AltSolution

AI researchers made a sarcasm detection model and it’s sooo impressive 😮🤨😏

Researchers in China say they’ve created sarcasm detection AI that achieved state-of-the-art performance on a dataset drawn from Twitter. The AI uses multimodal learning that combines text and imagery since both are often needed to understand whether a person is being sarcastic.

The researchers argue that sarcasm detection can assist with sentiment analysis and crowdsourced understanding of public attitudes about a particular subject. In a challenge initiated earlier this year, Facebook is using multimodal AI to recognize whether memes violate its terms of service.

The researchers’ AI focuses on differences between text and imagery and then combines those results to make predictions. It also compares hashtags to tweet text to help assess the sentiment a user is trying to convey.

“Particularly, the input tokens will give high attention values to the image regions contradicting them, as incongruity is a key character of sarcasm,” the paper reads. “As the incongruity might only appear within the text (e.g., a sarcastic text associated with an unrelated image), it is necessary to consider the intra modality incongruity.”

On a dataset drawn from Twitter, the model achieved a 2.74% improvement on a sarcasm detection F1 score compared to HFM, a multimodal detection model introduced last year. The new model also achieved an 86% accuracy rate, compared to 83% for HFM.

The paper was published jointly by the Chinese Academy of Sciences and the Institute of Information Engineering, both in Beijing, China. The paper was presented this week at the virtual Empirical Methods in Natural Language Processing (EMNLP) conference.

The AI is the latest example of multimodal sarcasm detection to emerge since AI researchers began studying sarcasm in multimodal content on Instagram, Tumblr, and Twitter in 2016.

University of Michigan and University of Singapore researchers used language models and computer vision to detect sarcasm in television shows, a model detailed in a paper titled “Towards Multimodal Sarcasm Detection (An Obviously Perfect Paper).” That work was highlighted as part of the Association for Computational Linguistics (ACL) last year.
https://bit.ly/2IIKuJs

Google launches Document AI suite of parsing and processing tools in preview 🤔

Google this morning launched the Document AI (DocAI) platform, a console for document processing hosted in Google Cloud, in preview. The company says it’s aimed at automating and validating documents by extracting data from documents and making them available to business apps and users.

Companies spend an average of $20 to file and store a single document, by some estimates, and only 18% of companies consider themselves paperless. An IDC report revealed that document-related challenges account for a 21.3% productivity loss, and U.S. companies waste a collective $8 billion annually managing paperwork.

Google’s DocAI platform ostensibly solves this by providing access to document parsers, tools, and solutions via an API. It supports the creation and customization of document processing workflows built with Google Cloud’s predefined taxonomy without the need to perform additional data mapping or training. DocAI offers general processors including a form parser, W9 parser, optical character recognition, document splitter, and custom workflows for domain-specific documents. These reside in a unified dashboard from where they can be tested by uploading a document directly in the console.

The parsers can classify information in documents like addresses, account numbers, and signatures as well as extract data like supplier names, invoice dates, and payment terms. Google says it’s working on additional capabilities for the DocAI platform to grow its core capabilities and support additional toolsets.

General parsers such as optical character recognition, the form parser, and the document splitter are available from the DocAI platform console. Access to specialized parsers like W9, 1040, W2, 1099-MISC, 1003, invoice, and receipts must be requested on a per-customer basis.

The launch of Google’s DocAI platform comes after the release of Lending DocAI, a Google Cloud product for the mortgage industry that ostensibly provides “industry-leading” accuracy for documents relevant to lending and processing. Google also recently unveiled PPP Lending AI, an effort to help lenders expedite the processing of applications for the since-exhausted U.S. Small Business Administration’s (SBA) Paycheck Protection Program, and Procurement DocAI, which automates procurement data capture by turning docs like invoices and receipts into structured data.

Lending DocAI and Procurement DocAI are now a part of the DocAI platform.

“We believe that any company that has to manually extract data from complex documents at scale can greatly benefit from Google Cloud AI,” Google product manager Lewis Liu and product marketing manager Yang Liang wrote in a blog post. “Transforming documents into structured data increases the speed of decision making for companies, unlocking measurable business value and helping develop better experiences for customers.”
https://bit.ly/3ndbS0q