INTERACTION DESIGN ANALYSIS AND IMPROVEMENTS OF TIKTOK’S CONTENT PREFERENCE FEATURE

• Main Author: Yolanda Bernadetta, Naomi
• Format: Final Project
• Language: Indonesia
• Online Access: https://digilib.itb.ac.id/gdl/view/66648
• Institution: Institut Teknologi Bandung

TikTok is a social media by ByteDance company where each user can create short videos by utilizing the features in it. The number of TikTok users in Indonesia is estimated to be around 38.7% of the Indonesian population. In the TikTok application, there are limitations to the content preference settings that can be made by users. The possibility of users setting up a homepage is only through selecting the category of interest when they first create a TikTok account, selecting not interested for videos that are not liked, and increasing the factors that influence the recommendation system. The problems investigated in this final project include the appearance of racially offensive content, content that is not up to date, content that does not match user preferences, content that has a negative effect on users, and content that has just been watched is lost due to For Your Page (FYP) being refreshed accidentally. The result of this final project is a high-fidelity prototype on a mobile app with the Android operating system. Based on thatA, this final project is expected to produce an interaction design for the TikTok application to overcome the problems felt by users regarding the limitations of setting content preferences in the TikTok application. The interaction design approach used is user-centered design (UCD) by centering the user as the design decision maker. This approach is used because this interaction design is based on satisfaction and emphasizes the user's view. Based on testing, it is found that the effectiveness of the design is met based on the participant's completion rate of 100% for each test. The design fulfills easy to learn based on SUS questions related to learnability of participants worth 1.73 out of 5 related to the need for help in understanding the design, 4.73 out of 5 related to ease of understanding, and 1.93 out of 5 related to the need to learn before using the design. In addition, the SEQ value shows 6.68 out of 7 in the low fidelity prototype test, indicating that all tasks are easy to learn from the first test. The design meets the easy to remember how to use based on the participants' learning curve in the first and second iterations which showed rapid improvement. The design has also met the goals helpful based on the IMI subscale usefulness/value score which shows 6.94 out of 7 and a SUS score of 95.5 in the high fidelity prototype test which shows good usability in the design.