|
Evaluating the ethical nature of utterances generated by a conversational agent Ovidiu Constantin Gheorghe, Ştefan Trăuşan-Matu |
1-12 |
|
SmartMeal: A scalable catering application Mohamed-Ata-Radu Abu Ras, Paul Stefan Popescu, Mihai Mocanu, Marian Cristian Mihăescu |
13-22 |
1 University Politehnica of Bucharest
313 Splaiul Independentei, Bucharest, Romania
2 Institutul de Cercetări în Inteligenţa Artificială
Calea 13 Septembrie nr. 13, Bucureşti
3 Academy of Romanian Scientists
Splaiul Indpendentei 54, Bucharest, Romania
Abstract. The ethical evaluation of utterances in a chat using artificial intelligence (AI) is not an easy task and a decision cannot be drawn as easily as for hate speech or other types of offensive behavior. This paper aims to scratch the surface of “ethical evaluation” tasks, making use of a piece of AI, a software technology that nowadays is used in almost everything. Using deep neural networks, three experiments were performed trying to evaluate the ethical nature of utterances generated by an AI conversation agent. The approach focuses on classifying utterances as being ethical or not. The corpus used for the experiments is “ETHICS”, which was developed by Hendricks et al. (2021). The results of the experiments were compared to those presented in the above-mentioned paper.
Keywords: Ethical behavior, Ethics; Artificial Intelligence, Natural Language Processing, Neural Networks, Conversational Agents
Cite this paper as:
Gheorghe, O. C., Trăuşan-Matu, S. Evaluating the ethical nature of utterances generated by a conversational agent.
International Journal of User-System Interaction 15(1),
1-12, 2022.
University of Craiova, Department of Computers and Information Technology
Blvd. Decebal nr. 107, RO-200440, Craiova, Romania
Abstract. In this paper, we present the design and development and a demo of SmartMeal, an application designed for an I.T. company that allows employees to order their meals for lunch. The application's audience is not limited and can be easily used by any company that regularly demands meals. This application uses a mix of frameworks, programming languages, and technologies to provide all the required functionalities and run on desktop computers and mobile devices. Another benefit of this application is that it was designed and implemented to be scalable for more users. The interface of the application was designed to be easy to use and let users order meals fast and efficiently.
Keywords: User interface, User Experience, Application design
Cite this paper as:
Abu Ras, M.-A.-R., Popescu, P. S., Mocanu, M., Mihăescu, M. C. SmartMeal: A scalable catering application.
International Journal of User-System Interaction 15(1),
13-22, 2022.