Effective communication becomes an important part in developing a child’s character and to prevent the violence of parents towards children. This research aims to develop a communication instrument of parents towards children. There are two communication aspects, which are the opened communication aspect and the closed communication aspect. There are two stages in developing this instrument: first, an introductory study stage to determine the prototype instrument and the second, expert validation examination stage and field examination. Based on the expert analysis results, the instrument has passed the construct validity and it is declared as fit for use. The results of the field examination show that from 34 items, 24 are declared as valid, with the validity value of ?0.3 and they are reliable as the r value= 0.84?0.7. From the expert analysis and the field examination, it is concluded that the instrument is fit for research use.

During this pandemic, most of people’s activities are carried out through digital media. Both learning and working processes are using the video-conference platform, a platform deemed effective to facilitate the needs of distance communication. One of the limitations of using video-conference lies in difficulty in understanding emotional conditions based on solely camera video. Hence, speakers generally do not know their interlocutors’ feelings related to the materials being presented. Grounded on this issue, we examined a facial expressions-based emotion recognition tool. Micro expression is one of the micro-languages of communication. Machine learning model developed in this study was Deep Learning with Convolutional Neural Network (CNN). The library that was used was Keras, this was used to recognize micro expression pattern. Additionally, OpenCV was also used for the general face recognition process. Both libraries were operated using Python programming language. The result of the micro expression test involving thirty participants detected three types of facial expression, namely joy, sadness, and anger expression. However, face recognition applied in the present study still needed some improvements, especially for anger and sadness expression. With regard to joy expression, 89% of the expression were recognized. Based on the recorded data, it is necessary to improve the recordings criteria to obtain a clearer expression.