Large Language Models (LLM)-based systems, i.e. interconnected elements that include an LLM as a central component (e.g., conversational agents), are typically monolithic static architectures that rely on a single LLM for all user queries. However, they often require different preprocessing strategies, levels of reasoning, or knowledge. Generalist LLMs (e.g. GPT-4) trained on very large multi-topic corpora can perform well in a variety of tasks. They require significant financial, energy, and hardware resources that may not be justified for basic tasks. This implies potentially investing in unnecessary costs for a given query. To overcome this problem, a routing mechanism routes user queries to the most suitable components, such as smaller LLMs or experts in specific topics. This approach may improve response quality while minimising costs. Routing can be expanded to other components of the conversational agent architecture, such as the selection of optimal embedding strategies. This paper explores key considerations for integrating routing into LLM-based systems, focusing on resource management, cost definition, and strategy selection. Our main contributions include a formalisation of the problem, a novel taxonomy of existing approaches emphasising relevance and resource efficiency, and a comparative analysis of these strategies in relation to industry practices. Finally, we identify critical challenges and directions for future research.

Data mining algorithms are now able to efficiently deal with huge amount of data. Various kinds of patterns may be discovered and may have some great impact on the general development of knowledge. In many domains, end users may want to have their data mined by data mining tools in order to extract patterns that could impact their business. Nevertheless, those users are often overwhelmed by the large quantity of patterns extracted in such a situation. Moreover, some privacy issues, or some commercial one may lead the users not to be able to mine the data by themselves. Thus, the users may not have the possibility to perform many experiments integrating various constraints in order to focus on specific patterns they would like to extract. Post processing of patterns may be an answer to that drawback. Thus, in this paper we present a framework that could allow end users to manage collections of patterns. We propose to use an efficient data structure on which some algebraic operators may be used in order to retrieve or access patterns in pattern bases.