DISTANCED LAUGHTER — MCAA 2026 READING GUIDE Yue Xing, University of Illinois Urbana–Champaign October 4, 2026. Session 6, Panel 26: Media, Technology, and Cultural Representation in East Asia. Companion page: https://yue-xing.pages.dev/mcaa/ Follow the talk Slides 2–7: /mcaa/#corpus — corpus and scope Slide 8: /mcaa/#words — selected final-model vocabulary Slide 9: /mcaa/#dictionary — Syuzhet mean and median scores Slide 10; discussion slide 15: /mcaa/#emotions — interactive primary emotion distribution Slide 11: /mcaa/#emoji — emoji incidence and composition Slides 12–13: /mcaa/#interpretation — distanced laughter and the illustrative comment Slide 14; discussion slide 16: /mcaa/#discussion — scope and checks How to read the numbers 16,683 comments across four content groups. Genre orientation describes marketed content, not viewer gender. LLM labels: 250 sampled comments per group. Excluding none_unclear gives denominators 242 long/female, 231 long/male, 244 short/female, 238 short/male. The interactive page can instead use all 250. Male-oriented short drama: mean sentiment 0.033; 107/238 = 45.0% primary critical/negative; tears-of-joy emoji 571/1280 = 44.6% of emoji occurrences. Units differ. These group-level comparisons do not establish co-occurrence within individual comments. The illustrative comment on slides 3 and 12 comes from a female-oriented short drama. The site paraphrases it and identifies its group. Short-form laughter emoji (tears of joy, grinning face with sweat, rolling on floor laughing): 2955 + 476 + 362 = 3793 of 7063 emoji occurrences, 53.7% when rounded from counts. Slide 11 reports 53.6%. Classifier lift: final format 10.4 percentage points above baseline; final genre 2.9 points above baseline. Lift is not total accuracy; vocabulary coefficients are not frequency counts. Slide 16 reports 55 label changes among 100 humor-labeled comments after emoji removal. The supplied package contains the procedure but not robustness_comparison.csv; this result is presentation-reported and not independently reproduced here. Downloads /mcaa/syuzhet_scores.csv: supplied outputs/tables/syuzhet_scores.csv /mcaa/top_words_drama_type_clean.csv: supplied final manually filtered vocabulary table; signed coefficients associate with the classifier's format category. /mcaa/classification_final_comparison.csv: supplied classifier lifts in percentage points. /downloads/distanced-laughter-analysis.csv: existing public feature release, 16,683 observations. /downloads/DATA-DOCUMENTATION.txt: existing public release documentation. All links are relative to https://yue-xing.pages.dev Scope English-language, relevance-ranked, top-level public comments on selected YouTube videos. Channels, stories, and upload contexts differ. These are descriptive comparisons, not causal effects or claims about private feelings or silent viewers. Contact: yxing16@illinois.edu