The 16 personalities dataset is publicly available and contains 8675 rows of data.

The data collection process begins as users foremostly complete a quesionnaire or survey, allowing professionals to identify their personality type. Each entry in the dataset is given as a tuple of the corresponding user's personality type and a section from their latest 50 posts separated by 3 pipes ("||") characters. Four different dimensions were available inside each personality label. For instance, if the personality type label is ITSJ, we can conclude the first personality dimension of this person is Introvert (I), the second, third, and fourth dimensions are Thinking (T), Sensing (S), and Judging (J).

It is also interesting to note that this dataset provides a classification of human personalities in 16 ways via four dimensions, namely, (1) introversion/extroversion (I/E), (2) sensing/intuition (S/N), (3) thinking/feeling (T/F) and (4) judging/perceiving (P/J)
## Files

*   **raw_train.csv** - the training set of unprocessed text
*   **clean_train.csv** - the training set of processed text
*  **raw_test.csv** - the test set of unprocessed text you have to make predictions on
*   **sample_solution.csv** - a sample submission file in the correct format

## Columns

*   `posts` - text of inputs
*   `type` - definition of personality type (label)