As misunderstood as it is for the uninitiated, a computer’s understanding of natural language is not new. The first attempts and theories in this field began in the 1960s, shortly after the invention by John McCarthy of the term Artificial Intelligence. It proves, if necessary, the essence of the analysis of language in the progress of robotics, and this, from its beginnings. It is still the center of research on artificial intelligence.
Language, innovation in everyone’s mouth
Connected objects, social media, digital communications, data are everywhere and more and more. From 2 zettabytes in 2010, the volume of global data has grown to 64 zettabytes in 2020, and forecasts predict an exponential increase until 2025, at least. The two big issues related to it are, of course, its treatment and storage.
However, with much of this data tied to an idiomatic dimension, the fundamental nature of progress in language comprehension is indisputable. Its complexity too. In fact, the NLU applies to a variety of data sources. A challenge for Wacim Belblidia, Head of Data Science at Illuin Technology, a French startup specializing in artificial intelligence, particularly in understanding and processing natural language (NLU / NLP) for companies: ” A company’s textual data can be found in many media: emails, documents, instant messaging, business tools, and more. And there are also large volumes of voice data from numerous telephone exchanges, meetings, videos, and so on. which are rarely captured in a structured way. The NLU is the center of the exploitation of all this data, either to automate conversations with a chatbot or a callbot, to extract information from documents such as invoices, contracts, receipts or to fill a CRM directly from a voice recording . or live during a telephone conversation, for example. »
Behind the acronym NLU is an ever-evolving ecosystem and value chain, both in terms of research and development and in the design of innovative products. In addition, it is at the crossroads of these two fields where Illuin intervenes, specifies its co-founder and CEO Robert Vesoul: Our DNA lies on the border between research and industry. In these markets it is essential to find the right balance between technological innovation and product application. As proof of this ambition, Robert Vesoul is mainly the co-director of CentraleSupélec’s Chair in Digital Innovation. A way to capitalize on an academic anchor that, as is often the case, is fertile ground for innovation.
For people, for people
The brains of tomorrow are, therefore, the key to technological progress around the understanding of natural language. To emphasize the technological dimension and the importance of data in contemporary societies, we would almost forget that it is the human being who is both the creator and the recipient of these innovations as well as their uses.
However, we are perhaps the most essential factor, as it conditions the whole approach of the field. The main function of the NLU is to help humans (and their organizations) in their productivity and efficiency. Wacim Belblidia insists: Thanks to advances in data analytics, architecture, and resulting algorithms, our software is supplanting human performance. However, it is very rare to entrust a decision to artificial intelligence. It is above all a tool for analysis and decision support. »
The applications, although at the beginning of their implementation, are legion in all business functions. At the heart of the customer relationship, omnichannel robots capable of understanding all the subtleties of natural language allow self-service to be extended to a multitude of new situations for maximum customer satisfaction. And for call center agents, real-time voice flow analysis technologies are revolutionizing the work relationship by freeing them from repetitive, low-value-added tasks such as typing. In the HR departments, the automation of administrative tasks (vacations, payroll, etc.) but also the automated analysis of resumes or the generation of interview reports allows to focus more on the experience of the candidate and of employees. An overview of the proliferation of possible use cases in all areas of business.
What do you have in common? They all involve a symbiosis between the value of technology and its contribution, on the one hand, and business expertise or the understanding of implicit, human-specific elements, on the other. A way of remembering that human labor and machine labor are not opposites.
A matter of European technological sovereignty
The fact that the machine learning which is a key variable in the progress of NLU technologies, reinforces this observation. By setting up artificial intelligence so that it can constantly learn and improve as it is used, it becomes, at every stage, more useful to humans, but in no way more favorable to replace it. Choosing a good NLU tool therefore means acquiring technology at the service of equipment, capable of processing, structuring and analyzing data in a much more efficient and productive way than humans, thus reorienting its missions into high value-added tasks.
Around the challenges of storing, classifying and processing colossal volumes of data in the service of understanding natural language, the place of Europe and France on the world map of technological innovation is also at stake. Robert Vesoul recalls the need for European sovereignty: ” Companies must now ask the question of the trajectory of their data. The United States and China in particular are investing heavily. However, the culture of technology is not the same, the core values of the tools designed do not always fit well with European uses, especially in matters of security or data protection. It is important to be able to offer an alternative. »
Following the European technology giants, Illuin is positioning itself as an additional advocate of a strong and voluntary Europe in terms of technology investment. An ambition that is even more fundamental because data at the service of people, concludes Robert Vesoul, is without a doubt “The key capital of today’s business “.
Maddyness, media partner at Illuin Technology.