Skip to content

Emotional Traits Knowledge Model

Overview

The Emotional Traits Knowledge Model (display name: Emotional Traits EN v#) classifies documents in terms of the feelings—joy, surprise, irritation, etc.—expressed in the text. They can recognize 39 different emotional traits divided into eight groups.

Mostly aimed at documents written in 1st person (tweets, messages, interviews, reviews), but not only, the model can distinguish between relatively subtle differences in emotions.

The model returns leaf categories, that is 2nd level categories like Excitement and Disillusion, but if the Output rules extra data functional option of the workflow block if turned on, it also returns the main groups of emotional traits. Main groups are the category tree groups corresponding to the most relevant emotional traits expressed in the text. They provide an easy-to-read indication of the clusters of emotional traits the text is more about, similarly to an abstract.

Category tree

Categories are divided in eight groups containing emotions with similar characteristics:

0100 Group Rage
    0101 Anger
    0102 Irritation
    0103 Exasperation
0200 Group Apprehension
    0202 Anxiety
    0203 Fear
    0204 Stress
    0205 Worry
0300 Group Distress
    0301 Disgust
    0302 Repulsion
    0311 Guilt
    0312 Shame
    0313 Embarrassment
    0322 Regret
    0331 Boredom
0400 Group Resentment
    0402 Hatred
    0403 Offence
    0411 Jealousy
    0412 Envy
0500 Group Dejection
    0501 Sadness
    0502 Torment
    0503 Suffering
    0511 Disappointment
    0512 Disillusion
    0513 Resignation
0600 Group Surprise
    0601 Surprise
0700 Group Delight
    0701 Happiness
    0702 Excitement
    0703 Joy
    0704 Amusement
    0705 Well-Being
    0711 Satisfaction
    0721 Relief
0800 Group Fondness
    0801 Like
    0802 Trust
    0803 Affection
    0804 Love
    0805 Passion
    0812 Empathy
    0813 Compassion

Output structure

The model output has the same structure as any other model and is affected by the functional properties of the workflow block.
The peculiar parts of the output are the result of categorization, i.e. the categorization array, and the extraData object: to have extraData it's necessary to turn on the Output rules extra data functional option of the workflow block.

The extraData object contans the main gropus of emotional traits found in the text, for example:

"extraData": {
    "groups": [
        {
            "id": "0500",
            "label": "Group Dejection",
            "position": 1
        }
    ]
}

Each item of the group array corresponds to a group of emotional traits and has these properties:

Property Description
id Identification number or category group name in the category tree
label Group category description
position Ranking of the group

The group with the highest rank has position set to 1. Other groups, when present, have consecutive values.

Example

Considering the following text:

We're very busy #coding a whole network manager for #unity3d based on #steamworks networking. #gamedev #indiedev #3amDeadTime #horror #game #amusement

The categorization output is:

"categories": [
    {
        "frequency": 23.06,
        "hierarchy": [
            "Group Apprehension",
            "Fear"
        ],
        "id": "0203",
        "label": "Fear",
        "namespace": "emotional_traits_en",
        "positions": [
            {
                "end": 133,
                "start": 127
            }
        ],
        "score": 3,
        "winner": true
    },
    {
        "frequency": 76.91,
        "hierarchy": [
            "Group Delight",
            "Amusement"
        ],
        "id": "0704",
        "label": "Amusement",
        "namespace": "emotional_traits_en",
        "positions": [
            {
                "end": 150,
                "start": 141
            }
        ],
        "score": 10,
        "winner": true
    }
]