mlsanityData to 7 Oct 2026

A Relational Tsetlin Machine with Applications to Natural Language Understanding

arXiv 2102.10952 ↗First posted 22 Feb 2021cs.CL cs.AI cs.LGcited by 8

Not on the map. Why?

Abstract (arXiv)

TMs are a pattern recognition approach that uses finite state machines for learning and propositional logic to represent patterns. In addition to being natively interpretable, they have provided competitive accuracy for various tasks. In this paper, we increase the computing power of TMs by proposing a first-order logic-based framework with Herbrand semantics. The resulting TM is relational and can take advantage of logical structures appearing in natural language, to learn rules that represent how actions and consequences are related in the real world. The outcome is a logic program of Horn clauses, bringing in a structured view of unstructured data. In closed-domain question-answering, the first-order representation produces 10x more compact KBs, along with an increase in answering accuracy from 94.83% to 99.48%. The approach is further robust towards erroneous, missing, and superfluous information, distilling the aspects of a text that are important for real-world understanding.

Key facts. Cited by 8 papers in the map (data through 7 Oct 2026). Builds on 17 of its 57 references that are in the map. First posted on arXiv 22 Feb 2021.

Builds on

17 of 57 references.

References · 17

  1. 2019BART: Denoising Sequence-to-Sequence Pre-training for Natural Language Generation, Translation, and Comprehensioncited by 5,874
  2. 2016SQuAD: 100,000+ Questions for Machine Comprehension of Textcited by 4,115
  3. 2014Memory Networkscited by 354
  4. 2015Towards AI-Complete Question Answering: A Set of Prerequisite Toy Taskscited by 309
  5. 2019TransferTransfo: A Transfer Learning Approach for Neural Network Based Conversational Agentscited by 188
  6. 2017ParlAI: A Dialog Research Software Platformcited by 139
  7. 2018The Tsetlin Machine -- A Game Theoretic Bandit Driven Approach to Optimal Pattern Recognition with Propositional Logiccited by 71
  8. 2020Turning 30: New Ideas in Inductive Logic Programmingcited by 42
  9. 2019The Convolutional Tsetlin Machinecited by 37
  10. 2018Using the Tsetlin Machine to Learn Human-Interpretable Rules for High-Accuracy Text Categorization with Medical Applicationscited by 24
  11. 2020Extending the Tsetlin Machine With Integer-Weighted Clauses for Increased Interpretabilitycited by 16
  12. 2020Measuring the Novelty of Natural Language Text Using the Conjunctive Clauses of a Tsetlin Machine Text Classifiercited by 15
5 more
  1. 2021Low-Power Audio Keyword Spotting using Tsetlin Machinescited by 11
  2. 2020Closed-Form Expressions for Global and Local Interpretation of Tsetlin Machines with Applications to Explaining High-Dimensional Datacited by 6
  3. 2014Probabilistic Inductive Logic Programming Based on Answer Set Programmingcited by 4
  4. 2020Machine Reasoning Explainabilitycited by 4
  5. 2019The Weighted Tsetlin Machine: Compressed Representations with Weighted Clausescited by 3

Built on by

Cited by year

Citing papers per year: 2023: 5; 2024: 1; 2025: 2.

8 papers.

  1. 2023Building Concise Logical Patterns by Constraining Tsetlin Machine Clause Sizecited by 14
  2. 2023TMComposites: Plug-and-Play Collaboration Between Specialized Tsetlin Machinescited by 4
  3. 2023An FPGA Architecture for Online Learning using the Tsetlin Machinecited by 1
  4. 2024TPFL: Tsetlin-Personalized Federated Learning with Confidence-Based Clusteringcited by 1
  5. 2025Uncertainty Quantification in the Tsetlin Machinecited by 1
  6. 2023Learning Minimalistic Tsetlin Machine Clauses with Markov Boundary-Guided Pruningcited by 0
  7. 2023Efficient Data Fusion using the Tsetlin Machinecited by 0
  8. 2025A Methodology for Transparent Logic-Based Classification Using a Multi-Task Convolutional Tsetlin Machinecited by 0