AutoRecLab automates building recommender system experiments from prompts
AutoRecLab: Describe the Experiment, Get the Code!
Artificial IntelligenceInformation RetrievalMachine Learning
Summary
Recommender system experiments usually take a lot of manual coding, which can be slow and prone to mistakes. The authors introduce AutoRecLab, a tool that can take simple natural language descriptions of an experiment and automatically create the necessary code to run it. This reduces errors and speeds up the research process by turning ideas directly into working experiments. AutoRecLab uses advanced language models along with verification checks to ensure the experiments work correctly.
What this means in practice
- •For machine learning engineers: Automatically generate reproducible recommender system experiments from research ideas described in natural language.
- •For software development teams: Reduce manual coding and errors in building recommender system prototypes by using a system that validates and expands experiments iteratively.
- •For e-commerce platform teams: Support rapid testing of recommendation algorithms and feedback models to enhance personalized shopping experiences.$Commercial implications: Enables quick, automated evaluation of recommendation strategies for e-commerce platforms, accelerating deployment of better user experiences.
Authors
Moritz Baumgart, Philipp Meister, Justus Krell, Michael Schmidt, Bela Gipp, Joeran Beel
Abstract
Empirical evaluation is central to recommender-systems (RecSys) research, but turning experimental designs into executable code remains a manual and error-prone task. We present AutoRecLab, a Python-based autonomous RecSys lab that automates RecSys experiments from natural-language prompts. Given a research idea, AutoRecLab derives explicit experiment requirements, builds and validates a prototype, and iteratively expands it into the requested full experiment. The workflow combines retrieval-augmented generation (RAG) for documentation lookup, static type verification, and execution-steered tree search. In our demonstration, AutoRecLab autonomously implements an explicit-to-implicit feedback conversion study. In a baseline comparison across six algorithms and three datasets, 8 of 9 runs succeed at an average cost of approx- imately $1 per run with GPT-5.4-mini.