Table 1
Word Distribution per Corpus Type and Newspaper.
| NEWSPAPERS | CORPUS | TOTAL COMMENTS | TOTAL WORDS1 |
|---|---|---|---|
| MindaNews | Main | 88 | 5444 |
| Side | 25 | 941 | |
| The Mindanao Times | Main | 2 | 41 |
| Side | 0 | 0 | |
| Sunstar Davao | Main | 306 | 7350 |
| Side | 69 | 5911 | |
| Cebu Daily News | Main | 55 | 2251 |
| Side | 68 | 2488 | |
| The Freeman | Main | 186 | 8057 |
| Side | 41 | 1399 | |
| Philippine Daily Inquirer | Main | 7419 | 441517 |
| Side | 962 | 48723 | |
| Manila Bulletin | Main | 214 | 8468 |
| Side | 231 | 19273 | |
| Manila Times | Main | 433 | 31803 |
| Side | 73 | 6478 | |
| Philippine Star | Main | 1688 | 55045 |
| Side | 399 | 20123 | |
| Sunstar Cebu | Main | 136 | 5504 |
| Side | 83 | 2456 |
Table 2
Data Sources.
| FEATURE | SOURCE |
|---|---|
| object_id, message, message_proc (processed message), from_name (public page sources), created_time, newspaper | Facepager, Graph API |
| region (Luzon/Visayas/Mindanao), corpus (main/side), administration (President Benigno Aquino III/President Rodrigo Roa Duterte), year (year of posting), month_year (month and year of posting), count (word count) | Manually entered |
| lang_label (Tagalog (Filipino), English, Cebuano, Taglish, Bislish, Bislog, Other) | Computational |
Figure 1
Distribution of Languages in the Corpus.
Figure 2
Distribution of Languages per Newspaper. Normalization occurred row-wise (per newspaper).
Figure 3
Distribution of Comments per Date. The data here is skewed towards the months encompassing the Mamasapano Clash (January–February 2015) and the Marawi Siege (May–June 2017), due to the high concentration of key words.
Figure 4
Distribution of Comments per Region.
Figure 5
Distribution of Comments per Administration.