Alapfogalmak
The EBER chatbot combines Artificial Intelligence Modeling Language, Natural Language Generation, and Sentiment Analysis to create an "intelligent radio" that reads news and engages elderly users in short, empathetic dialogues to improve their information access and abstraction capabilities.
Kivonat
The paper presents the design and implementation of the EBER chatbot, which aims to improve digital inclusion and information access for elderly users. The key aspects are:
News Service:
EBER reads a variety of news topics that may interest the elderly, such as accessibility, environment, health, leisure, public services, retirement, social services, sport, technology, and transport.
The news is obtained from the Spanish National Radio and Television (RTVE) API.
Natural Language Generation (NLG) Module:
A three-stage NLG module generates coherent, grammatically correct responses based on keywords extracted from user utterances and news content.
Sentiment Analysis (SA) Module:
A three-stage SA module classifies user responses into positive, negative, or neutral sentiment.
The SA knowledge is used to adapt the chatbot's responses and facial expressions to the user's mood, creating a more empathetic interaction.
Chatbot Design:
The chatbot personality is designed using Artificial Intelligence Markup Language (AIML) to maintain a controlled dialogue flow with short questions about daily routines and mood.
Accessibility features include voice-based interaction, simple graphics, and visual cues to guide the user.
The chatbot aims to engage the user by alternating news readings with short, empathetic dialogues.
The experiments involved 31 elderly users and showed that the system was able to improve the users' information abstraction capabilities, even for those with limited technological skills. The analysis of user behavior and satisfaction scores validated the effectiveness of the "intelligent radio" approach in reducing the digital gap for the elderly.
Statisztikák
The system was tested with 31 elderly users (20 women and 11 men), with an average age of 75.5 ± 6.95 years.
10 users had some basic technology skills, 8 had hearing problems, 23 were highly focused during the experiments, and 5 felt awkward or stressed during the interactions.
Idézetek
"The understanding of the bot."
"The system moves its eyes and mouth to show a frame of mind."
"The systems detects the mood of my opinions."
"The chatbot understands me and adapts itself to my answers."