askubuntu

# Dataset Card for AskUbuntu The AskUbuntu dataset (Lei et al., 2016) is a collection of preprocessed questions taken from AskUbuntu.com 2014 corpus dump. It also comes with 400*20 mannual annotations, marking pairs of questions as "similar" or "non-similar". The dataset is sourced from the original [GitHub repository](https://github.com/taolei87/askubuntu). Note that for the `train` split, the "positive" is the list of similar questions according to AskUbuntu, and "negative" is a list of randomly selected questions. For `dev` and `test`, the "positive" are human-annotated and may be empty. Unlike the original dataset, the "positive" data is *not* a subset of the "negative" data, they're mutually exclusive. See also [sentence-transformers/askubuntu-questions](https://huggingface.co/datasets/sentence-transformers/askubuntu-questions) for all AskUbuntu questions in this dataset. ## Dataset Details * Columns: "query", "positive", "negative" * Column types: `str`, `list[str]`, `list[str]` * Examples: ```python { "query": "system running in low graphic mode ( ubuntu without monitor )", "positive": [ "getting system to boot in headless mode set-up without display problems" ], "negative": [ "software center not progress showing mergelist error", "how to install google earth or draftsight for 64-bit os ?", "how to install a huawei ec-226 usb modem ?", ... ] } ``` * Collection strategy: Downloading the `train_random.txt`, `dev.text`, and `test.txt` files from the original source, and mapping the query IDs to the query texts. * Deduplicated: No