Technology and the online universe are constantly reinventing themselves.The digital earthquake revolutionizes itself several times a year.However, the roots of this online and hyper-communicated world are in the past; and a series of recurring ideas are repeated since then. The semantic web is one of them .This is how it is helping artificial intelligence and voice control systems.
The language, as natural as possible
"The voice is natural, it is our way of communicating.We have been using it for hundreds of thousands of years.And now we are using it to change the way we exchange information with machines." Related
For Max Amordeluso, lead evangelist of Amazon, the future of the conversation with the machines goes through talking with them just as we do with other human beings .It is true that Amazon's most revolutionary product, Alexa, combines voice and artificial intelligence to try to (re) evolve the way we talk with our devices, but that does not downplay the statement of Amordeluso.
"The big problem-continues the Amazon executive in a talk during 4YFN 2018 in Barcelona-is that the machines are not good at disambiguation" .From small, human beings (Spanish speakers) we learn that if our parents ask us to buy a bar, they probably won't refer to the bar.Or if they tell us to feed the cat, they probably don't want us to look for anything in the garage.(or feed a madrileno).We use the context to specify the meaning of the words and it hardly costs us any effort.
For machines, however, it is a complicated task.If voice recognition systems have to learn to understand the so-called natural language, and, in fact, they are already doing it, they need to process much more information that the meaning of the word in the dictionary .Here is where an ambitious idea fallen into disuse is throwing a cable to developers and programmers.
Web 3.0 that never arrived (or did)
It was the year 2001 when Internet father Tim Berners-Lee, along with James Hendler and Ora Lassilla, baptized the semantic web in the Scientific American magazine .He named it like that for an idea that had been around the computer world for years, as old as the first proposals of hypertexts developed in the 40s of the 20th century, during World War II.
Google or Facebook bet on semantic databases to improve their products; and Cortana, Siri or Alexa have assumed the principles of the semantic web to learn human language
The semantic web, also known as web 3.0 , gathers a series of technologies to publish data readable by computer applications, by machines .This data includes semantic and ontological data that describe the context, content and relationship between the data itself.The main problem with this idea is that, today, it is impossible to adapt all Internet content to be semantic In addition, the volume of data generated by each online document is triggered.
Thus, the semantic web did not end up being as disruptive as it was intended.Its name was lost in the mists of the digital revolution.And, nevertheless, semantic technologies are the order of the day in the main search engines such as Google, Bing or Baidu, social networks, artificial intelligence systems and data analysis such as Watson (IBM) or, as we have seen, voice control platforms, such as Alexa, Cortana (Microsoft) or Siri (Apple).
The renaissance of the semantic web
"The good ones news is that today there is evidence that the semantic web is emerging rapidly .Far from the ambitious idea of a beginning, but through small versions of itself and intelligent assistants, "he explains Peter Sweeney, founder of the Primal artificial intelligence company, in an article published in Medium.
The key is in the need for the machines to understand the natural language .That is, to understand our orders, even if they are full of ambiguities and colloquial structures.For this the concept of linked data or linked data , a more affordable evolution of the semantic web.This technology focuses on creating open databases , distributed and decentralized interconnected with each other.
"Linked data opens the door for machines to combine information in very complex ways," Sweeney continues.More complex information, more context, more nuances and better understanding.And that's how an idea fallen into disuse has permeated the revolution of artificial intelligence and has reinforced our communication with the machines, helping them understand, increasingly, our imperfect and ambiguous language.
Talking to the coffee maker
More and more people connected and chatting through the Internet and billions of objects doing the same thing.This is the present of the network.In the near future, the numbers seem not to stop growing.In this world of unimaginable amounts of data, artificial intelligence will become increasingly intelligent. The more information the machines and their algorithms receive, the more efficient they will become .
The semantic web or, better, the technologies derived from this concept and the linked databases will gain strength.The volume of linked data has not stopped increasing in recent years: Giants like Google or Facebook bet on the semantic databases to improve your products ; and smart assistants like Cortana, Siri, Google Now or Alexa have assumed the principles of the semantic web to learn human language .
Also, there seems to be no going back."Let's think about one thing.Today, we already have a complete generation of young people who are getting older assuming that they will always be able to talk to their devices , talk to your house, "reflects Max Amoderluso, from Amazon.Who knows if there will come a day when who knows best to listen to us is our coffee maker.Will we remember then the semantic web?
Images | iStock, Pixabay, Wikimedia Commons
Comments
Post a Comment