The 1st Interdisciplinary Workshop on Algorithm Selection and Meta-Learning in Information Retrieval (AMIR) brings together researchers from the fields of algorithm selection and meta-learning as well as information retrieval (IR) including related disciplines like recommender systems (RecSys) and natural language processing (NLP). The AMIR workshop is held on the 14th of April 2019 in conjunction with the 41st European Conference on Information Retrieval (ECIR) in Cologne, Germany.

AMIR aims at achieving the following goals:

  1. Raise awareness of the algorithm selection problem in the information retrieval community
  2. Find solutions to address and solve the algorithm selection problem in IR
  3. Familiarize the IR community with existing research and tools from the field of algorithm selection and meta-learning
  4. Identify the potential for automated algorithm selection and meta-learning for IR applications
  5. Explore if and how information retrieval techniques can be applied to solve the algorithm selection problem.

More precisely, topics relevant for the workshop include but are not limited to

  • Algorithm Selection
  • Algorithm Configuration
  • Automated Machine Learning / Automatic Machine Learning / AutoML
  • Meta-Learning
  • Neural Network Architecture Search / Neural Architecture Search (NAS)
  • Hyper-Parameter Optimization and Tuning
  • Evolutionary Algorithms
  • Evaluation Methods and Metrics
  • Benchmarking
  • Meta-Heuristics
  • Learning to Learn
  • Automated Information Retrieval (AutoIR)
  • Automated Natural Language Processing (AutoNLP)
  • Automated User Modelling (AutoUM)
  • Algorithm Selection as User Modeling Task
  • Recommender Systems for Algorithms
  • Automated Recommender Systems (AutoRecSys)
  • Search Engines for Algorithms
  • CASH Problem (Combined Algorithm Selection and Hyper Parameter Optimization)
  • Automated Evaluations (AutoEval)
  • Automated A/B Tests (AutoA/B)
  • Auto* Tools in Practice (e.g. AutoWeka, AutoKeras, librec-auto, auto-sklearn, AutoTensorFlow, …)
  • Transfer Learning, Few-Shot Learning, One-Shot Learning, …

Submission types

  1. Full Papers (max. 12 pages + references)
  2. Short Papers / Posters (max. 6 pages + references)
  3. Nectar papers (1 page + references)

Full & short papers shall present original research, novel datasets, real-world applications (demonstrations), literature surveys, or critical discussions (position paper). Nectar papers shall summarize substantial research results that were already published at high-impact journals or conferences.

All accepted submissions shall be published in the CEUR Workshop Proceedings Series.

Important Dates

Please refer to


Joeran Beel – Trinity College Dublin – School of Computer Science & Statistics – ADAPT Centre – Ireland

Lars Kotthoff – University of Wyoming – Department of Computer Science – Meta Algorithmics, Learning and Large-scale Empirical Testing Lab – USA




Email: 2019 AT