Skip to main content
. Author manuscript; available in PMC: 2014 Aug 11.
Published in final edited form as: Cartogr Geogr Inf Sci. 2013 Apr 19;40(2):90–102. doi: 10.1080/15230406.2013.776210

Table 3. Obesity-related themes from the LDA model.

Topic Most Likely Topic Components (bi-grams) Theme Percentage
Tweets
2, 20 Childhood obesity, school lunches, pizza vegetable, pizza
tomato, diet soda, vending machine, parent education,
tomato paste, obese kid, fight childhood, help childhood,
physical activity
Childhood
obesity and
schools
4.79
7, 13,
17
Calorie diet, calorie burn, obesity program, eating habit,
eat rule, obesity awareness, weight loss, body mass,
program prevent, eat salad, food stamp, achieve healthy,
fight obesity, health news, avoid holiday, burn calorie,
curb weight, food industry, avoid weight, food desert,
health center
Obesity
prevention
6.96
16,
21,
25,
29
Corn syrup, fructose corn, corn sugar, coca cola, food
addict, get fat, lay bed, thanksgiving dinner, video game,
holiday weight, McDonald, fries, junk food, ice cream, fat
junk, food effect, eat chocolate, food epidemic, soft
drink, grocery store, food night, super bowl, eat candy,
Chinese food
Obesity and
food habits
9.68

Notes: Column 1 shows the topic numbers, column 2 shows the most likely topic components or bi-gram terms from each topic, column 3 describes the topic themes, and the last column is the percent of tweets that used any of the bi-gram words within each theme.