KeyMoji 關鍵情緒偵測 (SENSE2、SENSE8、Tension) 採用不同於其它「素人標記」和「純機器學習」的文本情緒偵測分析工具,結合了「句型」、「邏輯語意」和「詞彙模型」,設計出一個完整的「情緒計算過程」。
pip3 install KeyMojiAPI
or
python3 -m pip install KeyMojiAPI
KeyMoji Website Demo
KeyMojiAPI Documentation
from KeyMojiAPI import KeyMoji
# 若您是使用 Docker 版本,無須填入 username, keymoji_key 參數
keymoji = KeyMoji(username="", keymojiKey="")
inputSTR = "他逃離了危險的災難"
# Sense2
sense2Result = keymoji.sense2(inputSTR, model="general", userDefinedDICT={"positive":[], "negative":[], "cursing":[]})
print(sense2Result)
# Sense2 Visualization
status = keymoji.keymoji2visual(sense2Result, filename="kimetsu.png")
{
"status": true,
"msg": "Success!",
"results": [
{
"score": 0.2798,
"sentiment": "positive",
"input_str": "他逃離了危險的災難",
"cursing": false
}
],
"sense": "sense2",
"version": "v101"
}
from KeyMojiAPI import KeyMoji
# 若您是使用 Docker 版本,無須填入 username, keymoji_key 參數
keymoji = KeyMoji(username="", keymojiKey="")
inputSTR = "他逃離了危險的災難"
sense8Result = keymoji.sense8(inputSTR, model="general", userDefinedDICT={"positive":[], "negative":[], "cursing":[]})
print(sense8Result)
# Sense8 Visualization
status = keymoji.keymoji2visual(sense8Result, filename="kimetsu.zip")
{
"status": True,
"msg": "Success!",
"results": [
{
"input_str": "他逃離了危險的災難",
"Joy": 3.7486,
"Trust": 5.1776,
"Surprise": 6.7238,
"Anticipation": 0.9618,
"Fear": 0.9505,
"Sadness": 0.9108,
"Anger": 0.9516,
"Disgust": 0.8876
}
],
"sense": "sense8",
"version": "v101"
}