NeuCLIR TREC Track 2022 Planning Meeting

[oard]

Cross-language Information Retrieval (CLIR) has been studied at evaluation forums for more than twenty years, but recent advances in the application of deep learning to information retrieval (IR) warrant a new, large-scale effort that will enable exploration of classical and modern IR techniques for this task. A new track, NeuCLIR, will be part of TREC 2022. Please join us at one of the planning sessions on December 14 at 900 EST† and 2100 EST‡ to learn more about the track and provide feedback to the organizers. Send email to neuclir-o...@googlegroups.com to receive the zoom invitation to the meetings and announcements about the track.

The purpose of the NeuCLIR track is to address the following research questions:

• What are the best neural CLIR approaches?

• How does neural CLIR compare to the combination of machine translation and monolingual IR?

• How does it compare to the strongest statistical approaches to CLIR?

• How do the resource requirements for the various approaches compare?

• Can separate ranking and reranking phases improve performance over a single-phase approach?

• What resources are most useful for training CLIR systems?

In its first year, NeuCLIR will focus on three languages: Chinese, Persian, and Russian. 

More information is available on our webpage: https://neuclir.github.io/ 

† 900 EST is recommended for Eastern North America, South America, Europe, Africa, and Western Asia.

‡ 2100 EST is recommended for Western North America, Oceania (15 Dec), Australia (15 Dec), and Eastern Asia (15 Dec).