Verbal search engines still can’t interact naturally with the user. A new mechanism has been developed to understand and predict where a conversation between a human and an AI is going.
Thans to personalized AI-based digital assistants, users are increasingly interacting with search engines using conversational search queries, rather than typing questions on the keyboard.
Verbal interaction with AI allows for more targeted, specific and personally relevant questions. In this way, the user aims to get what he or she specifically needs more quickly, by having a real conversation with the online search system.
But how does an AI converse? How does it narrow down the answers to the user’s specific area of interest?
This is the question that information science and engineering are currently focusing on.
From a study by the Department of Information Engineering of the University of Pisa in collaboration with the University of Padua and CNR comes a method to evaluate and predict the effectiveness of Artificial Intelligence systems, such as voice assistants, in carrying on a conversation that satisfies the user's expectations.
Image
Image
"The starting point - explains Nicola Tonellotto, professor of computer engineering at the University of Pisa - is the need for conversational agents to be both effective, that is, to give correct answers, and able to maintain the user's interest during the conversation. Our mechanism is able to give the system immediate feedback on the progress of the conversation, for example by seeing if the range of answers falls outside the area of interest of the person asking the question. We do this by analyzing the mathematical properties and relationships between question-answer models”
The work was awarded the ACM SIGIR 2023 Best Paper Award Honorable Mention at the SIGIR conference (Conference on Research and Development in Information Retrieval), which annually brings together the leading academic and industrial experts in information retrieval, the science behind search engines.
“The study of the interaction between humans and artificial intelligence is part of the activities of FoReLab (forelab.unipi.it), the Department's laboratory that conducts cutting-edge research on technologies for the new paradigm of society and industry 5.0. - continues Tonellotto - Research on an ever-increasing customization of artificial intelligence systems on individual users is in fact destined to have a great impact also in the new industry, for example in the field of design, generating demonstrators starting from textual descriptions, for customer support through automatic responders with customization based on social media, for latest-generation human-machine interfaces, capable of simplifying control and inspection processes, to optimize the management and production chain, and in various other fields of society, such as personalized medicine, where personalized interfaces can be designed that automatically adapt to patients”.
Image
