Digital Humanities Vocabulary in English

Text mining (also called text analytics) is the automated process of extracting meaningful patterns, trends, and insights from large bodies of text using computational methods. In digital humanities, it is used to analyse corpora of thousands or millions of documents impossible to read manually. Techniques include topic modelling, sentiment analysis, word frequency analysis, and named entity recognition. Text mining enables ‘distant reading’ — the analysis of large-scale patterns across many texts simultaneously, complementing traditional close reading of individual texts.

Digitisation is the process of converting physical objects, documents, or media into digital formats that can be stored, searched, and shared electronically. In cultural heritage institutions, digitisation projects aim to improve access to collections, preserve fragile materials, and enable new research possibilities. Mass digitisation projects such as Google Books and the Internet Archive have made millions of texts freely searchable. Digitisation raises questions about ownership, copyright, access inequality, and the long-term preservation of digital files.

Metadata is data that describes other data — in digital archives, it is structured information about a digital object that makes it discoverable, understandable, and usable. Metadata for a digitised manuscript might include title, author, date, language, provenance, and rights statement. Standards such as Dublin Core, MARC, and TEI provide common frameworks that enable interoperability between different digital collections. Poor metadata makes digital objects effectively invisible to researchers, so quality and consistency are crucial.

A corpus (plural: corpora) in digital humanities is a structured collection of texts assembled for research purposes. Digital corpora enable computational analysis impossible with physical texts — searching for word patterns across thousands of documents, tracing the evolution of language over time, or comparing the style of different authors. Famous digital corpora include the British National Corpus and the HathiTrust Digital Library. Corpus design decisions — what texts to include and how to annotate them — fundamentally shape the research questions that can be asked.

Topic modelling is a machine learning technique used to discover abstract themes within a large collection of documents. The most widely used algorithm is Latent Dirichlet Allocation (LDA), which identifies groups of words that tend to appear together. Historians have used it to trace the emergence and decline of political themes in newspapers, while literary scholars have used it to study genre conventions. The results require careful human interpretation, as the computer identifies statistical patterns rather than understanding meaning.

Open access refers to the free, unrestricted online availability of research outputs — including articles, datasets, and digital editions — so that anyone can read, download, and reuse them without financial or legal barriers. The open access movement argues that publicly funded research should be freely available to the public. In digital humanities, open access is particularly important because DH projects often produce digital editions, datasets, and tools that need to be shared widely. Preprint servers, institutional repositories, and open-access journals are the main vehicles.

A digital edition is a scholarly edition of a historical text that uses digital technologies to offer features not possible in print, such as multiple versions displayed in parallel, interactive annotations, hyperlinks to related documents, image facsimiles alongside transcriptions, and encoded markup enabling computational searching. The Text Encoding Initiative (TEI) provides XML-based guidelines for marking up digital editions. Famous examples include the Walt Whitman Archive and the Samuel Beckett Digital Manuscript Project.

GIS (Geographic Information Systems) is a technology for capturing, storing, analysing, and visualising spatial data. In digital humanities it is used to map historical phenomena such as the spread of diseases, trade networks, migration patterns, and battle locations. Historical GIS projects layer digitised maps with data about past events to reveal spatial patterns that text-based analysis cannot easily show. The Mapping the Republic of Letters project used GIS to visualise correspondence networks of early modern scholars across Europe.

Group terms by domain: archival (digitisation, metadata, finding aid, provenance), computational (text mining, corpus, algorithm, visualisation), publication (open access, digital edition, TEI, repository), and theory (distant reading, digital preservation). Reading the Digital Humanities Quarterly (open access) and the Programming Historian (free tutorials) exposes you to DH vocabulary in both theoretical and practical contexts. Joining DH online communities and following the #DH hashtag connects you with researchers using this vocabulary in real discussions. Trying beginner-level text analysis tools such as Voyant Tools helps you internalise computational vocabulary through hands-on practice.