The project VVU (Ubiquitous Virtual Seller) initiated by PICOM (the competitivity institute on Trading Industries) aims at ameliorating the integration of a conversational agent in commercial trading websites in order to improve their transformation ratio (sales/visitors). The main goal is to adapt interactions to particular selling situations. The project, carried out for the University Lille 1, by the SMAC team (Multi-Agents Systems and Behaviors) of LIFL (the Fundamental Computer Science Laboratory), will be done in close collaboration with, among other, Artificial Solutions, 3 Suisses International and Oxylane Group.
Our objective is the construction of a selling ontology and the reasoning of virtual seller agents in selling situations defined by the project. This computer science research is related to agents technologies, ontological engineering, knowledge reasoning (in particular argumentation logic). We aim at to developing, implementing, evaluating and validating our approach through software prototypes. These prototypes will be evaluated by the non-academic partners of the project.
Under construction
Under construction
MARGO.sf.net is an argumentation-based engine for decision making. Since MARGO argues over multiple motivations which may not be fulfilled all together by a set of non-conflicting decisions, high-ranked goals must be preferred to low-ranked goals and MARGO is prepared to concede. Different qualitative preferences are attached to the motivations. These concrete data structures consist of information providing the backbone of arguments. Due to the abductive nature of decision making, we built arguments by reasoning backwards, possibly by making suppositions. Moreover, arguments are defined as tree-like structures. In this way, MARGO evaluates the possible decisions, suggests some solutions, and provides an interactive and intelligible explanation of the choice made.