#Tag
BREAKING: There will be an #IRRJ paper session at the 48th European Conference on Information Retrieval #ECIR2026 on 1 April in Delft, The Netherlands. https://ecir2026.eu
BREAKING: There will be an #IRRJ paper session at the 48th European Conference on Information Retrieval #ECIR2026 on 1 April in Delft, The Netherlands. https://ecir2026.eu
This year, the #IRforGood track will be a part of the main #ECIR2026 conference program:
https://ecir2026.eu/calls/call-for-ir-for-good-papers
https://bhaskar-mitra.github.io/posts/2025/09/01/what-is-ir-for-good/
I registered for #ECIR2026. Cool to see they have an online social platform for the conference attendees!
It has a long list of the usual suspect things to fill in: Affiliation, Current Position, Twitter, Facebook, LinkedIn, Github, Google Scholar, 🤮 ...
... so, I filled in:
Affiliation: Radboud University
Current Position: Find me on Mastodon: https://idf.social/@djoerd 😎
#ECIR2026
To be presented at #ECIR2026 by @gijs; with @arjen: "Open Web Indexes for Remote Querying"
https://djoerdhiemstra.com/2026/open-web-indexes-for-remote-querying/
I registered for #ECIR2026. Cool to see they have an online social platform for the conference attendees!
#ECIR2026 notifications part two:
2. Tutorial "Reasoning for IR & IR for Reasoning" tutorial led by @mhoveyda with Maarten de Rijke, Panagiotis and myself:
Many real-world information needs demand more than semantic relatedness: enforcement of logical constraints, multi-step inference, and synthesis of multiple pieces of evidence. Addressing these requirements is, at its core, a problem of reasoning.
Come to the tutorial if you want to learn more!
2/4
#ECIR2026 notifications part three:
3. Full paper "OrLog: Resolving Complex Queries with LLMs and Probabilistic Reasoning" with @mhoveyda, Jelle Piepenbrock, MdR and @fhasibi.
OrLog is a neuro-symbolic retrieval framework that combines LLM with probabilistic logic programming to resolve natural language queries with logical constraints by decoupling predicate-level plausibility estimation (=LLM) from logical reasoning (=ProbLog).
Revisit Van Rijsbergen's vision of logic for IR!
3/4
#ECIR2026 notifications part four:
4. WOWS returns one more year at ECIR 2026:
The Third International Workshop on Open Web Search (WOWS)
Organizers (from #OpenWebSearchEU):
Laura Caspari, Maik Fröbe @maik_froebe, Sebastian Heineking, Michael Granitzer, Gijs Hendriksen @gijs, Djoerd Hiemstra @djoerd, Martin Potthast @potthast, Arjen de Vries @arjen, and Saber Zerhoudi
(Quite some idf.social crowd :-))
Expect the CfP soon! We offer a new Web Test Collection drawn from Curlie.
4/4
#ECIR2026 notifications part two:
2. Tutorial "Reasoning for IR & IR for Reasoning" tutorial led by @mhoveyda with Maarten de Rijke, Panagiotis and myself:
Many real-world information needs demand more than semantic relatedness: enforcement of logical constraints, multi-step inference, and synthesis of multiple pieces of evidence. Addressing these requirements is, at its core, a problem of reasoning.
Come to the tutorial if you want to learn more!
2/4
#ECIR2026 notifications part three:
3. Full paper "OrLog: Resolving Complex Queries with LLMs and Probabilistic Reasoning" with @mhoveyda, Jelle Piepenbrock, MdR and @fhasibi.
OrLog is a neuro-symbolic retrieval framework that combines LLM with probabilistic logic programming to resolve natural language queries with logical constraints by decoupling predicate-level plausibility estimation (=LLM) from logical reasoning (=ProbLog).
Revisit Van Rijsbergen's vision of logic for IR!
3/4
#ECIR2026 notifications part two:
2. Tutorial "Reasoning for IR & IR for Reasoning" tutorial led by @mhoveyda with Maarten de Rijke, Panagiotis and myself:
Many real-world information needs demand more than semantic relatedness: enforcement of logical constraints, multi-step inference, and synthesis of multiple pieces of evidence. Addressing these requirements is, at its core, a problem of reasoning.
Come to the tutorial if you want to learn more!
2/4
#ECIR2026 notifications were friendly to me 🤗
1. Full paper "Open Web Indexes for Remote Querying" with @gijs and @djoerd.
Can we let ppl query the Terabytes of Web Index we collect in #OpenWebSearchEU in new ways, making good use of Parquet, S3, DuckDB?
Turns out the answer is a big YES!
Pre-print of the paper w/ code coming soon!
1/4
