--- annotations_creators: - expert-generated language_creators: - machine-generated languages: - en licenses: - unknown multilinguality: - monolingual pretty_name: BOUN size_categories: - unknown source_datasets: - original task_categories: - structure-prediction task_ids: - structure-prediction-other-word-segmentation --- # Dataset Card for BOUN ## Dataset Description - **Repository:** [ardax/hashtag-segmentor](https://github.com/ardax/hashtag-segmentor) - **Paper:** [Segmenting hashtags using automatically created training data](http://www.lrec-conf.org/proceedings/lrec2016/pdf/708_Paper.pdf) ### Dataset Summary Automatically segmented 803K SNAP Twitter Data Set hashtags with the heuristic described in the paper "Segmenting hashtags using automatically created training data". ### Languages English ## Dataset Structure ### Data Instances ``` { "index": 0, "hashtag": "BrandThunder", "segmentation": "Brand Thunder" } ``` ### Data Fields - `index`: a numerical index. - `hashtag`: the original hashtag. - `segmentation`: the gold segmentation for the hashtag. ### Citation Information ``` @inproceedings{celebi2016segmenting, title={Segmenting hashtags using automatically created training data}, author={Celebi, Arda and {\"O}zg{\"u}r, Arzucan}, booktitle={Proceedings of the Tenth International Conference on Language Resources and Evaluation (LREC'16)}, pages={2981--2985}, year={2016} } ``` ### Contributions This dataset was added by [@ruanchaves](https://github.com/ruanchaves) while developing the [hashformers](https://github..com/ruanchaves/hashformers) library.