TREND: Advances in artificial intelligence will enable a) next generation of web browsers to move beyond key word analysis and evaluate the specific content of websites/pages (the semantic web); b) networked devices to combine speech recognition, machine translation and speech synthesis to support real time multilingual voice translation; and c) cloud based crowd sourced translation checking of webpage text
Research continues to enable the next generation of search engines and web browsers to evaluate and assess the specific content of pages/sites (as opposed to simply reading meta data, tags or identifying key words). If implemented effectively the semantic web would revolutionise the efficiency of search with a correspondingly positive impact on access to information and research productivity. However, this same technology could have negative implications in relation to tracking, censorship and monitoring/blocking content.
In a context where three quarters of the global population (and just under a three quarters of global Internet users) do not speak English, language still represents a significant barrier to access to information where English remains (closely followed and soon to be surpassed by Chinese) the leading language on the web (Internet World Stats). However, recent advances in combining speech recognition, machine learning, machine translation and speech synthesis technology may have the capacity to support real time multilingual voice translation via any Internet enabled device within the near future.
Despite the perennial problems of adapting to slang, regional accents and culturally specific idioms and concepts – pioneering approaches using deep neural networks (Microsoft) and cloud-based crowd sourced sentence checking (Google) are showing significant promise. In conjunction with related developments in webpage translation methods, this trend has the potential to dissolve many of the barriers which limit access to multicultural content, and has particularly exciting implications for visually impaired Internet users.
