When it comes to providing information services and managing relationships with users, the need for simple and personalised automated systems is nowadays widely acknowledged. This need arises from two main assertions:
1) Automated systems are â€œcomplicated to useâ€
2) Automated systems â€œnever provide the answers you needâ€
For this reasons and with a view to reducing this resistance to automated systems, Key2 proposes, as an innovative solution, the development of more â€œhumanâ€ automated systems called Virtual Assistants. In this way, the issue of complexity is solved through the use of more natural instruments that offer users the option to interact with human-like avatars which communicate using natural or semi-natural language.
The Virtual Assistant acts as a liaison between the â€œknowledge-baseâ€ and end users, to whom it is dedicated, thus assisting users in their search for information, and often in defining the need itself, which sometimes cannot be or is not clearly formulated.
Data Access Layer (DAL) access to multiple data sources via Web Services.
Business Logic Layer (BLL) Data management is carried out based on the logic of Bayesian filters.
The recurrence of responses considered exhaustive by users allows the system to handle probability based algorithms: a certain type of responses are matched to a given question received. This procedure allows for an intelligent evolution of the system based on previous choices made by users, thus over time the system improves its searching capabilities. Results are based not only on â€œkeywordsâ€, but also on the logic of â€œrelevanceâ€ in relation to the search. Through a system that â€œranksâ€ the relevance of occurrences over time, the conceptual links between data are identified irrespective of the existence of an exact correspondence with the searched â€œkey wordsâ€.
User Interface (UI )
The way results are presented is straightforward and users have no difficulty in understanding them helped by an intuitive and guided interactivity. The application is designed for web, desktop and mobile environments.
SHARE THIS PROJECT