Natural language processing with machine learning
Natural language processing is not very distant from creating a robot like Jarvis (Just A Rather Very Intelligent System)from the movie Iron Man
Latest News »
- Telcos woes: Ministerial panel to submit recommendations to Telecom Commission in July
- AAP changes tack: No direct attack on PM Modi, slams BJP instead
- Ivanka Trump must testify in shoe knockoff suit against company, says US court
- Hopes fade in China for 118 still missing day after landslide
- Western, southern states control over 60% of bank credit: RBI
There have been numerous examples over the last two decades of how Natural Language Processing, or NLP, is being used by companies to provide an intelligent voice to gadgets and searches. Think, for instance, how the world of search engines—from Yahoo, Microsoft and Google—have changed the Internet with text-based search algorithms driving and augmenting the World Wide Web. NLP, though, does much more than just that and text analytics. NLP exploration on our current digital planet includes voice searches on automobiles and then, of course, the dictation mechanics of the software world.
Voice recognition and speech-to-text
I have an 18-month-old who thrives on YouTube searches asking for ‘Peppa Pig’ or ‘Mickey Mouse’ series while my 5-year-old is exploring the world of content on YouTube (of course, with restricted parental control)—from watching the world of KungFu Panda to how to make dummy videos.
NLP became a popular framework when Apple Inc. introduced Siri and Google introduced its YouTube voice search. Apple’s NLP on voice-to-text in Siri works in 16 global languages and it is meant for languages from English to French to Chinese.
This has had an impact on the iPhone community when it comes to uses of searches, looking for your content in your phone, making life easier, lazier, yet efficient. Multilingual internationalisation exists for YouTube, too.
Impact of social networks and social listening
With the ongoing growth of the World Wide Web and exponential growth of social network platforms, text analytics of billions of data are required by marketeers, traders and customer services team and political verticals to understand the sentiment of the general public.
Understanding the emotions of all the demographics and text analytics leap frogged into what we now focus on social listening with numerous tools including Radian 6 and Sprinkler. The world is analysing petabytes of text everyday with text-based analytics to understand social networks.
With the advancement of Google and Siri, listening, combined with social listening, is a data scientist’s dream, following which the vertical of management information systems to data warehouses is seeing a ten-fold increase.
Data scientists are becoming much more sought-after and the thin red line between Big Data analytics and text analytics is blurring.
NLP with machine learning
A strong case for NLP is emerging globally and in India amid the smart cities ecosystem. Realtors and builders are building condos and apartments which includes machine learning to understand spoken commands. So a Star Trek, or an I, Robot, is becoming a reality. You talk and your commands based on the lexicon usage of your search takes the command and reacts to your demand.
Machines being trained on voice search will command the next generation of human interface in our day-to-day lives, commanding your washing machine to start, or instructing your music system to begin your favourite track.
Smart cities in India by top realtors and the Internet of Things firms are trailing these out.
Firms like GE and Siemens are championing this space in Nordic countries and Western Europe. SAP’s acquisition of Plat One and Telit will augment the next generation of NLP to artificial intelligence.
But India has 22 languages...
India has always been a research hub of NLP; organizations such as Owler are great but the mantle to crack NLP for 22 languages by voice to text and augment voice commands to machines via key lexicon datasets on universal remotes and many more use cases is taken by an organization called Mihup Communications.
If Mihup, funded by Accel Partners, can crack 22 languages and ensure that machines and device commands get augmented through their algorithms, then India leapfrogs as a sub-continent on NLP innovation.
Uneducated farmers talking on platforms on crop remedy, tackling farmer suicides is a never ending opportunity and a company like Mihup intends to crack this issue. I hope we soon have more such companies to spur innovation on language processing for machines.
We live in a world where one sees movies like Iron Man where Tony Stark talks to his Just A Rather Very Intelligent System, or Jarvis. I believe the world of NLP will soon match this expectation.
Sauvik Banerjjee, Global CTA & Innovation Lead (Omni Channel, IoT, Industry 4.0 and AI), Digital Business Services, Asia Pacific & Japan, SAP Asia Pte Ltd.