Claudia Diamantini

Ontology-Driven KDD Process Composition

One of the most interesting challenges in Knowledge Discovery in Databases (KDD) field is giving support to users in the composition of tools for forming a valid and useful KDD process. Such an activity implies that users have both to choose tools suitable to their knowledge discovery problem, and to compose them for designing the KDD process. To this end, they need expertise and knowledge about functionalities and properties of all KDD algorithms implemented in available tools. In order to support users in this heavy activity, in this paper we introduce a goal-driven procedure for automatically compose algorithms. The proposed procedure is based on the exploitation of KDDONTO, an ontology formalizing the domain of KDD algorithms, allowing us to generate valid and non-trivial processes.

Bayes Vector Quantizer for Class-Imbalance Problem

The class-imbalance problem is the problem of learning a classification rule from data that are skewed in favor of one class. On these datasets traditional learning techniques tend to overlook the less numerous class, at the advantage of the majority class. However, the minority class is often the most interesting one for the task at hand. For this reason, the class-imbalance problem has received increasing attention in the last few years. In the present paper we point the attention of the reader to a learning algorithm for the minimization of the average misclassification risk. In contrast to some popular class-imbalance learning methods, this method has its roots in statistical decision theory. A particular interesting characteristic is that when class distributions are unknown, the method can work by resorting to stochastic gradient algorithm. We study the behavior of this algorithm on imbalanced datasets, demonstrating that this principled approach allows to obtain better classification performances compared to the principal methods proposed in the literature.

Quantizing for minimum average misclassification risk

In pattern classification, a decision rule is a labeled partition of the observation space, where labels represent classes. A way to establish a decision rule is to attach a label to each code vector of a vector quantizer (VQ). When a labeled VQ is adopted as a classifier, we have to design it in such a way that high classification performance is obtained by a given number of code vectors. In this paper we propose a learning algorithm which optimizes the position of labeled code vectors in the observation space under the minimum average misclassification risk criterion.

Semantic enrichment of strategic datacubes

In the information system view, the reference architecture for strategic and decision support is based on the Data Warehouse architecture, that enables flexible and multidimensional analysis of strategic indexes by means of OLAP tools and reports. In this paper we propose a novel model for semantic annotation of Data Warehouse schema that takes into account domain ontologies as well as a mathematical ontology. Such an ontology describes mathematical formulas underlying elements of the datacube schema, including the semantics of operands and operators. In particular, we discuss and apply the proposed model for the semantic annotation of the schema of a datacube, that is the basis for OLAP analysis and contains information derived from Data Warehouse schema. In the paper, an illustrative case study together with some examples of analysis based on this kind of annotation are provided.

A conceptual indexing method for content-based retrieval

A great deal of work has been done to define index structures to support feature-based similarity queries. However, other kinds of content-based retrieval, namely keyword-based and concept-based, are founded on different properties of the data space, which make these methods ineffective. Nevertheless, similarity notions are still needed, in order to manage incompleteness and imprecision in the representation of multimedia data, as well as in user query specification. In the paper, we present an indexing method which is based on partitioning the data space. We introduce the binary counterpart of the notions of minimum volume and minimum overlap, and combine them in a global hierarchical clustering criterion. We also show how the index structure induced by the clusterization can be exploited to deal with incompleteness and imprecision expressed in terms of answer precision and recall.

A Logic-Based Formalization of KPIs for Virtual Enterprises

Open innovation is gaining increasing interest as a model to foster innovation through collaboration and knowledge sharing among organizations, especially in the context of Virtual Enterprises (VE). One of the main issues to overcome in such distributed settings is the integration of heterogeneous data, and the need to evaluate common Key Performance Indicators (KPI) capable to measure overall performances of the VE. In this paper we propose a conceptualization of KPIs into an ontology, to provide a common vocabulary to semantically annotate data belonging to different organizations. KPIs are described in terms of dimensions and a mathematical formula. In order to support reasoning services over KPIs formulas we refer to a logic-based formalization in Prolog, where formulas are translated as facts, and several predicates are included to support both mathematical functionalities for formula manipulation and higher-level functions especially suited for VE setup.

Mining usage patterns from a repository of scientific workflows

In many experimental domains, especially e-Science, workflow management systems are gaining increasing attention to design and execute in-silico experiments involving data analysis tools. As a by-product, a repository of workflows is generated, that formally describes experimental protocols and the way different tools are combined inside experiments. In this paper we describe the use of the SUBDUE graph clustering algorithm to discover sub-workflows from a repository. Since sub-workflows represent significant usage patterns of tools, the discovered knowledge can be exploited by scientists to learn by-example about design practices, or to retrieve and reuse workflows. Such a knowledge, ultimately, leverages the potential of scientific workflow repositories to become a knowledge-asset. A set of experiments is conducted on the my Experiment repository to assess the effectiveness of the approach.

Semantic Annotation and Services for KDD Tools Sharing and Reuse

Active KDD research groups typically make their software tools at disposal of others through the net. However, integration and reuse of these tools typically require a considerable amount of time to understand software scope and use, install it, transform data in a format compatible with the required input. This paper introduces a semantic based, service-oriented framework for tools sharing and reuse, giving advanced support for the semantic enrichment through semantic annotation of KDD tools, deployment of the tools as Web services and discovery and use of such services. A concrete implementation of the framework by Web service technologies is presented.

A Semantic Framework for Knowledge Management in Virtual Innovation Factories

Knowledge management is a crucial aspect for enterprises that want to effectively cope with business innovation. However, the full control of the knowledge asset is often missing due to the lack of precise organizational models, policies, and proper technologies, especially in Virtual Enterprises VEs, which are characterized by heterogeneous partners with different policies, skills and know-how. For such reasons, the need for technologies that enable knowledge sharing, efficient access to knowledge resources, and interoperability is felt as primary. This work proposes a semantics-based infrastructure aimed at supporting effective knowledge management for business innovation in VEs. Knowledge resources are formally represented and stored in a semantic layer, which is exploited by a set of semantic services for enabling efficient retrieval and reasoning capabilities to derive additional knowledge.

SemPI: A semantic framework for the collaborative construction and maintenance of a shared dictionary of performance indicators

Abstract Collaboration at strategic level entails the sharing of Performance Indicators (PIs) to measure the achievement of common objectives and evaluate performances. PIs are synthetic measures calculated starting from transactional data. Given their compound nature, it is difficult to achieve an agreement on their definitions and heterogeneities arise that make sharing and exchange a difficult task. Semantic techniques can help to address these challenges by providing a common layer of formal definitions and automatic reasoning tools to maintain its consistency. In this paper, we present a novel semantic framework for representing Performance Indicators that supports the construction and maintenance of a minimal and consistent dictionary. The distinctive feature of the approach is the logical representation of formulas defining PIs, allowing to make algebraic relationships among indicators explicit, and to reason over these relationships to derive PI identity and equivalence and to enforce the overall consistency of the dictionary. We also present a web application implementing the framework for collaborative construction and maintenance of the dictionary. We provide experimental evidence of the efficiency and effectiveness of the approach on synthetic and real data.