The Matrix: connecting and re-using digital records of archaeological investigations - EAA2022 conference

Editor's Notes

• #2: This paper will present work undertaken by The Matrix project [AH/T002093/1] which is addressing some of the current problems caused by the lack of standardized approaches to analysis and digital archives of archaeological stratigraphic and phasing data. The Matrix project (AHRC AH/T002093/1) is investigating how digital data from archaeological excavations can be made more consistent and useful thereby more interesting and cost-effective to a range of users and audiences. It is working towards a shared plan and methods to get such data more consistently recorded, analysed, disseminated and archived in a way that is Findable, Accessible, Interoperable and Re-useable (FAIR).   The Matrix project has four key areas of activity: 1) Digital Standards; 2) Characteristics of digital Heritage Data; 3) Stratigraphy Standards; 4) Research Tools;   Areas for investigation in the project include enhancing the recording of implicit and explicit spatio-temporal and temporal relationships in the digital stratigraphic records of archaeological investigations and the use of LOD vocabularies to deduce or make explicit dating and phasing records derived from temporal Periods and/or other types of reference data (e.g. LOD site types and periods from HeritageData.org). The project also reviews current stratigraphic matrix analysis activities, including stratigraphic and temporal recording and analysis processes, digital stratigraphic data archiving, and requirements for reuse of archaeological data by chronological modellers.   The Matrix project aims to address 2 major research questions: 1. How can we encourage the sharing, linking and interoperability of archaeological data and information, particularly information derived from the commercial development funded archaeology sector in order to maximise public value and enhance the research potential of archaeological data? 2. How can we ensure the consistent development, application and enforcement of existing technical information and data standards and their promotion to others?   The Matrix project is addressing the current problems caused by the lack of standardized approaches to digital archiving of archaeological data using the particular case study of stratigraphic and phasing data. Stratigraphic data form the backbone of all the related archaeological records from each excavated site and are essential for integrated analysis, wider synthesis and accessible archiving of the growing body of archaeological data and reports generated through the commercial archaeological sector in the UK and internationally. The stratigraphic record, usually in the form of a stratigraphic matrix, with associated relationships and data, acts as a primary, if not the primary piece of 'Evidence' for how, and in what order, the site was excavated. As such the stratigraphic matrix is the key mechanism that enables anyone less familiar with the site, to re-visit the excavation records, understand what data is most relevant for any particular research questions, or problems encountered, and piece together the underlying details of how the interpretations by the excavator(s) were actually arrived at. However, such records are often only held on paper or scanned copies of matrix diagrams that cannot easily be re-used with associated data. Often the key phasing data needed for synthesis work and interpretive understanding is not well documented or archived consistently, if at all, in written reports. This results in key records being unsearchable or remaining unconnected to other data and at best usually requires lengthy and wasteful re-keying if any one wishes to work with the archives from such sites. The focus of digital archives and museums is now switching from simply providing better access to digital archives, to questions of how users in commercial units, curatorial organizations and academia, along with the general public, can make best use of this growing body of digital information and data (May 2020).   A number of research projects in the last 10 years have attempted to draw together archaeological archives from different excavation teams to analyse the temporal sequences and use the stratigraphic relationships recorded on site to cross-search for artefacts and structures from related phases e.g. the STAR project (Tudhope et al. 2011). But a lack of consistent practice in digital deposition of such records has placed severe limitations on the amounts of archaeological records available for such analyses. For Chronological modelers of archaeological data these problems are exacerbated by a lack of standardized approaches to the archiving of stratigraphic data, often held in hard-copy matrix diagrams or inconsistently structured database tables. The outcomes of the Matrix project will help inform decisions on digital archiving standards and best practice for stratigraphic data deposition and re-use, so that such digital data can be held and re-used in a form most suitable as input for Bayesian calibration software such as BCal, OxCal, or Chronomodel. The use of such Bayesian chronological modelling techniques has become critical in the more accurate dating of archaeological sites and phases in the last 10 years, but the way such information is analysed is quite painstaking and often involves many hours of laborious manual data preparation for key staff involved (Dye and Buck 2015). The techniques this project develops will considerably reduce the inefficiencies in this process. Stratigraphic laws and principles underpin the archaeological records from excavated sites and are essential for integrated analysis, wider synthesis and accessible digital archiving of the growing body of archaeological data and reports generated through the commercial archaeological sector in the UK and internationally. On most excavated sites, the stratigraphic record, most often visualized, and to a degree quantifiable, in the form of a stratigraphic matrix, acts as a primary, if not the primary piece of evidence for how, and in what order, the site was excavated. As such the stratigraphic record is the key mechanism that enables anyone less familiar with the site, to re-visit the excavation records, understand what data is most relevant for any research questions, or problems encountered, and piece together the underlying details of how the interpretations by the excavator(s) were arrived at.   However, such primary records are often only held on paper or scanned copies of matrix diagrams that cannot easily be re-used with associated data. Often the key phasing data needed for re-use in synthesis work and interpretive understanding, let alone Bayesian Chronological modelling, is not consistently documented, if at all, in written reports. This results in key records being unsearchable or remaining unconnected and lacking interoperability with other data (unFAIR).   The focus of digital archives and museums is now switching from simply providing better access to digital archives, to questions of how users in commercial units, curatorial organizations and academia, along with the general public, are going to make best use of this growing body of digital information and data.   AHRC Leadership Fellow – The Matrix https://www.researchgate.net/project/The-Matrix-connecting-and-re-using-digital-records-and-archives-of-archaeological-investigations Analytics Department Policy & Evidence Historic England Keith.May@HistoricEngland.org.uk May, K. 2020 The Matrix: Connecting Time and Space in Archaeological Stratigraphic Records and Archives, Internet Archaeology 55. https://doi.org/10.11141/ia.55.8 The Matrix project web page: https://stratigraphic.github.io/matrix/
• #3: Introduction to Matrix project Stratigraphic Standards – Need for interchange and archive formats for stratigraphic data derived from Analysis and phasing processes Process Modelling for Analysis – review of existing processes for digital stratigraphic recording and analysis Phaser – prototype web based software tool developed to explore the user needs for stratigraphic analysis and suitability for converting and migrating legacy stratigraphic matrix data via JSON export LOD vocabularies – e.g. Linked Open Data (LOD) terminologies from PeriodO for temporal data to improve the interoperability of resulting digital archives. Phasing methodologies - use of the prototype Matrix tool for dating records and spatio-temporal matrix phasing Project outcomes aim at improving the re-usability of stratigraphic data for Bayesian chronological modelling and increasing the FAIR-ness especially for interoperability and reusable synthesis of archaeological data.
• #4: Issues with Digital Data in Analysis & Stratigraphic Standards in Digital Archive contents Stratigraphy is “the skeleton” of the archaeological site How do people document Stratigraphic Analysis? How is an Harris Matrix archived? – just PDF image? Kept as images or more reusable data? How readily able to re-use strat relationships? Need a consistent format for preservation, interchange sharing and re-use of the Stratigraphic RELATIONSHIPS E.g. Data as CSV can easily convert to RDF/XML/JSON for use by semantic technologies
• #5: Issues with stratigraphic data not appearing in reports or archives of archaeological investigations in the UK.
• #6: It is commonly understood that the archaeological excavation removes, and effectively destroys, the stratigraphy that it is at the same time attempting to record. As Phillip Barker put it the excavation process is an “unrepeatable experiment”. This can become an excuse for lack of quality But is this necessarily true of the data we manage and analyse after excavation, and the additional records and interpretive analysis that we make after the site is excavated? – No! Surely a more consistent model for Analysis practice is possible?
• #7: Could this form of Phasing Index provide the basis for a more standardized approach to archiving stratigraphic grouping and phasing data? Do other organizations in the UK, or internationally, have equivalent forms of record, or only Museum of London Archaeology (MOLA)? Such data is crucial for anyone trying to re-visit and re-use the data in digital archives.
• #8: In 2015 I got copied into an email exchange between Caitlin Buck & Julian Richards at the Archaeology Data Service (ADS) at York University https://archaeologydataservice.ac.uk/ asking about how to get hold of examples of digitally archived stratigraphic data. One big issue for interoperability that had emerged from STAR was that the lack of a common approach to how people archive their data results from the Analysis. This results in digital outputs on the analysis process being very fragmented.   Tales From Two River Banks Commercial Vs Research Funded archaeological project data.   I decided I needed to look for commonalties of process and asked the question Would Process modelling for Analysis help? – The aim was three-fold: Identify common steps in the Analysis process Identify and clarify differing approaches to process Enable semantic mappings between common concepts and terms used in Analysis process(es)
• #9: Under the current Matrix AHRC funded research project I have Interviewed staff from all of the 10 most active archaeological Units in England and produced a draft Process model for Analysis – Principally Grouping & Phasing which has helped with initial R&D work on prototype software. Reduced the most common approaches to PX Phasing to 5 key iterative steps
• #10: There also seem to be emerging differences and divergence in process, and therefore differing digital archive products, as GIS based technologies and more recently Structure from Motion (SFM) have become more prevalent in last 10-20 years, especially in large-scale infrastructure projects undertaken in non-urban, or less deeply stratified, sites.
• #11: 11. The terms “period” and “phase” are defined variously and sometimes interchangeably by archaeologists. In this case a Period can be nested temporally within a Period In the Harris Matric Composer help file “Units can be grouped into Phases and Periods. Phases represent structural relations and periods temporal relations.”
• #12: Silbury Matrix - What became clear as I worked on more of this was that the other Allen Operators were represented in the written reports and implicitly within the hardcopy versions of Harris Matrix diagrams that had gone through full PX Analysis processes, but unfortunately there was no formal way that such temporal reasoning and relationships between the data were included in the final project publications, and particularly not in the digital data sets in any way that would enable use by the semantic reason tools. Potential benefits if the stratigraphic records, including groups/phases were more consistently recorded in archive data. • Potential of more explicit Phasing data methodology approaches • Semantic opportunities of using Allen operators[KM1] to cross-search related data • Semantic search and e.g. for dated material - Bayesian potential enhanced by additional dating parameters from additional temporal relationships. [KM1]Slide of Silbury Hill matrix with Allen operators would work here
• #20: Are “The old ways” always the best?
• #21: Conclusions & Challenges Reduce - Need more consistent standards in digital records of stratigraphic and temporal relationships (amongst others) Reuse - Matrix data should be re-usable effectively e.g. minimum as CSV files, rather than as images of matrix diagrams buried in a PDF document. Recycle - Is there a need for an International Convention on Archaeological Stratigraphic & Chronological Methods and Data? To facilitate better understanding and communication. FAIR stratigraphic data standards - Make this fundamental archaeological data more sustainably Findable, Accessible, Interoperable and Reusable (FAIR Principles 2016) across present day geo-political (period), and spatio-temporal, boundaries?
• #22: Allen, J.F. (1983) 'Maintaining Knowledge about Temporal Intervals'. Communications of the ACM 26, 11, 832-843. Binding, C. (2010) Implementing archaeological time periods using CIDOC CRM and SKOS .The Semantic Web: Research and Applications : 7th Extended Semantic Web Conference, ESWC 2010. Cripps, P., Greenhalgh, A. Fellows, D., May, K., David Robinson, D. (2004) Ontological Modelling of the work of the Centre for Archaeology, CIDOC CRM technical paper: pdf file. Also available: The CRM Diagram, pdf file. Dye, T.S. & Buck, C.E. (2015) Archaeological sequence diagrams and Bayesian chronological models.Journal of Archaeological Science, 63. 84 - 93. ISSN 0305-4403 Harris, E.C. (1979) Principles of Archaeological Stratigraphy (1st Edition). London: Academic Press. Huggett, J. Reuse remix recycle: repurposing archaeological digital data. Advances in Archaeological Practice (2018), doi:10.1017/aap.2018.1 May, K. 2020 The Matrix: Connecting Time and Space in Archaeological Stratigraphic Records and Archives, Internet Archaeology 55. https://doi.org/10.11141/ia.55.8 Papadakis, M., Doerr, M. and Plexousakis, D. "Fuzzy times on space-time volumes," eChallenges e-2014 Conference Proceedings, Belfast, 2014, pp. 1-11. Roskams, S. (2001) Excavation. Cambridge: Cambridge University Press. Taylor, J.S. (2016) Making Time For Space At Çatalhöyük: GIS as a tool for exploring intra-site spatiotemporality within complex stratigraphic sequences. PhD thesis, University of York. Tudhope, D., May, K., Binding, C. and Vlachidis, A. (2011) 'Connecting Archaeological Data and Grey Literature via Semantic Cross Search', Internet Archaeology 30. https://doi.org/10.11141/ia.30.5 HeritageData.org LOD vocabularies: http://www.heritagedata.org/blog/ Perio.do LOD gazetteer of Periods: https://perio.do/en/ SKOS standard for RDF vocabularies: https://www.w3.org/2004/02/skos/ FAIR Principles https://www.force11.org/group/fairgroup/fairprinciples
• #23: Paul Cripps, Anne Greenhalgh, Dave Fellows, Keith May, David RobinsonOntological Modelling of the work of the Centre for Archaeology, September 2004Available: pdf file (207 Kb)Also available: The CRM Diagram, pdf file (65 Kb)