A Feasibility Study of Available Natural Language Chatbot Technologies
Tsikora, Alona (2023)
Julkaisun pysyvä osoite on
https://urn.fi/URN:NBN:fi:amk-2023060220803
https://urn.fi/URN:NBN:fi:amk-2023060220803
Tiivistelmä
This thesis investigates the feasibility of deploying natural language processing (NLP) chatbot artificial intelligence (AI) toolkits in industrial organisation to improve service and product quality as well as enhance product functionality. The study compares features, pricing models, deployment options, and other factors to identify suitable AI solutions for commercial deployment by conducting a review of the literature and market-available toolkits.
Evaluating the available NLP chatbot AI toolkits for commercial deployment supports organizations in identifying and implementing the most suitable solutions for their needs, leading to more efficient information management and improved human-computer interactions. Rasa emerged as the most suitable solution due to its robust data privacy provisions, extensive customization options, deployment flexibility, and active development.
Alternatively, a hybrid solution combining Rasa's framework with the advanced NLP capabilities of OpenAI API was suggested. This approach offers enhanced language understanding while maintaining data privacy and customizability, though cost and additional data privacy concerns related to the OpenAI API integration must be considered. The study contributes to the broader fields of AI integration, workflow optimization, and customer experience enhancement.
Evaluating the available NLP chatbot AI toolkits for commercial deployment supports organizations in identifying and implementing the most suitable solutions for their needs, leading to more efficient information management and improved human-computer interactions. Rasa emerged as the most suitable solution due to its robust data privacy provisions, extensive customization options, deployment flexibility, and active development.
Alternatively, a hybrid solution combining Rasa's framework with the advanced NLP capabilities of OpenAI API was suggested. This approach offers enhanced language understanding while maintaining data privacy and customizability, though cost and additional data privacy concerns related to the OpenAI API integration must be considered. The study contributes to the broader fields of AI integration, workflow optimization, and customer experience enhancement.