SCoCESLE Corpus — Metadata

Metadata for the SCoCESLE learner corpus: a collection of learner essays annotated with learner background and proficiency.

Overview

This repository holds the metadata spreadsheet describing 307 learner essays, including essay prompts, learner L1, CEFR proficiency level, and demographics. It accompanies work on a "problem of computational linguistics" act (акт проблемы комплинг).

Repository contents

File	Description
SCoCESLE-metadata-spreadsheet.csv	307 rows × 11 columns of essay metadata.
SCoCESLE_metadata_spreadsheet.xls	The same spreadsheet in legacy Excel format.

Schema

Column	Meaning
Essay ID	Unique essay identifier (e.g. 01b.txt)
Total number of word types	Type count
Total number of word tokens	Token count
Essay type/topic	Essay sub-type/prompt letter
Essay question	Full prompt text
Course	Course the essay was written for (e.g. IELTS Preparation)
L1	Learner's first language (e.g. Colombian Spanish)
CEFR level as per course being studied	Proficiency (e.g. B2)
Gender	Learner gender
Year of birth	Learner birth year

Quick start

import pandas as pd
df = pd.read_csv("SCoCESLE-metadata-spreadsheet.csv")
print(df[["Essay ID", "L1", "CEFR level as per course being studied"]].head())

License

Research use. The corpus metadata is provided for academic purposes; please contact the author before redistribution.