Automating Knowledge Acquisition for Content-Centric Cognitive Agents Using LLMs (2312.16378v1)
Abstract: The paper describes a system that uses LLM technology to support the automatic learning of new entries in an intelligent agent's semantic lexicon. The process is bootstrapped by an existing non-toy lexicon and a natural language generator that converts formal, ontologically-grounded representations of meaning into natural language sentences. The learning method involves a sequence of LLM requests and includes an automatic quality control step. To date, this learning method has been applied to learning multiword expressions whose meanings are equivalent to those of transitive verbs in the agent's lexicon. The experiment demonstrates the benefits of a hybrid learning architecture that integrates knowledge-based methods and resources with both traditional data analytics and LLMs.
- Linguistics for the Age of AI. Mit Press.
- ONTOS: An Ontology-Based Knowledge Acquisition and Maintenance System. In Proceedings of the Second Workshop on Knowledge Acquisition. Banff, Canada. August.
- Hybrid ML/KB Systems Learning through NL Dialog with DL Models. In AAAI-Make Workshop on Challenges Requiring the Combination of Machine Learning and Knowledge Engineering.
- Content-centric computational cognitive modeling. Advances in Cognitive Systems.
- Learning by reading by learning to read. In International Conference on Semantic Computing (ICSC 2007), 694–701. IEEE.
- Toward human-style learning in robots. In AAAI Fall Symposium on Natural Communication with Robots.
- Language models are unsupervised multitask learners. OpenAI blog, 1(8): 9.
- TELeR: A General Taxonomy of LLM Prompts for Benchmarking Complex Tasks. arXiv preprint arXiv:2305.11430.
- Attention is all you need. Advances in neural information processing systems, 30.
- Acquisition semi-automatique du lexique. Proceedings of\normal-\\backslash\Quatri emes Journ ees scientifiques de Lyon”, Lexicologie Langage Terminologie, Lyon, 95.
- The ecology of lexical acquisition: Computational lexicon making process. In Proceedings of Euralex, volume 96.
- Chain-of-thought prompting elicits reasoning in large language models. Advances in Neural Information Processing Systems, 35: 24824–24837.
- Steps Towards Automated Knowledge Acquisition. Towards Very Large Knowledge Bases: Knowledge Building & Knowledge Sharing (KB&KS’95), 97–102.
- React: Synergizing reasoning and acting in language models. arXiv preprint arXiv:2210.03629.
- Large language models are human-level prompt engineers. arXiv preprint arXiv:2211.01910.