Automated Crossword Solving

TL;DR

This paper introduces the Berkeley Crossword Solver, a system that combines neural question answering, belief propagation, and local search to automatically solve crosswords with high accuracy, outperforming previous methods and even humans.

Contribution

The paper presents a novel integrated system for crossword solving that significantly improves accuracy and includes a large dataset for future research.

Findings

Achieved 82% exact puzzle accuracy on New York Times crosswords.

Obtained 99.9% letter accuracy on themeless puzzles.

Outperformed human competitors at the American Crossword Puzzle Tournament.

Abstract

We present the Berkeley Crossword Solver, a state-of-the-art approach for automatically solving crossword puzzles. Our system works by generating answer candidates for each crossword clue using neural question answering models and then combines loopy belief propagation with local search to find full puzzle solutions. Compared to existing approaches, our system improves exact puzzle accuracy from 71% to 82% on crosswords from The New York Times and obtains 99.9% letter accuracy on themeless puzzles. Additionally, in 2021, a hybrid of our system and the existing Dr.Fill system outperformed all human competitors for the first time at the American Crossword Puzzle Tournament. To facilitate research on question answering and crossword solving, we analyze our system's remaining errors and release a dataset of over six million question-answer pairs.