How machine learning may help fight illegal wildlife trade on social media

The illegal wildlife trade is one in all the biggest threats to biodiversity. Demand for species and wildlife products, like rhino horn, elephant ivory and pangolin scales, have triggered a rise in unsustainable harvesting of species. This causes necessary population declines, threatening the existence of certain species.

Traditionally, illegal wildlife trade thrived in physical markets. But today it has also moved online. In China, greater than half of the trade in elephant ivory items happens on e-commerce platforms. Of this, recent investigations showed that social media platforms are the preferred ways to advertise, source, and trade species and wildlife products.

Social media offers good conditions for wildlife trafficking to thrive: the platforms are easily accessible and have a high variety of users. It’s also good for sellers who can put up photographs and detailed information concerning the products.

But monitoring illegal wildlife trade on social media is a challenge. There are an enormous number of various groups and profiles that data law enforcers have to research. Traders also often use code words to cover illegal transactions. So far, law enforcement efforts by governments, international organisations and NGOs have mainly focused on manually searching social media content for information. This is completed on a number of key species and wildlife products.

In our recent study we proposed a system that uses machine learning – a style of artificial intelligence that gives systems with the power to routinely learn and improve from experience – to research illegal wildlife trade on social media platforms. To our knowledge, that is the primary time automatic content identification methods have been used to research illegal wildlife trade on social media.

These recent methods and data can provide us with fresh insights on illegal wildlife trade at a worldwide scale. Because it’s done by a pc, it’s in a position to process an enormous amount of data in a brief space of time. It also caters for a wide selection of various species and wildlife products.

How it really works

The system consists of three stages: mining (examining large databases of data), filtering, and identifying relevant information on illegal wildlife trade on social media.

The whole system might be automated in order that data are mined directly from social media platforms, the content is filtered and only relevant content is kept for further investigations by a pc or an individual.

Neural networks make this all occur. These are a particular set of algorithms – or machine rules – that might be trained to recognise and classify species and wildlife products, like a pangolin or rhino horn, from images contained in social media posts. They can do that while also picking up the pictures’ settings – like a pangolin in a cage, in a market place.

Neural networks can be trained to look for extra information in videos – like specific bird calls in audio – or in text – just like the writing style characteristic of illegal transactions or special interest groups, like exotic pet owners.

Natural language processing, which analyses language, is used to infer the meaning of a text description in social media posts and assess social media users’ preferences, and their sentiment towards species and wildlife products.

By comprehensively taking a look at all this content we imagine we are going to unveil patterns in the information that we’d not have the option to uncover otherwise – for instance combos of specific words and pictures.

In a previous study we demonstrated the worth of such a system. We trained a deep neural network to find out whether Twitter posts with the word “rhino”, in 19 different languages, contained images of rhinoceros species. In doing so, we were successfully in a position to routinely discover all images regarding rhinos from Twitter content. This reduced the variety of images that a human expert would must review by over 90%.

Only a number of examples of automatic online content identification exist. In one other study, researchers were in a position to detect illegal elephant ivory items on the market on an e-commerce platform by utilizing metadata.

Machine learning might be used to avoid wasting experts’ time when manually classifying content on social media. It can be utilized by social media platforms to discover and promptly remove suspicious posts.

Way forward

The biggest challenge in developing efficient monitoring algorithms is to create adequate training datasets. For instance, when classifying images algorithms should be trained to pair images with a corresponding label – like “rhino horn”.

But for some rare species, and for a lot of wildlife products, there may be an entire ignorance that might be used to create training datasets, for instance rhino horn powder. In cases like this, we aim for a less accurate result which requires less training data and improve the models with time as more data becomes available.

The advancement of the system would require working in close collaboration with social media platforms, while respecting ethical and privacy requirements. With time, success in curbing illegal wildlife trade on social media, and other digital platforms, will hopefully help to cut back poaching of species within the wild.