Gugagayanan, Jananan (2022) The Effects of a Chatbot Solving Mathematical Equations using NLP. Masters thesis, York St John University.
Preview |
Text (MSc by Research thesis)
The Effects of a Chatbot Solving Mathematical Equations using NLP - Gugagayanan Jananan.pdf - Published Version Available under License Creative Commons Attribution. | Preview |
Abstract
For many years, the field of chatbots have made an evolutionary expansion. However, chatbots have yet to be challenged against any higher-level academic knowledge. What this means is that chatbots have yet to be challenged with higher level information such as find the area of a circle and round the answer to the first significant value. The definition of higher-level academic knowledge for this project means that the experiment will be using college based (A-level) mathematical questions, this is because this level of academics provides a heavier amount of knowledge like calculus symbols for NLP to be examined with in the experiment phase of the thesis. Natural Language Processing (NLP) is a method in artificial intelligence that enables the chatbot to take a big dataset and divide the dataset into little tokens (IMB, 2023). In addition to having the dataset being broken down into pieces of tokens, the accuracy of how NLP answers mathematical questions is a vital principle that is in high demand by users who are using higher level mathematics. In this research paper, the concept of chatbots solving mathematical questions will be explored as well as analysing the accuracy levels when challenged with mathematical topics. The thesis has experimented the NLP chatbot with the following mathematical topics: trigonometry and logarithms. In addition to the NLP chatbot, there has been a pre-programmed chatbot called MATH that will also be used in the experiment to compare the accuracy of answering questions.
The experiment results confirms that NLP chatbot has a better accuracy for answering the selected topics than math library. This is based upon the number of questions each chatbot got correct when answering the questions. Trigonometry was the sole topic to have more than one question answered from both chatbots.
| Item Type: | Thesis (Masters) |
|---|---|
| Status: | Published |
| Subjects: | T Technology > T Technology (General) |
| School/Department: | School of Science, Technology and Health |
| URI: | https://ray.yorksj.ac.uk/id/eprint/8523 |
University Staff: Request a correction | RaY Editors: Update this record
CORE (COnnecting REpositories)
CORE (COnnecting REpositories)
