LSJ-based Greek–English Dictionary for Lemmatized Ancient Greek Texts

This digital Greek–English dictionary is designed to support computational analysis of lemmatized Ancient Greek texts, particularly in connection with the LAGT (Lemmatized Ancient Greek Texts) dataset. It provides English definitions for lemmatized Greek words, facilitating semantic and lexical exploration of Ancient Greek corpora. The dictionary was created by matching a cleaned LAGT-derived vocabulary (n ≥ 2) with entries from the LSJ lexicon included in the open-source Diogenes software. Definitions were extracted from the XML-encoded “grc.lsj” file using a custom Python parser. In addition, 170 entries were added manually, mostly proper names. Unmatched items were excluded from the final version. The dictionary is distributed as a JSON file. Each entry contains the original lemma, a normalized form, and a list of English definitions: { "lemma": "φιλοσοφία", "lemma_normalized": "φιλοσοφια", "definitions": [ "love of knowledge, pursuit thereof, speculation", "philosophy", "systematic, methodical treatment" ] } The first release includes 64,774 unique raw lemmas and 46,693 normalized lemmas. A sample of 150 entries was manually reviewed to assess overall accuracy. However, due to the automated nature of the process, some errors may remain. Future versions will aim to increase the proportion of manually validated entries. The publication also includes a small Python script (lookup.py) that can be used from the command line to retrieve definitions by supplying a Greek lemma, and it supports both raw and normalized forms. This dictionary is optimized for offline use in Python workflows and can be easily integrated into a wide range of computational analyses of Ancient Greek texts. This work was supported by research grants from the Finnish Cultural Foundation and the Emil Aaltonen Foundation.