BLOG

Amsterdam's AI initiative in the spotlight

Search Engines Amsterdam

The ELLIS unit Amsterdam is proactively promoting local excellence by extending support to local initiatives here in Amsterdam which focus on different themes within the AI landscape. To date, we are funding seven Amsterdam’s initiatives on Machine Learning, and the Search Engines Amsterdam (SEA) is one of them. The SEA is the most longstanding initiative in our AI community with its first series organised in 2014. This month, we take the opportunity to spotlight the thriving initiative by asking several fundamental questions. 

What is the guiding principle of the Search Engines Amsterdam?

Information is central to everything we do as people. And increasingly, we depend on technology to connect us to information. The guiding principle underlying the Search Engines Amsterdam is that in an open democracy, ownership of technology that plays such a key role in society should be shared, between academia, industry, government, and NGO’s. That’s why SEA always strives to bring different voices to the table. Since our start in 2014, we’ve hosted well over 200 presentations with speakers from many corners of the planet and society.

How do you see the SEA different from other talks on AI? 

At Search Engines Amsterdam, regular talks are organized around algorithms and models of information retrieval, in search engines, recommender systems, or conversational assistants. Our hybrid meetups are usually held on the last Friday of the month, at Science Park Amsterdam. Usually, we have two talks in a row, one industrial, the other academic, 25+5 minutes each, no marketing, just algorithms, followed by drinks. We also host ad hoc “single shot” events whenever an interesting visitor stops and shares their work. And every now and then we have bigger events for a whole afternoon.

The main target audiences of SEA are students, researchers, developers, anyone interested in information retrieval technology and its implications. Furthermore, researchers from NLP, speech, and computer vision backgrounds may also relate to some SEA events. Before the covid pandemic, SEA meetups were physical only, narrowing the audience down to participants from Amsterdam or nearby cities. However, currently with our hybrid setting, we have a wider geographical spectrum of audience. Over 2600 members are now part of the meetup.

How does the SEA create a meaningful impact?

In SEA, we help IR enthusiasts, mainly in the Amsterdam ecosystem, to extend their knowledge of the trends and ongoing work, in academia, industry and government, on a diverse set of topics on technology to connect people to information. These talks are especially helpful for junior students to shape their research direction with a better understanding of the current challenges and existing solutions in the field. Importantly, they also help people expand their network and find partners for collaborative projects.
Search, recommendation, and conversational technology attracts interest from many angles. Interesting ideas to help solve longstanding problems in information retrieval emerge around the problem. Bringing together people eager to share ideas with people eager to find out about new ideas is a lot of fun for everyone involved… also for the organizers.

What are the crucial topics within Information Retrieval the SEA would like to further highlight?

In our regular monthly SEA talks, we usually try to have a diversified set of topics and not focus on a particular topic for several consecutive months. This is because the main objective of SEA is to give a fair overview of what is happening within the information retrieval community in the industry and academia. That being said, when planning for each month’s topic, we definitely consider emerging hot topics and try to cover them in the coming months. For example, few months back we had two talks on generative information retrieval, which is a trending topic with three related workshops at this year’s SIGIR, the leading international conference in our area.
What do you think needs to be taken up next in the upcoming research in this field? How do you envision the future of the SEA?

There are various topics that are getting attention in the information retrieval field. To name a few, conversational search and recommendation, that is, algorithms that let users search for and discover information more naturally, using spoken or written language. Then, there is personalized and context-aware search which tries to make the search results more tailored to individual users and more related to the context. As another example, fairness, accountability, transparency, and explainability are increasingly important considerations in the design and deployment of information retrieval systems, especially as these systems are becoming more complex and impactful and these factors can help to ensure that these systems are trustworthy, reliable, and ethical. With the emergence of foundational models and large language models, the information retrieval field, similar to other AI fields, has shown huge interest in applying them for different tasks such as query understanding and expansion, document ranking, question answering, improving conversational search and recommendation, and, providing data for semi-supervision.

We’ll continue to support the local information retrieval community with our monthly talks, occasionally experimenting with new topics, new formats, and new locations. We started the season with an XL event on product search and an event dedicated to information retrieval in Amsterdam. 

Interested to attend one of the SEA talks? join the meetup community here. 

Follow @irlab_amsterdam on Twitter for the latest updates.

The ELLIS unit Amsterdam thank Prof.dr. Maarten de Rijke and Ali Vardasbi for the contribution.