Protection Order Sentiment Analysis - Initial Exploration
Purpose/Context:
Apply R's "sentimentr" package to initial batch of transcribed protection order text in order to get a sense of what analysis is doable and interesting.
Average Sentiment Per Protection Order
For each Protection Order, produce a word count, a sentiment score and corresponding standard deviation (e.g. the variability of positive & negative emotions in a single protection order).
| 1 |
248 |
0.3919165 |
-0.2896926 |
| 6 |
323 |
0.2565173 |
-0.1678728 |
| 3 |
116 |
0.3297618 |
-0.1260139 |
| 2 |
764 |
0.2835390 |
-0.0646693 |
| 4 |
366 |
0.1243202 |
-0.0063708 |
| 7 |
0 |
NA |
0.0000000 |
| 5 |
139 |
0.1532549 |
0.0544215 |
Density Plot of Sentence Sentiment Across all Protection Orders
Show the distribution of positive and negative emotions for every sentence across the whole dataset. 
##
## Call:
## density.default(x = senti$sentiment)
##
## Data: senti$sentiment (73 obs.); Bandwidth 'bw' = 0.1168
##
## x y
## Min. :-1.4659 Min. :0.0005279
## 1st Qu.:-0.8260 1st Qu.:0.0525892
## Median :-0.1862 Median :0.2437542
## Mean :-0.1862 Mean :0.3903267
## 3rd Qu.: 0.4536 3rd Qu.:0.6740276
## Max. : 1.0935 Max. :1.1708733
Top Emotions Referenced in all POs
Plot the most common emotions across the whole dataset. 
## # A tibble: 14 × 2
## emotion_type ave_emotion
## <fct> <dbl>
## 1 anger 0.0188
## 2 anger_negated 0.00246
## 3 anticipation 0.0190
## 4 anticipation_negated 0.00227
## 5 disgust 0.00592
## 6 fear 0.0237
## 7 fear_negated 0.00337
## 8 joy 0.00953
## 9 sadness 0.0173
## 10 sadness_negated 0.00205
## 11 surprise 0.00519
## 12 surprise_negated 0.000825
## 13 trust 0.0214
## 14 trust_negated 0.00263
