phil_emote

A ready-to-use sentiment analysis and emotion classification model for Philippine tweets.

Using tweets collected from the Philippine, an LSTM model was trained for sentiment analysis and emotion classification. This model was trained into 11 classes specifically:

Emoji	Emotion	Sentiment
📝	neutral	neutral
😄	happy	positive
😌	relief	positive
😑	unammused	negative
😘	love	positive
😜	playful	positive
😞	sad	negative
😡	angry	negative
😱	shocked	negative
😷	sick	negative
🤔	pondering	neutral

Labels used were the 11 emojis which were selected by clustering a set of facial expression emojis into ten clusters. Then, tweets containing news were added to the dataset to represent the neutral class.

Quickstart

Create an instance of PhilEmoteModel. Then, pass a list of tweets to predict_dataset to predict its respective emotion. Alternatively, the model can be used to predict sentiment by setting output_type='sentiment'. Also, setting output_type='emoji' will output the emoji labels used.

Running the following code:

from phil_emote.models import PhilEmoteModel

dataset = [
    "Nakakaasar nman! ahhh... ang dami dami kong gagawin... kulang time...",
    "Asar... Badtrip!",
    "May bagong bagyo ang namataan sa kanlurang bahagi ng bansa.",
    "Thank you po Lord Jesus at gumana din sya.",
    "Congrats po sa inyong lahat!",
]

model = PhilEmoteModel()
pred = model.predict_dataset(dataset, output_type='emotion')
print(pred)

will output the following:

['unammused', 'angry', 'neutral', 'love', 'love']

Setting output_type='sentiment' will output the following:

['negative', 'negative', 'neutral', 'positive', 'positive']