SMART CUSTOMER CARE USING AI-BASED APPROACH
Published In: 6TH INTERNATIONAL CONFEROMPUTING, COMMUNICATION AND INFORMATION TECHNOLOGY
Author(s): PASAPITCH CHUJAI , WIROOT KLAKHAENG
Abstract: Artificial Intelligence is getting as smart as human and it is capable of replacing humans in many fields, for example, in medical and in education fields or in business. One of the reasons to use AI was for customers service and care. In this study, the researcher has developed a simulation program ChatbotĀ to assist the authorities on the fishing license, by collecting some samples and information from a helpdesk center. The researcher then used this data to analyze and design responding format of the program, which was developed by Dialogflow and Python. Program efficiency and users satisfaction can be examined through an evaluation form, with is classified into five rating scales with two points of view: program design and use as well as program response and accuracy. The program design and use are focused on efficiency, feature and process, chat design and reliable analysis. The results of the program design and use can be explained as follows: practical delivery was 4.35 in ratin
- Publication Date: 29-Apr-2018
- DOI: 10.15224/978-1-63248-181-8-11
- Views: 0
- Downloads: 0
TIME SERIES MODEL FOR PREDICTING GROUND-LEVEL OZONE
Published In: 9TH INTERNATIONAL CONFERENCE ON ADVANCES IN COMPUTING, COMMUNICATION AND INFORMATION TECHNOLOGY
Author(s): KUNLAWEE MANWONG , PASAPITCH CHUJAI
Abstract: This research is a comparison for forecasting methods for ground-level ozone using ARIMA (Auto-Regressive Integrated Moving Average) and GARCH (Generalized Autoregressive Conditionally Heteroskedastic) for the forecasting of four places in Thailand, from January 2013 to August 2018. The results obtained compare between the two models above mentioned in order to find the most accurate one, considering the lowest RMSE (Root Mean Square Error) and the lowest MAPE (Mean Absolute Percentage Error). According to the experiment, the most suitable method is GARCH which is good for the 1-2 hours early forecasting.
- Publication Date: 08-Dec-2019
- DOI: 10.15224/978-1-63248-181-8-12
- Views: 0
- Downloads: 0