首页 /研究 /Credibility Analysis of Robot Speech Based on Bangla Language Dialect
OTHER

Credibility Analysis of Robot Speech Based on Bangla Language Dialect

Md. Nahid Hasan, Raiyan Azim, Sadia Sharmin

发表年份
2024
引用次数
1

摘要

Our daily life now largely depends on the assistance of different machines. We perform various tasks using machines which make our lives easier. As machines are getting closer to having a perfect imitation of the human voice, it’s very important to give machines enough credibility for different cultural settings. Our research tends to find the robot speech’s credibility in persuasiveness in the context of different factors that affect credibility as well as local dialects of Bangladesh. From the previous research, we found that practical knowledge-based speech and rhetoric cues can increase the credibility of speech. It is also seen that these element’s importance can vary from culture to culture. Our research is based on previous studies where they tried to estimate robot speech credibility in different languages [1].We have performed three studies. One study is to find the effect of knowledge and rhetoric linguistics on the credibility of a robot that speaks in Bangla. Next is to find the effect of credibility based on accents. In both of these studies, standard Bangla shows the go-to option while making the speech credible. To understand why it is feaver over other accents of Bangla we conducted a third survey where we selected five Bangladeshi local dialects such as Jessore, Sylheti, Chittagong, Barisal, and formal Chalito dialect to perform our study. We assessed the credibility of the speech when the robot’s voice had accents of these given dialects while talking. Our research is based on the survey we conducted to understand the significance of accents in robot speech. Our major findings give us an idea about the most credible Bengali accent and their feedback about using local accents in robot speech. Most participants are open to using the local dialects in robot speech as a secondary option. From these results, it can be concluded that, although local dialects are welcomed by the people, they still feel uncomfortable using them for convenient use while interacting with a machine for important work. The results acquired from the experiment can be useful to improve the user experience in the context of human computer interaction across Bengali culture.

关键词

BengaliComputer scienceCredibilityNatural language processingArtificial intelligenceSpeech recognitionSyllable

相关论文

查看 OTHER 分类全部论文