An LLM-Based Multi-Turn Task-Oriented Dialogue System for Machine Learning Algorithm Selection in Data Mining Dr. Ing. h.c. F. Porsche AG

Format:

Book

Conference/Event

Author/Creator:

Hörtling, Stefan, author.

Contributor:

Albers, A.

Bause, Katharina

Conference Name:

2025 Stuttgart International Symposium (2025-07-02 : Stuttgart, Germany)

Language:

English

Physical Description:

1 online resource cm

Place of Publication:

Warrendale, PA SAE International 2025

Summary:

For the systematic application of machine learning during data mining in product development processes, selecting a suitable algorithm is crucial for success. During an empirical study in the automotive industry, a team applying data mining to develop battery systems for battery electric vehicles was accompanied. Here, it could be observed that data mining tasks are often unique during product development processes and can differ in boundary conditions. Depending on these tasks, suitable machine learning algorithms must be selected. Because of the variety of machine learning paradigms, problems, and algorithms, it is often hard to select a suitable algorithm, especially for inexperienced data miners. This paper presents a large language model (LLM)-based, multi-turn, task-oriented dialogue system to support data miners in selecting machine learning algorithms that are suitable for their specific data mining tasks. This approach, called "Algorithm Selection Assistant" (ASA), enables free text communication in natural language and uses an algorithm selection process to analyze the boundary conditions systematically. The user, id est, the data miner, doesn't have to know this process since the ASA guides and leads through the process. Therefore, software is presented based on a web application architecture, including frontend, backend, LLM interfaces to closed- and open-source models, and a description of the algorithm selection process. The execution of the ASA, including the multi-turn steps to select a suitable algorithm, is shown as an example coming with a data mining task out of the empirical study. The ASA combines the advantages of novel LLMs, like generative abilities, contextual understanding, and their implicit comprehensive knowledge base, with a systematic process using a chain of thought and a multi-turn conversation. Thereby, the approach can support data miners in selecting machine learning algorithms that suit their specific data mining task

Notes:

Vendor supplied data

Publisher Number:

2025-01-0289

Access Restriction:

Restricted for use by site license