Finding the original mass - A machine learning model and its deployment for lithic scrapers

Aug 5, 2026

Predicting the original mass of a retouched scraper is a central objective in lithic analysis, tied to how stone-tool morphology, standardization during reduction, use life, and site occupation patterns can be inferred. Previous methods relied on attributes expected to stay stable through retouch, but they showed limited accuracy and were not tested across repeated resharpening. This study trains four machine-learning models on experimentally resharpened flint scrapers, achieving strong performance with Random Forest and deploying it as an open-source Shiny app.

LanguageEnglish

CategoryResearch & Report

FormatPDF

Pages27

Reading Time1 h 8 min

AccessFree to download

Introduction

Reduction model and curation

Why original scraper mass matters

OPEN ACCESS Citation: Bustos-Pérez G (2025) Finding the original mass: A machine learning model and its deployment for lithic scrapers. PLoS One 20(7): e0327597. [https://doi.org/10.1371/](https://doi.org/10.1371/)[ ](https://doi.org/10.1371/)[journal.pone.0327597](journal.pone.0327597) Editor: Enza Elena Spinapolice, Sapienza University of Rome: Universita degli Studi di Roma La Sapienza, ITALY Received: November 18, 2024 Accepted: June 17, 2025 Published: July 28, 2025 Copyright: © 2025 Guillermo Bustos-Pérez. This is an open access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited. Data availability statement: The repository with all code and data for published app is: [https://doi.org/10.5281/zenodo.15771479](https://doi.org/10.5281/zenodo.15771479) [Funding:](Funding: GBP is postdoctoral scientist at)[ GBP is postdoctoral scientist at](Funding: GBP is postdoctoral scientist at)the Max Planck Institute of Evolutionary Anthropology, Department of Human Origins. RESEARCH ARTICLE Finding the original mass: A machine learning model and its deployment for lithic scrapers Guillermo Bustos-Pérez1,2,3* 1 Departament of Human Origins, Max Planck Institute for Evolutionary Anthropology, Leipzig, Germany, 2 Institut Català de Paleoecologia Humana i Evolució Social (IPHES-CERCA), Tarragona, Spain, 3 Universitat Rovira i Virgili, Departament d’Història i Història de l’Art, Tarragona, Spain * [guillermo_bustos_perez@eva.mpg.de](guillermo_bustos_perez@eva.mpg.de) Abstract Predicting the original mass of a retouched scraper has long been a major goal in lithic analysis. It is commonly linked to lithic technological organization of past societies along with notions of stone tool general morphology, standardization through the reduction process, use life, and site occupation patterns. In order to obtain a prediction of original stone tool mass, previous studies have focused on attributes that would remain constant or unaltered through retouch episodes. However, these approaches have provided limited success for predictions and have also remained untested in the framework of successive resharpening episodes. In the research presented here, a set of experimentally knapped flint flakes were successively resharpened as scraper types. After each resharpening episode, four attributes were recorded (scraper mass, height of retouch, maximum thickness and the GIUR index) . Four machine learning models were trained using these variables in order to estimate the mass of the flake prior to any retouch. A Random Forest model provided the best results with an r2 value of 0.97 when predicting original flake mass, and a r2 value of 0.84 when predicting percentage of mass lost by retouch. The Random Forest model has been integrated into an open source and free to use Shiny app. This allows for the wide spread implementation of a highly precise machine learning model for predicting initial mass of flake blanks successively retouched into scrapers. 1. Introduction Scrapers are some of the most common lithic implements among archaeological lithic assemblages. They are present from the first Oldowan stone tools [ 1–3] through to modern ethnographic studies of hunter gatherers [4–7] . The “reduction model” [8 ,9] suggests that some stone tools (including scrapers) can represent different stages of reuse and modification through retouch. When considering scrapers within the reduction model, an integral concept is that of curation. The initial definition of curation included a series of behavioral patterns related to provisioning strategies PLOS One | [https://doi.org/10.1371/journal.pone.0327597](https://doi.org/10.1371/journal.pone.0327597) July 28, 2025 1 / 27 The funders had no role in study design, data collection and analysis, decision to publish, or pr

FAQs

What problem does the study address in lithic analysis?

It targets predicting the original mass of a scraper before retouch, a major goal linked to interpreting technological organization and reuse.

How was the dataset created for training the machine learning models?

Flint flakes were experimentally knapped as scrapers and then repeatedly resharpened; after each episode four attributes were recorded to model pre-resharpening mass.

Which model performed best and how is it provided to users?

Random Forest gave the best results (r2=0.97 for original mass and r2=0.84 for percentage mass lost) and was integrated into an open-source Shiny app.

Aug 5, 2026

Upload by Evangeline