Ask HN: Does “tone detection for text” software exist? | Hacker News Reader

Ask HN: Does “tone detection for text” software exist?

6 pointsmrsmee8911 comments

I'm looking for a something that can detect and characterize the tone of a written text the way we do naturally when speaking. For example detecting sarcasm in a tweet.

View on HN

The general category of research is sentiment analysis:

https://en.wikipedia.org/wiki/Sentiment_analysis

However that's as much as I know. If you get lucky maybe an expert or user of these tools will see your post!

(2000) https://www.qualcomm.com/news/releases/2000/09/11/qualcomms-...

> MoodWatch, based on rhetorical theories developed by David Kaufer, chairman of the English department of Carnegie Mellon University, employs a very fast and efficient algorithm to identify words and phrases that might be offensive. As such, it's efficient at flagging potentially offensive messages, but it is up to the user to decide if a message 'deserves the chilies.'

This is from a release announcement of Eudora, a bygone, once-popular e-mail client. (Kudos to Qualcomm to keeping this online 22 years later!) "Deserve the chilies" refers to how the client would display one to three chili pepper icons if it felt your e-mail was rude or offensive. (Software used to be whimsical.)

There was a white paper that described how they trained the feature on postings from the Usenet group alt.flame. Also, the source code to Eudora is now available from the Computer History Museum, including MoodWatch: https://computerhistory.org/blog/the-eudora-email-client-sou...

You're looking for sentiment analysis and you won't get good results unless the text is in a specific domain. I did this work for a previous company where we analyzed restaurant reviews to extract actionable feedback. The trouble is that the written word is missing most of the nonverbal cues that give you the tone for spoken word (think air quotes or rolled eyes). So when you see a review that says "My steak came out well done...great job guys!" It looks positive but could be sarcastic. Or when someone says "We came to Taco Bell because we were looking for the most authentic Mexican cuisine" of course this is sarcastic but there is no way you could teach a computer to spot that.

Hey now, taco bell is the sh*t! But seriously, thanks for the insight. Do u think this is solvable? It would be so helpful for online discourse.

Like I said, if the domain is super specific, it is solvable. That Taco Bell example is real. A private equity company was considering purchasing them from Yum Brands and they wanted to know what the customers liked or disliked. If all of your text is Taco Bell reviews, then you can train a model to spot sarcasm by having a human tag a portion and letting the model extrapolate from there. The model will determine which words (e.g. authentic, gourmet, and homemade) correlate with sarcasm. This breaks down if you try to apply it to all restaurants because what is clearly sarcasm for Taco Bell may not be for Lupe's Mexican Cantina. In the end, we went with a more objective approach. I broke the reviews into sentences and analyzed independent and dependent clauses, identified nouns, verbs, and adjectives, and used those to determine the main idea of the clause. I noticed that 1st person nouns and verbs are what the reviewers said about themselves (e.g. I was hungry or I rolled into the drive-thru) and 2nd and 3rd person was about the restaurant (you need to train your employees, the taco was cold or the cashier gave me the wrong order). Looking at the sentiment of the clause and its subject/verb phrase gave us exactly what we needed. In the end we had a list of ideas like "cold taco - negative", "waited forever - negative", "super cheap - positive" which we used a logistic regression classifier to put into categories like price, service, selection, ambience, etc.

Taco Bell is food for chihuahuas, not humans.

To correctly analyze the sentiment of that, your software would have to have the "Yo quiero Taco Bell" ad series from, what, the 90s maybe, in its database. It's snarky, but it's not the outright slam that the words themselves would seem to indicate.

That's the problem. There's sarcasm. There's parody, which if done well can be impossible to tell from the real thing. (That's somebody's law - too lazy to look it up right now.) Those are bad enough. But when a sentence drags in cultural knowledge, you have to understand the cultural knowledge to accurately assess the sentence. It's tough to ask a program to "get" not just what the reference is, but how that affects what a sentence is actually saying.

If you think about this you'll realize that it is impossible to detect tone from text because it's not possible to convey tone with text alone. Tone is auditory, text is visual, and there is no way to recover one from the other. For an obvious demonstration of this watch a movie on mute and then listen to the audio of the movie without the accompanying visuals.

Are looking for this?

search: >'AI Content Moderation'< @DDG : <https://html.duckduckgo.com/html/?q='AI Content Moderation'>

IBM used to have one but it was deprecated this year. Looks like there’s a follow up product with a decent free tier. https://cloud.ibm.com/apidocs/tone-analyzer

Anyone know or used this?

search: >microsoft:'AI Content Moderation'< @DDG : <https://html.duckduckgo.com/html/?q=microsoft:'AI Content Mo...>

See this tutorial: https://www.youtube.com/watch?v=ZUqB-luawZg It kind of works but this is a very very difficult problem to solve.

Reply on news.ycombinator.com