Amélie Cordier

Engineering and learning of adaptation knowledge in case-based reasoning

Case-based reasoning (CBR) uses various knowledge containers for problem solving: cases, domain, similarity, and adaptation knowledge. These various knowledge containers are characterised from the engineering and learning points of view. We focus on adaptation and similarity knowledge containers that are of first importance, difficult to acquire and to model at the design stage. These difficulties motivate the use of a learning process for refining these knowledge containers. We argue that in an adaptation guided retrieval approach, similarity and adaptation knowledge containers must be mixed. We rely on a formalisation of adaptation for highlighting several knowledge units to be learnt, i.e. dependencies and influences between problem and solution descriptors. Finally, we propose a learning scenario called “active approach” where the user plays a central role for achieving the learning steps.

Taaable: a Case-Based System for personalized Cooking

TAAABLE is a Case-Based Reasoning (CBR) system that uses a recipe book as a case base to answer cooking queries. TAAABLE participates in the Computer Cooking Contest since 2008. Its success is due, in particular, to a smart combination of various methods and techniques from knowledge-based systems: CBR, knowledge representation, knowledge acquisition and discovery, knowledge management, and natural language processing. In this chapter, we describe TAAABLE and its modules. We first present the CBR engine and features such as the retrieval process based on minimal generalization of a query and the different adaptation processes available. Next, we focus on the knowledge containers used by the system. We report on our experiences in building and managing these containers. The TAAABLE system has been operational for several years and is constantly evolving. To conclude, we discuss the future developments: the lessons that we learned and the possible extensions.

A Model and Simulation of Early-Stage Vision as a Developmental Sensorimotor Process

Theories of embodied cognition and active vision suggest that perception is constructed through interaction and becomes meaningful because it is grounded in the agent’s activity. We developed a model to illustrate and implement these views. Following its intrinsic motivation, the agent autonomously learns to coordinate its motor actions with the information received from its sensory system. Besides illustrating theories of active vision, this model suggests new ways to implement vision and intrinsic motivation in artificial systems. Specifically, we coupled an intrinsically motivated schema mechanism with a visual system. To connect vision with sequences, we made the visual system react to movements in the visual field rather than merely transmitting static patterns.

Failure Analysis for Domain Knowledge Acquisition in a Knowledge-Intensive CBR System

A knowledge-intensive case-based reasoning system has profit of the domain knowledge, together with the case base. Therefore, acquiring new pieces of domain knowledge should improve the accuracy of such a system. This paper presents an approach for knowledge acquisition based on some failures of the system. The cbr system is assumed to produce solutions that are consistent with the domain knowledge but that may be inconsistent with the expert knowledge, and this inconsistency constitutes a failure. Thanks to an interactive analysis of this failure, some knowledge is acquired that contributes to fill the gap from the system knowledge to the expert knowledge. Another type of failures occurs when the solution produced by the system is only partial: some additional pieces of information are required to use it. Once again, an interaction with the expert involves the acquisition of new knowledge. This approach has been implemented in a prototype, called FrakaS , and tested in the application domain of breast cancer treatment decision support.

Similarity Measures to Compare Episodes in Modeled Traces

This paper reports on a similarity measure to compare episodes in modeled traces. A modeled trace is a structured record of observations captured from users’ interactions with a computer system. An episode is a sub-part of the modeled trace, describing a particular task performed by the user. Our method relies on the definition of a similarity measure for comparing elements of episodes, combined with the implementation of the Smith-Waterman Algorithm for comparison of episodes. This algorithm is both accurate in terms of temporal sequencing and tolerant to noise generally found in the traces that we deal with. Our evaluations show that our approach offers quite satisfactory comparison quality and response time. We illustrate its use in the context of an application for video sequences recommendation.

Knowledge continuous integration process (K-CIP)

Social semantic web creates read/write spaces where users and smart agents collaborate to produce knowledge readable by humans and machines. An important issue concerns the ontology evolution and evaluation in man-machine collaboration. How to perform a change on ontologies in a social semantic space that currently use these ontologies through requests? In this paper, we propose to implement a continuous knowledge integration process named K-CIP. We take advantage of man-machine collaboration to transform feedback of people into tests. This paper presents how K-CIP can be deployed to allow fruitful man-machine collaboration in the context of the Wikitaaable system.

Inverting the Interaction Cycle to Model Embodied Agents

Cognitive architectures should make explicit the conceptual begin and end points of the agent/environment interaction cycle. Most architectures begin with the agent receiving input data representing the environment, and end with the agent sending output data. This paper suggests inverting this cycle: the agent sends output data that specifies an experiment, and receives input data that represents the result of this experiment. This complies with the embodiment paradigm because the input data does not directly represent the environment and does not amount to the agents perception. We illustrate this in an example and propose an assessment method based upon activity-trace analysis.

Contextual trace-based video recommendations

People like creating their own videos by mixing various contents. Many applications allow us to generate video clips by merging different media like videos clips, photos, text and sounds. Some of these applications enable us to combine online content with our own resources. Given the large amount of content available, the problem is to quickly find content that truly meet our needs. This is when recommender systems come in. In this paper, we propose an approach for contextual video recommendations based on a Trace-Based Reasoning approach.

Constructing Phenomenal Knowledge in an Unknown Noumenal Reality

Abstract In 1781, Immanuel Kant argued that cognitive agents ignored the underlying structure of their world “as such” (the noumenal reality ), and could only know phenomenal reality (the world “as it appears” through their experience). We introduce design principles to implement these theoretical ideas. Our agents input data is not a direct function of the environments state as it is in most symbolic or reinforcement-learning models. The agent is designed to discover and learn regularities in its stream of experience and to construct knowledge about phenomena whose hypothetical presence in the environment explains these regularities. We report a proof-of-concept experiment in which the agent constructs categories of phenomena, and exploits this knowledge to satisfy innate preferences. This work suggests a new approach to cognitive modeling that focuses on the agents internal stream of experience. We argue that this approach complies with theories of embodied cognition and enaction.

How Case-Based Reasoning on e-Community Knowledge Can Be Improved Thanks to Knowledge Reliability

This paper shows that performing case-based reasoning (CBR) on knowledge coming from an e-community is improved by taking into account knowledge reliability. MKM (meta-knowledge model) is a model for managing reliability of the knowledge units that are used in the reasoning process. For this, MKM uses meta-knowledge such as belief, trust and reputation, about knowledge units and users. MKM is used both to select relevant knowledge to conduct the reasoning process, and to rank results provided by the CBR engine according to the knowledge reliability. An experiment in which users perform a blind evaluation of results provided by two systems (with and without taking into account reliability, i.e. with and without MKM) shows that users are more satisfied with results provided by the system implementing MKM.