The BIRDS were flying again – Bridging Gaps at CHIIR 2021

Can Data Science, Information Science, Information Retrieval and Human-Computer Interaction get together and learn from each other? Bringing together these different communities is the aim of the BIRDS workshop. While in Information Retrieval we learnt over the last decades how to meld user- and system-oriented approaches, one of the questions is how we can make use of the results and experiences gained through this long process in Information Retrieval in a broader Data Science and Data Exploration context.

The BIRDS workshop – Bridging the Gap between Information Science, Information Retrieval and Data Science – took place for the second time. While the first edition of BIRDS took place at SIGIR 2020, BIRDS was held again as a virtual event, this time in the context of CHIIR 2021. The choice of different venues is deliberate – BIRDS is an interdisciplinary workshop that tries to reach beyond the silos of Data Science (DS), Information Science (IS), Information Retrieval (IR) and Human-Computer Interaction (HCI) and tries to explore what their communities can learn from each other. It was therefore straightforward to try the more HCI- and user-oriented CHIIR conference as host, which supposedly attracts a different audience than SIGIR, where the workshop took place in 2020. In Information Retrieval, we observe the emerging and more and more successful integration of cognitive and human-centred aspects, for instance by works trying to integrate theories such as Information Foraging into their models, or just building systems that are informed by such theories. With the emergence of more data-driven methods, in particular, in the era of deep and machine learning, as well as the need of data scientists to explore, find, combine and make sense of all sorts of heterogeneous internal and external data (be it textual or multimedia, unstructured data, data streams or structured database entries), one idea is to broaden the scope of classical IR and its user- and system-oriented methods, often rooted in IS and HCI, to broader DS concepts. BIRDS 2021 was organised by Haiming Liu (University of Bedfordshire, UK), Massimo Melucci (University of Padua, Italy) and Ingo Frommholz (University of Wolverhampton, UK).

To address these interdisciplinary challenges and the novelty of the topic at hand, BIRDS 2021 consisted of a mixture of invited talks and peer-reviewed, accepted papers. Invited talks covered areas such as user discovery and exploration in digital libraries (Ed Fox), Data Science and Information Access for social research (Emanuele Di Buccio), pragmatic ecosystems for practitioners to support transparent ‘fast’ and ‘slow’ searching (Tony Russell-Rose), the exploitation of clinical data (Lorraine Goeuriot), searching enterprise content (Martin White), query-by-example and explainable text categorisation (Tobias Eljasik-Swoboda) and the design of use case diagrams, personas and user interfaces for big data analysis processes (CRISP4BigData, Kevin Berwind). The program was completed by 6 accepted papers on visualizing tweet clusters for social media mining (Morshed Adnan), users’ attention spans for interactive image retrieval (Mahmoud Artemi), semantic query construction (Sebastian Wagenpfeil, Nicholas Collis), tag recommendation for UN humanitarian data exchange (Ghadeer Abuoda) and AI-based user empowerment for visual big data analysis (Thoralf Reis). The workshop day closed with a discussion on the relevance and the future of the BIRDS workshop series (spoilers: there will likely be more!).

BIRDS 2021 attracted an estimated number of around 30 delegates, attending on and off, from different time zones. Like many other conferences and workshops these days during the pandemic, it was held online only using Zoom.  As a true cross-community event, the organisers hope to conduct future BIRDS events in person. Recordings of selected talks can be found in the BIRDS 2021 YouTube playlist.

Leave a comment

Your email address will not be published. Required fields are marked *