Dr. Antonius Rachmat Chrismanto, a lecturer at Universitas Kristen Duta Wacana (UKDW) in Yogyakarta, developed a deep learning model specific to detecting spam comments on Indonesian social media platforms. The innovation in his approach involved utilizing posts and comments as paired data during the training process, a method showcased in his dissertation titled “Spam Content Detection Model on Social Media Through Post-Comment Pair Approach Using LSTM-CNN Based Attention and Emoji Features.”
Anton recently attained his Doctoral degree in Computer Science after completing his Ph.D. studies at Gadjah Mada University (UGM) in 3 years and five months, achieving a GPA of 4.0 and receiving cum laude honors. His academic journey began in September 2020 during the COVID-19 pandemic, focusing on researching spam detection on social media—a subject he has been dedicated to since 2017. The choice of this topic stems from the prevalence of spam content, including link spam and comment spam, especially on high-profile social media accounts. Emphasizing spam comments, which can distort information flow and mislead readers, he conducted a case study on Instagram due to its widespread occurrence on public figure accounts, notably among Indonesian celebrities.
Anton currently serving as the Head of the Institute of Academic Development and Learning Innovation (LPAIP) at UKDW, elaborated, “This research also incorporates emojis, a common feature in social media but rarely explored in previous studies. Besides developing the model, this research led to a browser extension and web-based services tailored for Instagram on desktop.”
For the publication, Anton has produced one national Sinta 2 journal article, one international journal article, two computer program copyrights, and a Scopus Q1-indexed international journal article. Another international journal article is under review for inclusion in a Q2-indexed journal. [humasukdw/Eng.drr]