We are thrilled to announce the publication of our latest research paper in the esteemed journal Discover Computing. This significant work is the result of our fruitful and ongoing research collaboration with Nokia Bell Labs.

The paper, co-authored by HSDSLab's Csaba Kiss and Dr. Marcell Nagy, alongside Péter Szilágyi from Nokia Bell Labs, introduces an innovative method called "Max–Min Semantic Chunking." This novel technique leverages semantic similarity and a sophisticated Max–Min algorithm to extract highly coherent semantic chunks from text.

The primary benefit of this approach is a significant enhancement in the performance of Retrieval-Augmented Generation (RAG) systems, a key area in modern AI development. By improving how text is segmented and understood, "Max–Min Semantic Chunking" paves the way for more accurate and contextually relevant AI-generated responses.

We extend our hearty congratulations to all the authors on this impactful publication! HSDSLab is immensely proud of our long-standing and successful research partnership with Nokia Bell Labs, and we eagerly anticipate the future developments stemming from this innovative line of work.

Read the full paper here: https://link.springer.com/article/10.1007/s10791-025-09638-7