Verbs of seeing in Akkadian
This repository contains scripts and data for the paper "Language Technology Approach to Seeing in Akkadian".
/data
- aug18-nolex.txt
Lemmatized dataset from Oracc - results-pmi2-top50.log
Script parameters for pmizer (see https://github.com/asahala/Pmizer) - results-pmi2-top50.tsv
Results in .tsv format. Fields in the file:- keyword
- translation from Oracc
- collocate
- translation from Oracc
- period distribution
- genre distribution
- period and genre distribution
- keyword freq
- collocate freq
- co-occurrence freq
- PMI2 score
- average distance between keyword and collocate (in words)
- url to Korp (all links may not return results, as Korp Oracc had a major update in 2019: see https://www.kielipankki.fi/corpora/oracc/ for more info and user guide). Note that the co-occurrence of words (a, b) is symmetric, meaning that (a, b) == (b, a). Thus, if you search results in Korp using the links and do not get any results, you may have to switch the search boxes in reverse order. Unfortunately at present Korp does not allow "(a Σ+{w} b | b Σ+{w} a)" queries at present, thus all queries are in format "a Σ+{w} b" by default ("Σ+{w}" = any word times the window size, positions of a and b excluded).
- period/genre-distribution-matrix.tsv
Distribution of seeing verbs in different genres and periods as a matrix representation