Computational Sarcasm: Understanding, Generating, and Detecting Sarcasm in AI Agents
The emerging theme of Computational Sarcasm explores the nuances of understanding, generating, and detecting sarcasm in AI agents. This involves using incongruity within text, sentiment annotation complexity, and emotion engines to develop sarcasm-enabled chatbots. Various approaches such as rule-based and context incongruity-based methods are being employed to enhance AI's ability to recognize and produce sarcasm effectively.
Download Presentation
Please find below an Image/Link to download the presentation.
The content on the website is provided AS IS for your information and personal use only. It may not be sold, licensed, or shared on other websites without obtaining consent from the author. Download presentation by click this link. If you encounter any issues during the download, it is possible that the publisher has removed the file from their server.
E N D
Presentation Transcript
Human: You are fast, just like a snail ---------------------- ALICE: Thank you for telling me I am fast like a snail Assistant.ai: A good assistant is whatever their boss needs them to be Turing Test An AI agent must be able to imitate humans Sarcasm A cutting, often ironic remark intended to express contempt or ridicule Thefreedictionary.com All images from wikimedia commons.
Computational Sarcasm Sprint Thesis Talk by Aditya Joshi adityaj@cse.iitb.ac.in PhD Advisors: Prof. Pushpak Bhattacharyya, IITB Prof. Mark J Carman, Monash
The emerging theme Computational Sarcasm Understanding sarcasm in humans Generating sarcasm Detecting sarcasm A chatbot module that always responds sarcastically Sarcasm using context incongruity within target text Sentiment annotation complexity (eye-tracking) Sarcasm annotation complexity (eye-tracking) An emotion engine that measures change in emotions Sarcasm using incongruity with historical text by the author: (a) Rule-based, (b) Topic model-based Sarcasm in different cultures A sarcasm-enabled chatbot Sarcasm using incongruity with historical text in a conversation Ongoing Planned 3
Sarcasm Generation: A chatbot that responds sarcastically! Based on SarcasmBot: An open-source sarcasm generation module for chatbots , WISDOM-KDD 2015, Sydney, August 2015 What do you think of Greg? User Input Entities: Greg: Name Tense: Present Evaluation Were human evaluators able to identify which of two responses is from our chatbot? Input Analyzer Type of question: Opinion question Generator Selector I <sentiment-word> <entity>. <Expression-of-opposite-sentiment> Sarcasm Generators (8) I like Greg. The way I love zero-accountability people.
Sarcasm Detection: Using incongruity for a sarcasm classifier Based on Harnessing Context Incongruity for Sarcasm Detection , ACL IJCNLP 2015, Beijing, July 2015, and Your sentiment precedes you... , WASSA EMNLP 2015, Portugal, 2015. Conclusion Our work presents approaches to sarcasm generation and detection. Being stranded in traffic is the best way to start a week. Words Implicit phrases ... Features # Sentiment Flips Lexical polarity I loovved participating in RISC this year. Nicki Minaj! Don t I looovve her! Also see: My poster based on Drunk-texting Prediction Features as given above + Features Sentiment agreement with historical tweets Historical tweet: Nicki Minaj is pathetic!
