Mood Analysis
After exploring the details on the data we wanted to try something completely new to us as GI - Students. So we thought (with the help and inspiration of Dr. Euro Beinat) of trying some mood-analysis.
After exploring the details on the data we wanted to try something completely new to us as GI - Students. So we thought (with the help and inspiration of Dr. Euro Beinat) of trying some mood-analysis.
The used methods can basically be applied to any set of message data (like in our case twitter data), we decided to keep it rather small and try it out on the filtered tweets with financial topics. In order to analyse the mood in the tweets containing the money-related terms, around 70 words (and some common smilies) expressing feelings were used. The table containing the relevant tweets is then queried with each of the expressions and the number of rows resulting from the query is counted.
The words are either classified as positive or negative. Positive words are assigned a value of 1 to signify the feeling, while the negative words are assigned a value of -1. The weight of each word is obtained by dividing the number of rows containing that particular word by the total number of rows resulting from all the terms expressing feeling.
The words can then be assigned a value to signify the strength of the feeling, either very positive or very negative (2), or just positive or negative (1). Therefore the moods can be obtained by multiplying the feeling value assigned to the word (simply 1 or -1) by the weight of the word and also by the strength of the word. Since there is uncertainty in the strength variable, the moods are computed in booth variants:
Term
|
Number
|
Positive
|
strength
|
Weight
|
Mood
|
Mood
(without
strength)
|
Happy
|
1247
|
1
|
2
|
0.03555
|
0.0711
|
0.03555
|
Excited
|
359
|
1
|
2
|
0.01024
|
0.02048
|
0.01024
|
Great
|
2032
|
1
|
1
|
0.05793
|
0.05793
|
0.05793
|
Thankful
|
27
|
1
|
1
|
0.00077
|
0.00077
|
0.00077
|
Merry
|
103
|
1
|
1
|
0.00294
|
0.00294
|
0.00294
|
Elated
|
63
|
1
|
2
|
0.00180
|
0.0036
|
0.0018
|
Satisfied
|
19
|
1
|
1
|
0.00055
|
0.00055
|
0.00055
|
Jubilant
|
0
|
1
|
1
|
0.00000
|
0
|
0
|
Fortunate
|
123
|
1
|
1
|
0.00351
|
0.00351
|
0.00351
|
Thrilled
|
5
|
1
|
1
|
0.00015
|
0.00015
|
0.00015
|
Glad
|
473
|
1
|
1
|
0.01349
|
0.01349
|
0.01349
|
Optimistic
|
12
|
1
|
1
|
0.00035
|
0.00035
|
0.00035
|
Wonderful
|
118
|
1
|
1
|
0.00337
|
0.00337
|
0.00337
|
Ecstatic
|
4
|
1
|
2
|
0.00012
|
0.00024
|
0.00012
|
Love
|
5338
|
1
|
1
|
0.15216
|
0.15216
|
0.15216
|
Satisfactory
|
1
|
1
|
1
|
0.00003
|
0.00003
|
0.00003
|
Gr8
|
48
|
1
|
1
|
0.00137
|
0.00137
|
0.00137
|
I like
|
309
|
1
|
1
|
0.00881
|
0.00881
|
0.00881
|
Confident
|
21
|
1
|
1
|
0.00060
|
0.0006
|
0.0006
|
Calm
|
169
|
1
|
1
|
0.00482
|
0.00482
|
0.00482
|
Vital
|
88
|
1
|
1
|
0.00251
|
0.00251
|
0.00251
|
Sure
|
1761
|
1
|
1
|
0.05020
|
0.0502
|
0.0502
|
Enjoy
|
499
|
1
|
1
|
0.01423
|
0.01423
|
0.01423
|
Celebrate
|
46
|
1
|
1
|
0.00132
|
0.00132
|
0.00132
|
Rich
|
1439
|
1
|
1
|
0.04102
|
0.04102
|
0.04102
|
Wow
|
549
|
1
|
1
|
0.01565
|
0.01565
|
0.01565
|
Cheerful
|
12
|
1
|
1
|
0.00035
|
0.00035
|
0.00035
|
Cool
|
723
|
1
|
1
|
0.02061
|
0.02061
|
0.02061
|
Party
|
598
|
1
|
1
|
0.01705
|
0.01705
|
0.01705
|
lol
|
4243
|
1
|
1
|
0.12095
|
0.12095
|
0.12095
|
:-)
|
2761
|
1
|
1
|
0.07871
|
0.07871
|
0.07871
|
:-(
|
445
|
-1
|
1
|
0.01269
|
-0.01269
|
-0.01269
|
Sad
|
1509
|
-1
|
2
|
0.04302
|
-0.08604
|
-0.04302
|
Disappointed
|
72
|
-1
|
2
|
0.00206
|
-0.00412
|
-0.00206
|
Upset
|
55
|
-1
|
2
|
0.00157
|
-0.00314
|
-0.00157
|
Furious
|
25
|
-1
|
2
|
0.00072
|
-0.00144
|
-0.00072
|
Irritated
|
1
|
-1
|
2
|
0.00003
|
-0.00006
|
-0.00003
|
Useless
|
78
|
-1
|
2
|
0.00223
|
-0.00446
|
-0.00223
|
Bitter
|
51
|
-1
|
2
|
0.00146
|
-0.00292
|
-0.00146
|
Provoked
|
0
|
-1
|
2
|
0.00000
|
0
|
0
|
Hate
|
1188
|
-1
|
2
|
0.03387
|
-0.06774
|
-0.03387
|
Enraged
|
1
|
-1
|
2
|
0.00003
|
-0.00006
|
-0.00003
|
Bad
|
2014
|
-1
|
1
|
0.05741
|
-0.05741
|
-0.05741
|
Poor
|
55
|
-1
|
1
|
0.00157
|
-0.00157
|
-0.00157
|
Incensed
|
1
|
-1
|
2
|
0.00003
|
-0.00006
|
-0.00003
|
Terrible
|
83
|
-1
|
2
|
0.00237
|
-0.00474
|
-0.00237
|
Disgust
|
68
|
-1
|
2
|
0.00194
|
-0.00388
|
-0.00194
|
Despair
|
4
|
-1
|
1
|
0.00012
|
-0.00012
|
-0.00012
|
Discouraged
|
3
|
-1
|
2
|
0.00009
|
-0.00018
|
-0.00009
|
Hurt
|
163
|
-1
|
2
|
0.00465
|
-0.0093
|
-0.00465
|
Crushed
|
9
|
-1
|
2
|
0.00026
|
-0.00052
|
-0.00026
|
Bored
|
203
|
-1
|
1
|
0.00579
|
-0.00579
|
-0.00579
|
Pained
|
0
|
-1
|
1
|
0.00000
|
0
|
0
|
Grieve
|
5
|
-1
|
1
|
0.00015
|
-0.00015
|
-0.00015
|
Grief
|
15
|
-1
|
1
|
0.00043
|
-0.00043
|
-0.00043
|
Mourn
|
11
|
-1
|
1
|
0.00032
|
-0.00032
|
-0.00032
|
Dismayed
|
1
|
-1
|
2
|
0.00003
|
-0.00006
|
-0.00003
|
Wary
|
6
|
-1
|
1
|
0.00018
|
-0.00018
|
-0.00018
|
Unhappy
|
11
|
-1
|
1
|
0.00032
|
-0.00032
|
-0.00032
|
Unsettled
|
0
|
-1
|
1
|
0.00000
|
0
|
0
|
Mad
|
3261
|
-1
|
2
|
0.09296
|
-0.18592
|
-0.09296
|
Unsure
|
13
|
-1
|
1
|
0.00038
|
-0.00038
|
-0.00038
|
Fool
|
103
|
-1
|
2
|
0.00294
|
-0.00588
|
-0.00294
|
Pissed off
|
41
|
-1
|
2
|
0.00117
|
-0.00234
|
-0.00117
|
Angry
|
118
|
-1
|
2
|
0.00337
|
-0.00674
|
-0.00337
|
Annoyed
|
44
|
-1
|
2
|
0.00126
|
-0.00252
|
-0.00126
|
Resent
|
772
|
-1
|
1
|
0.02201
|
-0.02201
|
-0.02201
|
Weary
|
5
|
-1
|
1
|
0.00015
|
-0.00015
|
-0.00015
|
Never
|
1458
|
-1
|
1
|
0.04156
|
-0.04156
|
-0.04156
|
Aggregate
|
35082
|
1
|
0.17367
|
0.32202
|
From the results, it is evident that the aggregate mood is positive, at a value of 0.17 when the strength of each word is considered and a value of 0.32 when the strength variable is ignored. Based on this, the hypothesis that money buys happiness is confirmed, people at least seem to like to share their financial happines instead of complaining about their financial situation via twitter.
No comments:
Post a Comment