Connect with us

News and Politics

Research Group Discovers Loophole in New Online ‘Hate Speech’ Detectors

Published

on

 

GETTY IMAGES

State-of-the-art detectors that screen out online hate speech can be easily duped by humans, shows new study.

Hateful text and comments are an ever-increasing problem in online environments, yet addressing the rampant issue relies on being able to identify toxic content. A new study by the Aalto University Secure Systems research group https://ssg.aalto.fi has discovered weaknesses in many machine learning detectors currently used to recognize and keep hate speech at bay.

Many popular social media and online platforms use hate speech detectors that a team of researchers led by Professor N. Asokan have now shown to be brittle and easy to deceive. Bad grammar and awkward spelling–intentional or not–might make toxic social media comments harder for AI detectors to spot.

The team put seven state-of-the-art hate speech detectors to the test. All of them failed.

Modern natural language processing techniques (NLP) can classify text based on individual characters, words or sentences. When faced with textual data that differs from that used in their training, they begin to fumble.

‘We inserted typos, changed word boundaries or added neutral words to the original hate speech. Removing spaces between words was the most powerful attack, and a combination of these methods was effective even against Google’s comment-ranking system Perspective,’ says Tommi Gröndahl, doctoral student at Aalto University.

Google Perspective ranks the ‘toxicity’ of comments using text analysis methods. In 2017, researchers from the University of Washington showed that Google Perspective can be fooled by introducing simple typos. Gröndahl and his colleagues have now found that Perspective has since become resilient to simple typos yet can still be fooled by other modifications such as removing spaces or adding innocuous words like ‘love’.

A sentence like ‘I hate you’ slipped through the sieve and became non-hateful when modified into ‘Ihateyou love’.

The researchers note that in different contexts the same utterance can be regarded either as hateful or merely offensive. Hate speech is subjective and context-specific, which renders text analysis techniques insufficient as stand-alone solutions.

The researchers recommend that more attention be paid to the quality of data sets used to train machine learning models–rather than refining the model design. The results indicate that character-based detection could be a viable way to improve current applications.

The study was carried out in collaboration with researchers from University of Padua in Italy. The results will be presented at the ACM AISec workshop in October.

The study is part of an ongoing project called Deception Detection via Text Analysis https://ssg.aalto.fi/research/projects/deception-detection-via-text-analysis in the Secure Systems group https://ssg.aalto.fi> at Aalto University.

CREDIT: AALTO UNIVERSITY SECURE SYSTEMS RESEARCH GROUP

Facebook Comments

News and Politics

CNN on-boards Andrew Yang as Political Commentator

Published

on

Credit: Saul Loeb

Former Democratic presidential candidate Andrew Yang is joining CNN as a political commentator, the network announced Wednesday.

“I’m excited to join @CNN to help shed light on the election and the candidates’ experiences,” Yang posted via Twitter.

“Learned a lot these past months and am glad to contribute to the public discussion.” He’ll appear on the network in his new capacity later Wednesday, he added.

The tech entrepreneur Andrew Yang suspended his campaign last Tuesday, Feb. 11 in his bid for the White House.

“We’ve accomplished so much together,” Yang said during a speech to supporters in Durham, N.H.

Thank you #YangGang.

Posted by Andrew Yang for President 2020 on Wednesday, February 12, 2020

Yang rose from obscurity to become a highly-visible candidate, rallying a coalition of liberal Democrats, libertarians and some disaffected Republicans to form a devoted group of followers known as the Yang Gang.

Facebook Comments

Continue Reading

News and Politics

New York Rapper Pop Smoke Shot and Killed at 20

Published

on

Image via Getty/Jeremy Moeller

The budding Brooklyn rap star Pop Smoke (Bashar Barakah Jackson) has reportedly been murdered. 

According to TMZ, Pop Smoke was shot and killed early Wednesday morning in what appears to be a home-invasion robbery.

LAPD confirmed the rapper’s death, claiming that two men wearing hoodies and masks broke into Pop Smoke’s Hollywood Hills house around 4:30AM, shooting Pop Smoke multiple times and fleeing the scene on foot.

He was pronounced dead upon arriving at Cedars-Sinai Medical Center in West Hollywood.

No suspects have been apprehended or identified at this time, and there is no word on whether the rapper had any connection to the them.

The young emcee rose to fame with his first single, “Welcome to the Party,” which also featured rap heavyweight Nicki Minaj on the remix. 

The rapper’s death came just days after he released his anticipated new mixtape, Meet the Woo 2. Pop Smoke was age 20 when he passed.

Facebook Comments

Continue Reading

News and Politics

U.S. Passengers released from Quarantined Cruise Ship in Japan

Published

on

AP

The United States has reportedly evacuated its citizens from the Diamond Princess cruise ship in Japan on Monday morning after they had been quarantined for more than a week in the Japanese port city of Yokohama. 

Other nations are preparing to follow suit as global concerns heighten over the spread of the deadly disease with a death toll of 1,700 at its Chinese epicenter.

Fourteen people who tested positive for the coronavirus were among the more than 300 American citizens and family members evacuated by plane from the quarantined ship, the US State Department revealed Monday.

More than 70,500 have been infected across China by the COVID-19 virus as the world battles to contain the outbreak that has sparked panic buying and jitters over the global economic impact.

Facebook Comments

Continue Reading

Trending