⌛Gentle reminder: #TREC2023DL Track run submission deadline is **JULY 31**. Get your run submissions in before the deadline! #TRECDL
Guidelines and more info: https://t.co/pyRUKYs87k
Announcing: Test queries for the TREC 2023 Deep Learning Track are now available, along with top-100s for (optional) use in reranking. #TRECDL#TREC2023DL@MSMarcoAI https://t.co/ChL846qTsr
Announcing: Deep Learning Track at TREC-2019. Coordinators: Nick Craswell, Bhaskar Mitra, Emine Yilmaz, Daniel Campos. Large training data! Two datasets! Guidelines: https://t.co/0oNuJgaHPX
The Deep Learning track for #TREC2019 will include a DOCUMENT RANKING task and a PASSAGE RANKING task—both with large training datasets. Check out the full announcement! #TRECDL2019
In the meantime, if you have any feedback or suggestions, we would love to hear them! Also, if you are working on a paper for @TheWebConf/@sigir2019/etc. and looking for a large dataset for training/evaluation, please feel free to grab the data from https://t.co/Bu17ZAAzDK.
We have released the passage collection and the queries on https://t.co/Bu17ZAAzDK where we will also be hosting a passage re-ranking challenge. Leaderboard and further announcements coming soon! See updated paper: https://t.co/24D6mpn6Ay.
#TREC2019 will feature a new Deep Learning track for benchmarking IR models that require large training datasets organized by @nick_craswell, Emine Yilmaz, Daniel Campos, and me. The task will be passage retrieval based on @MSMarcoAI data containing real Bing queries.
The December 2017 issue of the SIGIR Forum includes a report on the Second SIGIR Workshop on Neural Information Retrieval (Neu-IR’17). Available online now!
https://t.co/3f0UVYqh0P https://t.co/YrvphnVuew
The December 2017 issue of the SIGIR Forum is online! https://t.co/RZuLw1HezQ #IR#SIGIR As usual a healthy mix of articles, workshop/conference reports and dissertation abstracts, plus the SIGIR test of time award winner of 2017!
MatchZoo is a toolkit for deep text matching, which contains a number of state-of-the-art matching methods, such as DRMM, MatchPyramid, MV-LSTM, aNMM, DUET, ARC-I, ARC-II, DSSM, and CDSSM, designed with a unified interface. Check: https://t.co/pemVjTxl0i #NeuIR
It was fun! Thanks for the invitation! Q: will IR survive the neural wave? A: Yes it will, as long as we don't forget our IR fundamentals. https://t.co/wijsFsSDDk