Diego Magro

Multi-modal diagnosis combining case-based and model-based reasoning: a formal and experimental analysis

Integrating different reasoning modes in the construction of an intelligent system is one of the most interesting and challenging aspects of modern AI. Exploiting the complementarity and the synergy of different approaches is one of the main motivations that led several researchers to investigate the possibilities of building multi-modal reasoning systems, where different reasoning modalities and different knowledge representation formalisms are integrated and combined. Case-Based Reasoning (CBR) is often considered a fundamental modality in several multi-modal reasoning systems; CBR integration has been shown very useful and practical in several domains and tasks. The right way of devising a CBR integration is however very complex and a principled way of combining different modalities is needed to gain the maximum effectiveness and efficiency for a particular task. In this paper we present results (both theoretical and experimental) concerning architectures integrating CBR and Model-Based Reasoning (MBR) in the context of diagnostic problem solving. We first show that both the MBR and CBR approaches to diagnosis may suffer from computational intractability, and therefore a careful combination of the two approaches may be useful to reduce the computational cost in the average case. The most important contribution of the paper is the analysis of the different facets that may influence the entire performance of a multi-modal reasoning system, namely computational complexity, system competence in problem solving and the quality of the sets of produced solutions. We show that an opportunistic and flexible architecture able to estimate the right cooperation among modalities can exhibit a satisfactory behavior with respect to every performance aspect. An analysis of different ways of integrating CBR is performed both at the experimental and at the analytical level. On the analytical side, a cost model and a competence model able to analyze a multi-modal architecture through the analysis of its individual components are introduced and discussed. On the experimental side, a very detailed set of experiments has been carried out, showing that a flexible and opportunistic integration can provide significant advantages in the use of a multi-modal architecture.

Selecting Most Adaptable Diagnostic Solutions through Pivoting-Based Retrieval

The aim of the present paper is to investigate a retrieval strategy for case-based diagnosis called Pivoting Based Retrieval (PBR), based on a tight integration between retrieval and adaptation estimation. It exploits a heuristic estimate of the adaptability of a solution; during retrieval, lower and upper bounds for such an estimate are computed for relevant cases and a pivot case is selected, determining which cases have to be considered and which have not. Such a technique has been evaluated on three different domain models and very satisfactory results have been obtained both in terms of accuracy, space and retrieval time

Decomposition strategies for configuration problems

The paper introduces and discusses the notion of decomposition of a configuration problem within the framework of a structured logical approach. The paper describes under which conditions a given configuration problem can be decomposed into a set of noninteracting subproblems and how to exploit such a decomposition, both for improving the performance of the configurator and for supporting interactive configuration. Different kinds of decomposition are considered, but all of them exploit, as much as possible, the explicit representation of the partonomic relations in the language, a KL-One like representation formalism augmented with constraints for expressing complex interrole relations. The paper introduces a notion of boundness among constraints, which is used for formally specifying different types of decomposition. One decomposition strategy aims at singling out the components and subcomponents that are directly related to the constraints put by the users requirements; the configurator exploits such decomposition by first configuring that portion of the product and then configuring the parts that are not related to the users requirements. Another decomposition strategy verifies whether the set of constraints for the product to be configured can be split into a set of noninteracting problems. In such a case the configurator solves the configuration problem by splitting the whole search space into a set of smaller search spaces. Different combinations of these two decomposition techniques are considered, and the impact of the decomposition strategies on the performance of the configurator is evaluated via a set of experiments using the configuration of computer systems as a test bed. The results of the experiments show a significant reduction of the computational effort (both in terms of number of backtrackings and in CPU time) when decomposition strategies are used.

A core reference ontology for the customer relationship domain

Customer Relationship Management (CRM) has emerged as an important strategy that companies should implement in order to build profitable and stable relationships with their customer. The domain of CRM has peculiar characteristics: a CRM strategy is largely independent from the specific market sector, it requires multiple units cooperation, it implies the management of a huge amount of knowledge, it is fruitfully supported by software solutions, and finally it implies the integration of human and machine activities. These characteristics suggest that both companies aiming at implementing an efficient CRM strategy, and software houses offering ICT solutions supporting CRM would take great advantage from a common semantic model of CRM. The main contribution of this paper is thus the proposal of O-CREAM-v2, a core reference ontology lbr the CRM domain, specifically targeted to Small and Medium Sized Enterprises (SME). The design of O-CREAM-v2 has been based on requirements mainly elicited from a domain analysis, which considered the way the involved actors talk about CRM within their business, by analyzing documents and interviews with representatives of SME and ICT companies. Moreover, in order to guarantee accuracy in the definition of the basic concepts and to support interoperability within/between companies, O-CREAM-v2 has been developed within the framework provided by the well-known DOLCE foundational ontology, together with three DOLCE extensions, i.e. the ontology of Descriptions and Situations, the Ontology of Intbrmation Objects and the Ontology of Plans. O-CREAM-v2 is composed by two layers: an upper core, which models more general concepts and relations, and can be useful also in business domains other than CRM, and a lower core, representing concepts and relations specific to the CRM domain. The content requirements defined by the domain analysis pointed out that an ontology for the CRM domain has to account for both particulars (such as activities, otters, sales, etc.) and information about them (customer records, reports about sales, etc.). Moreover, since CRM is typically supported by software tools, O-CREAM-v2 includes the formal characterization of software applications. Thus, O-CREAM-v2 is structured into five modules: Relationships, Knowledge, Activities (all three spanning both the upper and the lower core), Software and Miscellaneous (both limited to the upper core). The five ()-CREAM modules arc described in details in the paper. The discussion is concluded by mentioning two possible exploitation perspectives for O-CREAM-v2, which could be the basis for building: (a) Web-based repositories supporting the mediation between supply and demand of CRM-related tools; (b) tools supporting users in building the formal representations of resources in ontology-based IR systems and in the semantic search engines application field.

The business knowledge for customer relationship management: an ontological perspective

This paper presents some results of an ongoing ontological analysis of the CRM field. In particular, it describes a fragment of O-CREAM, an ontology for CRM based on DOLCE and on other three DOLCE-based modules, i.e. DnS (for the representation of roles and for handling reification), OIO (for modeling information objects, the key concept for representing business knowledge), and OoP (whose notions are used to express the derivation of new business knowledge). Since the business knowledge plays a major role within CRM activities, a significant fragment of O-CREAM is devoted to the formal characterization of notions related to business knowledge; such a fragment is the focus of this paper.

Towards a first ontology for customer relationship management

This paper presents some results of an ongoing project aimed at modeling the main concepts related to Customer Relationship Management (CRM). More precisely, the paper presents O-CREAM, a CRM ontology based on DOLCE and on two DOLCE-based modules, DnS (exploited for modeling roles and for handling reification) and OIO (exploited for modeling business knowledge by means of information objects). The project relies on the belief that all the actors involved in CRM could benefit from an ontological investigation of this field, aimed at providing a core set of formally described concepts and relations, useful both for describing CRM processes and for specifying the functionality of CRM applications. In particular, a well-formed CRM ontology would support communication and interoperability both in intra-organization and in inter-organization CRM processes. The paper discusses in details the axiomatization for the sale and customer relationship concepts, as well as for the corresponding business knowledge items (i.e., sale and customer records). It concludes by sketching a possible concrete exploitation of O-CREAM.

Ontology-driven collaborative annotation in shared workspaces

In the last decade, collaboration and sharing on the Web have become mainstream. Digital, remote interaction happens on a daily basis, not only to share digital resources, but also to create, manage and discuss them, in every possible situation where collaboration is required: from work teams to groups of friends, from community committees to no-profit organizations. In this paper we address the task of collaborative management of digital resources within a team, with a special focus on the task of semantic annotation, where team members, possibly supported by automated reasoning, enrich resources with properties that help in organizing, retrieving and creating connections between contents of different types. We focus in particular on the problem of reaching an agreement on the annotation itself among the participants. The paper presents a qualitative user study aimed at observing users behavior when faced with this task. The results of the study are then analyzed in order to draw guidelines, which are then implemented in a tool for collaborative annotation. This study is carried out in the context of the Semantic Table Plus Plus (Sem T++) Project, a framework supporting collaboration over thematic workspaces, whose goal is to enhance cooperation through awareness, enhanced communication and easy sharing of digital content. Display Omitted Sem T++ is a framework for the collaborative management of shared online resources.Sem T++ supports collaborative semantic annotation, thanks to a formal semantic model.Implementation of collaborative annotation is based on results of a user study.User study analyzes three collaboration policies: consensual, authored, supervised.

Supporting Product Configuration in a Virtual Store

In the present paper we shall analyse the main requirements and the general architecture of a system able to support a customer in configuring a complex product in a virtual store on the Web. Moreover, we shall present the conceptual language adopted for modelling the application domains and we shall sketch the main reasoning mechanisms involved in the interactive configuration process.

Automatically Decomposing Configuration Problems

Configuration was one of the first tasks successfully approached via AI techniques. However, solving configuration problems can be computationally expensive. In this work, we show that the decomposition of a configuration problem into a set of simpler and independent subproblems can decrease the computational cost of solving it. In particular, we describe a novel decomposition technique exploiting the compositional structure of complex objects and we show experimentally that such a decomposition can improve the efficiency of configurators.

COCONF: Conceptual language-based configuration

This article deals with the configuration task. Configuring means assembling a set of predefined components in order to build a composite object that meets a set of requirements. Here we present COCONF, an approach to configuration based on a conceptual encoding of the configuration knowledge, directly exploitable by a software configuration system to compute configurations. In particular, we show how Conflict-Directed Backjumping can be adapted to the proposed framework and in which way the efficiency of the configurator may be still enhanced by two look-ahead mechanisms, which exploit the characteristics of the modeling language and the explicit representation of both the compositional structure and the taxonomic relations among component types. The configuration algorithms are explained in detail; the assumptions and the properties which they are based on are explicitly stated. Formal proofs are provided for the basic properties. A set of experimental results on three different real-world domains are presented, which prove the suitability of the approach.