Youssef Drissi

A perspective view and survey of meta-learning

Different researchers hold different views of what the term meta-learning exactlymeans. The first part of this paper provides our own perspective view in which the goal isto build self-adaptive learners (i.e. learning algorithms that improve their bias dynamicallythrough experience by accumulating meta-knowledge). The second part provides a survey ofmeta-learning as reported by the machine-learning literature. We find that, despite differentviews and research lines, a question remains constant: how can we exploit knowledge aboutlearning (i.e. meta-knowledge) to improve the performance of learning algorithms? Clearlythe answer to this question is key to the advancement of the field and continues being thesubject of intensive research.

Glossary extraction and utilization in the information search and delivery system for IBM technical support

In this paper we describe the practical aspects of extracting and using a glossary for a selected technical domain. We first describe the existing glossary extraction process, as applied to general corpora, and examine its shortcomings in the technical support domain. Then we propose a number of enhancements to it, including focusing the glossary on a selected domain context, providing support for multidomain glossaries, and importing domain-specific dictionaries. We apply our focused-glossary approach to the IBM Technical Support corpus and incorporate resulting glossaries within the information search and delivery system used by IBM Technical Support. We demonstrate the effectiveness of our approach by evaluating the quality of keywords and terms extracted from sample documents with the help of these glossaries.

Towards knowledge management in autonomic systems

The classical definition of knowledge management promises to get the right knowledge to the right people at the right time so they can make the best decision [G. Petrash, 1996]. Autonomic systems, on the other hand, are expected to find and apply the right knowledge for self-managing purposes without human intervention. This article discusses the components to be built around a system to enable self-healing and managing capabilities. These are defined and described in this article as self-knowledge, self-monitoring, self-learning, problem detection, diagnosis, and search and solution components. Interaction of these system components to make knowledge available for self-healing purposes is also discussed.

Supply Chain Scenario Modeler: A Holistic Executive Decision Support Solution

McKesson is Americas oldest and largest healthcare services company. IBM Research developed an innovative scenario modeling and analysis tool, supply chain scenario modeler SCSM, for McKesson to optimize its end-to-end pharmaceutical supply chain policies. Through integrated operations research OR models, SCSM optimizes the distribution network, supply flow, inventory, and transportation policies, and quantifies the impacts of changes on financial, operational, and environmental metrics. The modeling work spawned a roadmap of projects with quantified opportunities, including a new air freight supply chain path, and provided new insights that have been critical to improving McKessons performance as a pharmaceutical industry leader. A structured data model supporting the OR models has provided a basis for additional improvement projects. The model directly links OR modeling results to a detailed profit-and-loss statement by product category for the different supply chain paths that McKesson uses. Since this effort began in 2009, McKesson Pharmaceutical division has reduced its committed capital by more than

Enterprise transformation: an analytics-based approach to strategic planning

1 billion.

A framework for strategic financial risk management

The importance of strategic planning is universally recognized in the business world as an effective approach to enable achievement of enterprise business objectives over long time periods. However, despite the criticality of the task, the strategic planning process often does not take advantage of analytics to support the process in a consistent way. This paper describes a transformation of the strategic planning process for a globally integrated enterprise through development of a planning system that provides a planning framework as well as a set of analytic capabilities to improve both efficiency and effectiveness. The system consists of a hierarchically structured data store, rules that govern relationships among data elements, and an enterprise dashboard with reports and insights supported by analytic capabilities. The hierarchical nature of the planning model can account for the needs and opportunities of individual lines of business while yielding a coherent, executable strategy at the enterprise level. Integrated scenario and sensitivity analysis, enterprise simulation, and other analytic capabilities enable optimal planning in the face of volatile and uncertain market conditions characteristic of a long planning horizon. In addition, the system fosters collaboration among planners and exemplifies the key attributes of a Smarter Planet™ instrumentation, interconnectedness, and intelligence.

Techniques for personalized and adaptive search services

Strategic risks represent the largest challenge for corporate risk management, often due to lack of data or incompatibility with existing financial modeling frameworks. Indeed, industry surveys found that from 2002 to 2012 strategic risks accounted for over 80% of the cases of significant shareholder value loss among Top 1000 companies. To meet the challenge, we propose a practical framework that includes (i) a collaborative environment to build (in an intuitive yet structured fashion) a graph representing the relationships between mitigation actions, risks, and affected drivers of a companys financial model, (ii) a set of rules to transform this graph into a probabilistic model, allowing for the automatic generation of questionnaires to elicit risk information from domain experts, (iii) an aggregation technique to combine opinions of multiple experts, (iv) simulation to quantify the effects of the risks and mitigation actions on key financial outcomes, and to compute, for example, the probability of achieving a given profit target conditional on a portfolio of mitigation actions, (v) an evaluation of mitigation actions in terms of their risk reduction, and (vi) a cost-benefit analysis to help decision makers determine a mitigation investment strategy.