the course and improve learning outcomes. We report on our first steps towards creating synergies between information retrieval courses at four German universities by conducting a shared task assignment with a document collection combining the IR and ACL anthologies, which we enrich with relevance judgments. We prepared two versions of this collection. First, a minimal test collection with 100 documents for which students can manually obtain results from intermediate results. Second, a more extensive test collection of 126 958 documents is used in a shared task setup, where students create topics, relevance judgments, and retrieval systems. The shared task setup therefore covers a broad spectrum of applied IR research. Our collaborative teaching initiative can help students learn from their peers locally and across universities. The cross-university setup means that institutes and degree programs with different academic backgrounds are involved, which leads to a broad spectrum of perspectives in the construction of topics and also in system development.…