---
license: cc-by-4.0
language:
- sr
field: sents
pretty_name: Serbian Literary Text Collection
size_categories:
- 1M<n<10M
task_categories:
- text-generation
tags:
- literary
---

SrpELTeC is a corpus of old Serbian novels for the first time published in the period 1840-1920. years of digitized within COST ACTION CO16204: Distant Reading for European Literary History, 2018-2022.

The corpus includes 120 novels with 5,263.071 words, 22700 pages, 2557 chapters, 158,317 passages, 567 songs, 2972 verses, 803 segments in foreign language and 949 mentioned works.

Dataset is constituted of two text files that can be loaded via:

```python
from datasets import load_dataset
dataset = load_dataset("jerteh/SrpELTeC")
```

Preview:

```python
print(dataset["train"][100])
{'text': '„Mi besmo dobre gazde u selu, otac moj imao je kuću i sesiju zemlje, imali smo četiri vola, lepa četiri krstaša, rogovi u njih behu kao u jelena, ali još više u šire nego u jelena, a vrhovi behu bačeni natrag, dlaku na glavi imaše belu i kovrčavu, koja je skoro do očiju popadala bila na glavi, — pa posle sva ova četiri vola behu tako jednaka i u hodu i u veličini i debljini, da je to divota bilo pogledati — e ta divna ti je to stoka bila.'}
```

Citation:

```bibtex
@inproceedings{stankovic2022distant,
  title={Distant Reading in Digital Humanities: Case Study on the Serbian Part of the ELTeC Collection},
  author={Stankovi{\'c}, Ranka and Krstev, Cvetana and {\v{S}}andrih Todorovi{\'c}, Branislava and Vitas, Du{\v{s}}ko and {\v{S}}kori{\'c}, Mihailo and Ikoni{\'c} Ne{\v{s}}i{\'c}, Milica},
  booktitle={Proceedings of the Thirteenth Language Resources and Evaluation Conference},
  pages={3337--3345},
  year={2022}
}
```