The goal of this project is to develop a better understanding of how AI and humans can effectively communicate with each other, and to develop tools and frameworks that can enhance this communication. By using taxonomic structures to classify the actions of questions that arise during interactions with AI, we can create a more comprehensive understanding of how these systems work and how they can be improved. This research has practical applications in a wide range of fields, including natural language processing, machine learning, and human-computer interaction.
https://www.researchgate.net/publication/369366694_Taxonomic_structure_of_questions_that_can_be_asked_to_AI_Artificial_Intelligence
Why is it so complicated?
Because you will only complicate the project by creating and creating taxonomy objects. which you will check against the human conceptual apparatus, which is different for each person. What amount of resources a job like yours would use, how difficult and expensive it would be to maintain.
Thank you for your response. My intention is not to complicate the incident for humans, but rather to match the taxonomic structure of the questions with the answering mechanisms for artificial intelligence. This matching can make the process of answering questions by AI systems more systematic. In particular, an approach based on this structure can help to obtain more accurate and efficient results, especially in difficulties such as determining the correct type of question and selecting the appropriate data sources. Thank you again for your valuable feedback.
Additionally, this taxonomy used in human-AI interaction can help AI systems to effectively utilize their knowledge and skills at all levels, by providing a variety of question types. This can lead to more effective results in the creation and evaluation stages, and ultimately help to obtain more accurate and reliable results.
It is better to do it not by taxonomy. In my humble opinion, it should be a system based on the construction of associative relationships (a kind of network base). And which will be able to check the data at the level of concepts and meaning.
We are still at the very beginning of this direction. And we have learned to reconcile normative reference and nomenclature information automatically. And we are just approaching the analysis of documents. But we are at the very beginning of the way. And we did all this without neural networks.
You're right, technically there's room for development in this area. However, if the primary goal of artificial intelligence is to target structures capable of human-like thinking, etc., then I believe it's possible that some classifications and autonomies inherent in the natural processes of learning could also be applicable for machine learning.
No one completely abandoned or will abandon taxonomy, but it should not be made the mainstay. And I would not use machine learning, but direct learning for such systems. That is, the way we teach children at school.
- Badges
- Science topic
- Similar topics
- Phytochemistry
- Extracts