Ever wondered what an ambiguous text message or DM really meant? Think that a tweet might contain hidden notes of love (or hate) but can't be sure?
Sentimeter is the answer. It is the only Android app out there to detect as many as 64 emotions in a message. Sentimeter uses cutting-edge machine learning technology to tell you what emotions (and how much of it) went into a text!
Sentimeter is simple: Just type or paste in a text message and hit go. Our natural language model crunches the numbers and gives you its expert opinion on what feelings influenced a text.
The Sentimeter model was trained on over 1.2 billion tweets, which gives it a nuanced understanding of human emotion (including sarcasm). So much so, that it outperforms humans on several benchmarks.
Credits:
1) B. Felbo et al. for the base sentiment model
2) Freepik from FlatIcon for the app logo
Note that the model has learned about language from the raw, uncurated expressions of individuals on social media. We do not endorse in any way the emotional interpretation that the model has of any particular content.
有没有想过一个含糊不清的短信或DM究竟是什么意思?认为推文可能包含隐藏的爱情(或仇恨),但不能确定?
Sentimeter就是答案。它是唯一可以在消息中检测多达64种情绪的Android应用程序。 Sentimeter使用尖端的机器学习技术来告诉你什么情绪(以及它有多少)进入文本!
Sentimeter很简单:只需输入或粘贴短信然后点击即可。我们的自然语言模型会对数字进行处理,并为您提供关于哪些情感影响文本的专家意见。
Sentimeter模型接受了超过12亿条推文的培训,使人们对人类情感(包括讽刺)有了细致入微的理解。这么多,它在几个基准测试中表现优于人类。
积分:
1)B.Felbo等人。对于基本情绪模型
2)来自FlatIcon的Freepik用于应用徽标
请注意,该模型已经从社交媒体上个人原始的,未经表达的表达中学习了语言。我们不以任何方式认可该模型对任何特定内容的情感解释。