Can algorithms and artificial intelligences read? The answer, thanks to the latest developments in the semantic analysis applied to software programming, is yes. Not only do they read and understand written words, but they also know how to listen and decipher the spoken language. A skill that is fascinating the political world.
During his campaign as a Secretary of the Democratic Party candidate, Nicola Zingaretti relied on nCore HR, a technology platform developed for recruiting and managing human resources that the Secretary used to monitor the voters' sentiment through video analysis. Founded at the end of 2017 as a solution to simplify the process of applying and recruiting, nCore is the fifth company founded by Enrico Ariotti and his partners.
“From cognitive interviews to feedback from the company, our solution differs from others in terms of being comprehensive, since it manages both the application and response to the candidate phase. And among its features is semantic analysis of video interviews, a tool that is increasingly widespread especially in large companies,” they explain. nCore then looks at the words used and what they convey to carry out a profiling of respondents who are then organized into groups based on the responses provided. In addition to words, from the video, nCore HR can also infer other non-verbal information such as estimated age and gender. The aim is to go beyond C.Vs., which, additionally, are also the subject of semantic analysis, in order to identify what are the key words in the experience of the candidate. All while providing an agile and intuitive user experience.
Application fields can range from open innovation programs to crowdfunding, through analysis of social texts to medical records in hospital centers
Semantic analysis, however, is not just a tool to find the ideal resource to position in the organization chart. “There are several applications for this practice. Thanks to the flexibility in the definition of the questions and the information required, the fields of application can vary from open innovation programs to crowdfunding, from the analysis of social texts to medical records in medical centers” Enrico Ariotti explains.
It can be no coincidence, therefore, that several solutions are already in use. In the medical field, for example, the Azienda Ospedaliera – University of Modena received an award by PoliMi for the Analisi Semantica Referti Ambulatoriali per la Codifica delle Prestazioni (ASRACoP) – or Semantic Analysis Outpatient References for Performance Encoding- project, which involves the creation of software to analyze outpatient reports and support doctors in coding prescribed services.
As for the publishing industry, the multinational ADmantX has launched Home Page Intelligent, which allows an increase in advertising revenue by making home pages more accessible to advertisers and improving the profiling of their own readers.
From a historical and scientific point of view, the UN Global Pulse study of the United Nations used the machine learning software and the archive of speeches given before the Assembly in the 1970-2015 period to understand whether a machine could replace a person on specific topics. The result? In 90% of cases, the texts generated were comparable to those written by humans.
In regards to services to the citizen, semantic analysis has been used by the Emilia-Romagna Region and Lepida (house organ of the Region) to facilitate communication with the PA by focusing on three strands: the development of assistance tools and the support to citizens in the use of services (such as chatbots); improvement of the application service by companies and reduced investigative time; improvement of document management through categorization and optimization functions.