Global Content and AI Are Partners Ready for Prime Time

Global content is at the threshold of a Cambrian explosion—aided by artificial intelligence (AI). But let’s begin at the beginning. We can discern two epochs of global content, and we’ll call them global content (GC) 1.0 and GC 2.0.

The GC 1.0 era, roughly the last 20–25 years, was all about getting the basics right for multilingual applications. Globalization and its doppelgänger, localization, were the focus areas. After all these years, it is fair to say we have a good grasp of the best practices, and content globalization is a relatively mature field now. Our quest for the holy grail of high-quality machine translations, an unrealistic goal to begin with, remains inconclusive. While that has disappointed some, the bigger letdown has been that for many organizations, content globalization is a bit like going to the gym. Everyone agrees that it is a good idea, but not many follow through.

But that’s about to change. So far, we mainly viewed content globalization through the narrow lens of translation. Going forward, I predict that the canvas will be much larger. In fact, content globalization will be a prerequisite for practically all AI use cases. Welcome to GC 2.0.

Content globalization may no longer be a nice-to-have; it may have become a strategic priority for organizations because of the advances in AI. In the last few years, because of the availability of inexpensive computing resources (i.e., cloud), access to vast amounts of raw data, and breakthroughs in machine learning methods, there has been a quantum leap in the capabilities of AI technologies. The good news is that these AI tools and techniques are available for organizations of all sizes to use as they see fit.

When it comes to the core translation use cases, the newer techniques—such as artificial neural networks (aka deep learning)—improve the quality of content translations compared to previous-era machine translation efforts. It is tough to quantify such improvements, but market feedback suggests a 25%–30% reduction in translation errors. At such levels of accuracy, automated content translations become passable for low-fidelity translation applications.

However, GC 2.0 is much more than just improved automated content translations. Previously, the emphasis was on auto-translations of long text such as documents and website content. GC 2.0 spans a much broader spectrum. It contains the written text but also the spoken word and a wide array of other content formats. Organizations want to mine this content cornucopia to improve the customer experience and enhance customer engagement. In other words, content globalization becomes part and parcel of the core business process and not an activity to be undertaken separately. AI is the prime enabler here.

AI techniques such as text mining, sentiment analysis, emotion detection, speech recognition, document classification, and handwriting/character recognition are all language-specific. As their names suggest, natural language processing, natural language understanding, and natural language generation all draw from the fount of language. 

As you can see, GC 2.0 and AI are joined at the hip for many new AI use cases and applications—be it social media sentiment, analysis, conversational interfaces (aka chatbots), Q&A systems, or the voice-based digital assistants on your phone. GC 2.0 is really about supporting such contextual-translation applications.

A few words of caution. The AI/deep learning methods work well only where labeled data is available for a particular language. AI platforms support English and a few major languages well, but it’s otherwise a mixed bag. Also, be aware that language support varies within the modules/components (e.g., language recognition versus sentiment analysis) of the same platform. It is a considerably significant undertaking to add support for a new language—so check carefully that languages of interest to you are supported by the AI platform you are planning to use.

In closing, AI enables content globalization and fuels demand for global content. AI and GC are perfect partners getting ready together for prime time.   

Related Articles

If there is one technology topic that drew a great amount of media attention this year, it is artificial intelligence (AI). Thus far, you can say that only digital technology giants such as Amazon, Facebook, and Google—along with a few other early adopters—have tasted success in implementing AI technologies. Here are 10 commandments to finally get you started on your enterprise AI strategy.
Content Management is a very broad discipline and encompasses several areas. I find it useful to distinguish between content that helps enhance the customer experience or content that is critical to a business process. Here are the top use cases machine learning use cases for content management.
Discovery models (such as text search, voice search, visual search, chatbots, natural language search, and faceted search) may overlap in some areas of functionality, but also enhance different use cases and appeal to different types of users.
RPA is an imaginative acronym that alludes to robots but the reality is more prosaic. RPA robots are intended to mimic tasks performed by humans on their computers after they are trained or programmed. Training these "bots" involves showing them the steps of a workflow or business process such as opening a website, keying in data and clicking through screens.