Parent Interventions: Navigating through the Pandemic  |  Health Monitor: Attention and Memory Deficits in People Who Experienced Mild Covid  |  Parent Interventions: How can we Revert Peanut Allergies in Children?  |  Teacher Insights: Play Based Learning has a Positive Impact on Child's Learning and Development  |  Health Monitor: Social Media Use Likely to Affect the Physical Health of a Person  |  Parent Interventions: How to Deal with Developmental Language Disorder in Children  |  Health Monitor: Lifestyle Interventions from Early Childhood Prevents Cardiovascular Diseases  |  Teacher Insights: Teacher Expectations Can Have Powerful Impact on Students Academic Achievement  |  Policy Indications: Make Sure the Digital Technology Works for Public Good  |  Teacher Insights: The Significance of Social Emotional Learning Curriculum in Schools  |  Health Monitor: Forgetting is a Form of Learning  |  Higher Studies: University of Manchester Invites Application for LLB and LLM Programmes   |  Health Monitor: Is There a Blue Spot Inside our Brain?  |  Parent Interventions: Babies born during the Pandemic Performs Lower during Developmental Screening  |  Policy Indications: Invest in Structural Steel R&D : Prof BS Murty  |  
July 27, 2019 Saturday 09:51:27 AM IST

Statistical Method Identified to Detect Fake or Computer-Generated Text

By Felix Litchenfeld

Researchers at Harvard John A Paulson School of Engineering and Applied Sciences (SEAS) and IBM have developed a statistical method to detect computer-generated or fake text from human generated text.
Researchers Sebastian Gehrmann and Hendrik Strobert(IBM) found that natural-language generators are trained on tens of millions of online texts and mimic human language by predicting the words that most often follow one another. For eg I followed by 'have' and 'am'. Using this idea they developed a method that identifies predictable text instead of flagging errors. Gehrmann and Strobelt’s method, known as GLTR, is based on a model trained on 45 million texts from websites — the public version of the OpenAI model, GPT-2. Because it uses GPT-2 to detect generated text, GLTR works best against GPT-2, but it does well against other models, too.


https://news.harvard.edu/gazette/story/2019/07/researchers-develop-a-method-to-identify-computer-generated-text/?utm_medium=Feed&utm_source=Syndication



Comments