What are we thankful for?

  4 m, 5 s

Well, now that the folks in the US have had a chance to digest their turkey, it's time to find out what we were thankful for this year.

Using Datasift's spiffy historics service, I gobbled 10% of all tweets from midnight Thanksgiving through to 23:00 that mentioned "thankful" or "grateful."  This comes out to roughly 615,000 tweets.  (!)

I then ran them through Lexalytics' text and sentiment analysis engine, Salience, but instead of wanting to find out sentiment, I was really looking to do some lexical pattern matching.  One of the key underpinnings of all of our natural language processing is the ability to do "part of speech" (PoS) tagging, and then feed those PoS tags into something called our "relationship" extraction.   Basically, this is normally used to parse for things like gene regulation, or company mergers/acquisitions activity (company a "bought" company b.)  

I created a quick pattern file with the likely parts-of-speech that someone would use, like so:

Thankful
{Company} for (?<topic> (VB_ALL)? (DT)? (PRP$)? (JJ_ALL)* (NN_ALL)+ ) 
{Company} for (?<topic> (VB_ALL)+ (DT)? (PRP$)? (JJ_ALL)* (NN_ALL)?)
 
I kinda cheated - relationships are typically between two entities, and so I just made an entity of "thankful" and "grateful" in a separate configuration file, easy peasy.  Basically all these patterns are trying to do is to match particular patterns of verbs, determinate, possessive pronouns, adjectives, and nouns.  If you'd like to read more about the pattern format we use, you can check out the documentation on pattern files here.   If you're simply curious, you're always welcome to drop us a line at that "Contact" block up there and we can run you through what we can do.  It's spiffy.
 
There's clearly some results that could be broken down further (like "the opportunity"), but for the amount of work that went into this, it's a pretty good view on what people are thinking and saying.    We're picking up about 150k out of the 600k tweets.  There were about 42k total in the "thankful for" column that covered almost all of those 600k tweets, so the pattern was pretty solid.
 
So, without further ado, here's the top 250, straight from the mouth of Twitter:

Thankful for:Count of Tweets
my family14,925
everything10,947
everyone10,630
having5,078
my friends4,203
being4,043
the people3,838
family2,759
food2,384
life2,375
god2,369
today2,351
my mom2,162
the opportunity2,072
u2,010
our gifts1,933
a lot1,867
this year1,834
everybody1,705
my best friend1,589
anyone1,345
music1,254
this man1,196
my boyfriend1,175
my parents1,152
my baby1,125
these people1,109
my best friends1,053
my life1,045
my bestfriend1,037
pizza1,037
the love951
the life928
friends923
the booty916
my amazing family853
every single one846
my brother841
this guy834
y'all738
yoga722
dat717
the fact711
people701
this girl675
hot guys670
my team658
my mother647
my wonderful family631
the friends615
...593
my dad580
the family576
jesus563
this thanksgiving563
this year...561
these guys550
my niggas545
my fam522
da519
meeting516
my brothers509
my day502
many things494
my bed488
waking486
my bestfriends477
heart466
my amazing friends460
anything448
netflix442
booty435
the ones421
leggings416
my fans416
my sister416
my girls415
my boys414
my amazing boyfriend407
my loving family405
warm socks401
my girlfriend399
these girls396
twitter393
shit380
cute butts379
the cowboys377
another day368
my health365
my sisters358
the fans354
your passion351
football344
my bros343
every single person342
weed338
the things337
someone335
getting326
alcohol325
harry321
my nigga317
girls316
country314
the food312
every person306
the man306
alot300
my girl295
the boys294
my beautiful family293
something293
the gift292
coffee287
starbucks279
cute boys276
ya275
my babies274
my momma271
your life271
my beautiful girlfriend270
my squad269
my dog267
justin266
marijuana266
wine265
pleading255
my cousin253
my grandma248
yall246
oomf243
beer236
the roof235
my cat234
the little things234
the men233
this boy231
things230
this food229
seeing228
selena228
the fam224
this holiday224
drugs222
soccer222
thanksgiving221
my lil218
the time218
living217
basketball215
your family215
mashed potatoes213
bae212
my job212
big booty211
the ability211
wifi211
my bff207
my mommy204
chuck201
my boy199
your support199
sleep198
butts197
my wonderful boyfriend197
boobs196
kanye196
my son196
boys194
my love194
my wife193
our fans192
the blessings192
my teammates190
white girls189
the day188
new relationships187
sharkeisha187
this life186
my bro183
these fools181
my bae179
this kid178
mary177
my mama175
this day173
miley171
cats170
this beautiful girl170
kissing165
josh162
baseball159
slam dancing159
sweaters159
my crazy family158
the internet157
my cousins155
the person155
paul154
my god152
my second family152
my husband150
my daughter149
every moment148
chipotle147
taco145
the amazing people145
the opportunities145
my bitches144
calvin142
nothing142
my daddy141
our troops141
this picture141
a god140
love140
my babe140
nick140
the squad139
every day137
my real friends137
my swag136
the bae136
the bitches136
black friday135
my amazing girlfriend134
these boys133
my wonderful friends132
sweatpants132
my hair131
porn131
my homies130
the teammates130
the company129
your existence129
luke128
the game128
my followers126
the best fandom126
makeup125
my moms123
my ride123
nights123
our friendship122
paper122
the girls121
my ability120
my grandparents120
the memories120
this little guy119
Categories: Analysis, Natural Language Processing, Social Media