Hong Kong’s Surround App Translates Sina Weibo Content into English

Read any news report out of China today, and chances are it’ll include a smarmy comment or two from a netizen on Sina Weibo. The 400 million-strong microblogging site is almost synonymous with the voice of the country’s young urbanites – so much so that there’s an online cottage industry (Tea Leaf Nation, Ministry of Tofu and China Digital Times among others) dedicated to translating cross-sections of it for an English audience.

But without fluency in Chinese (and a street-smart awareness of online slang), most of Weibo is near incomprehensible. That’s where Hong Kong-based startup Surround App comes in. It’s an upcoming English mobile client for Sina Weibo, with its first release planned for April 2013, offering a mix of machine and human translation to help you make sense of the content on China’s vast social networks. From trenchant social commentary to…um…Brad Pitt.

“It was inspired basically by my own frustration,” says co-founder Jeffrey Broer over my Skype interview with him, admitting that he can only speak about five sentences in Mandarin. “I’ve been in and out of mainland China for over 10 years, and there are all these fascinating people online that I want to follow, but I can’t. It was a whole world that I couldn’t grasp.”

Broer, the founder of another Hong-Kong based startup, Grayscale, put together a team of four late in 2012, and Surround App received about US$15,000 in seed funding (plus office space and technical guidance) from HK Accelerator.

Targeted at “language students, bloggers, micro-celebrities and managers in China”, Surround App plans to follow a freemium model – the free app will allow awkward machine translations (like copy-pasting a block of text into Google Translate) of all Sina Weibo content, helped partly by a slang database that the team is planning to build into future releases. Broer adds:

We’re still playing around with the specifics. But we’ll have monthly subscriptions that will feature human translations of specific accounts, a ‘curated’ human-translated daily news section, and packages for corporate clients.

Even with translation, the two challenges in engaging with Sina Weibo are figuring out who to follow:

We’re developing our own strong search databases for that, and we’ll present users with recommendations based on keywords.

Users will be able to make sense of the dense web of punning and wordplay that the network is famous for, and understanding those images people keep swapping. For that, Broer says:

We have a few ideas, such as optical character recognition (OCR) that we could use for images, but not planning to address that in our first release.