Find research datasets worth reusing
Search datasets from major research repositories and use ShareScore to quickly assess how well each record supports discovery, access, and reuse.
8
datasets available to search
ShareScore release 0.9.0
Dataset results
8 results for “Social Text”
MHG4SNA: Middle high german texts annotated for social network analysis
<p><strong>Description</strong></p> <p>This corpus contains multiple middle high german texts with annotations for social network analysis. It contains annotations of: named entities and entity mentions (including partial coreference resolution), direct speech, narrator's comments. See below for further description.</p> <p>The annotated texts are part of my dissertation on social network analysis of arthurian romances. The research was developed in the context of the DH center <a href="https://www.creta.uni-stuttgart.de/">CRETA</a> at the University of Stuttgart.</p> <p> </p> <p><strong>Texts</strong></p> <ul> <li>Wolfram von Eschenbach: 'Parzival', in: Wolfram von Eschenbach: Werke, ed. by Karl Lachmann, 5th edition, Berlin 1891, pp. 11–388.</li> <li>Hartmann von Aue: 'Erec', ed. by Albert Leitzmann continued by Ludwig Wolff, 7th edition by Kurt Gärtner, Tübingen 2006 (Altdeutsche Textbibliothek 39).</li> <li>Hartmann von Aue: 'Iwein', ed. by G. F. Benecke and K. Lachmann, revised by Ludwig Wolff, 7th edition, part 1: Text, Berlin 1968.</li> <li>Wolfram von Eschenbach: 'Willehalm', in: Wolfram von Eschenbach: Werke, ed. by Karl Lachmann, 5th edition, Berlin 1891, pp. 421–640.</li> <li>'Das Rolandslied des Pfaffen Konrad', ed. by Carl Wesle, 3rd edition by Peter Wapnewski, Tübingen 1985 (Altdeutsche Textbibliothek 69).</li> </ul> <p>All texts are part of the MHDBDB (<a href="http://mhdbdb.sbg.ac.at/">Mittelhochdeutsche Begriffsdatenbank</a>).</p> <p> </p> <p><strong>Annotations</strong></p> <p>The texts contain annotations of different categories, as described in the following sections.</p> <p> 1.<em> Named Entities and Entity Mentions</em></p> <p>I annotated all namend entities and entity mentions that belong to the categories PER and LOC. PER stands for 'person' and refers to real persons as well as fictional characters. LOC stands for 'location' and includes real and fictional places.</p> <p>I annotated named entities (e.g. 'Parzival' as PER or 'Nantes' as LOC) as well as entity mentions referring to an instance of PER or LOC (e.g. 'the knight' for Parzival, or 'the city' for Nantes). I did not annotate pronouns. Entity references can contain multiple words, e.g. 'the lovely queen Ginover', and they can be nested, e.g. '[the son of [the king Gahmuret]]'.</p> <p>The annotations follow the guidelines created for multiple categories and disciplines in the context of CRETA. They are published <a href="https://www.creta.uni-stuttgart.de/cute/datenmaterial/annotationsrichtlinien-1-1/index.html">here</a>. </p> <p> 2.<em> Entity Grounding</em></p> <p>All annotated entity references are mapped to the entity instance that they refer to. E.g. the refences 'Parzival', 'Herzeloyde's son', 'the young man', 'the red knight' etc. all refer to the character instance 'Parzival'. The entity grounding takes into consideration the context of the entity mentions since one and the same expression can refer to different instances (in one context 'the king' refers to Arthur, in another context to Gahmuret)</p> <p> 3.<em> Direct Speech (DS)</em></p> <p>Passages of direct speech have been annotated by detecting quotation marks. They are tagged as 'DS'. There are a few cases of embedded direct speech (passages of direct speech containing another passage of direct speech); these cases are annotated as well.</p> <p> 4. <em>Narrator's comments (EK)</em></p> <p>As additional category I annotated passages that contain statements of the narrator, narrator's comments, extensive descriptions or digressions (e.g. an excursus to a specific topic). These passages are not part of the fictional world or lead to a pause in the timeline of events. The are annotated as 'EK' ('EK': passages that aren't part of the diegesis, 'EK2': passages that lead to a pause, e.g. comments or descriptions).</p> <p> 5. <em>Segmentation</em></p> <p>The texts are subdivided in passages of 30 verses. Since some text's editions ('Parzival', 'Willehalm') contain a formal segmentation in passages of 30 verses each, the same kind of segmentation has been transfered to the other texts. This means 'segment 1' contains the first 30 verses, 'segment 2' contains verses 31-60 and so on.</p> <p>According to the editions by Lachmann, 'Parzival' and 'Willehalm' are also subdivided in chapter-like books (Parzival: book 1 to 16, Willehalm: book 1 to 9). The other texts are similarly subdivided in chapter-like sections following common content-based divisions.</p> <p> </p> <p><strong>Social Network Analysis</strong></p> <p>The data can be used to explore and analyse the social network of the texts. SNA can be performed via gephi [4] using the gefx files.</p> <p>The social network is based on co-occurrences using a) the annotated and grounded entities, and b) the text segmentation in segments of 30 verses each. A relation between two or more entities is extracted whenever they co-occur in a segment.</p> <p> </p> <p><strong>Data downloads</strong></p> <p>The annotated texts can be downloaded in multiple formats: conll, csv, and gexf.</p> <p> 1. <em>Conll </em></p> <p>The files contain seven columns:</p> <ul> <li>(1) token,</li> <li>(2) POS-tag, tagged using a <a href="https://www.ims.uni-stuttgart.de/forschung/ressourcen/werkzeuge/pos-tag-mhg/">middle high german pos tagger</a>,</li> <li>(3) number of segment,</li> <li>(4) Entity reference annotation indicating the intance that the entity reference refers to. '-' if there is no entity reference,</li> <li>(5) EK: '1' in case there is an annotation of 'EK', '0' if not,</li> <li>(6) EK2: '1' in case there is an annotation of 'EK2', '0' if not,</li> <li>(7) DS: '1' if the token is tagged as direct speech, '0' if not.</li> </ul> <p> 2. <em>Csv</em></p> <p>The csv files contain all annotations of the category PER including entity grounding.</p> <p>The files contain the following columns:</p> <ul> <li>begin and end (start and end of the entity reference expression, character offset),</li> <li>doc_id (document id),</li> <li>buch (book number),</li> <li>quote (entity reference expression),</li> <li>coref (the entity instance that the expression refers to),</li> <li>overlap (indicates if there is an overlap, relevant for embedded entities),</li> <li>ek and ek2 (narrator's comment),</li> <li>ds (direct speech),</li> <li>space (annotations of the space where the story takes action, can be ignored here),</li> <li>segnr (number of segment),</li> <li>em (embedded),</li> <li>klasse (entity class),</li> <li>xrange (technical, relevant for annotation view).</li> </ul> <p> 3. <em>Gexf</em></p> <p>These files can be used to import the data to gephi. It is based on the annotation and grounding of entities (categorie PER). A relation between entities is based on co-occurrence (whenever two or more entities co-occur in a segment, they have a relation; with more relations, the intensitiy of their relation grows). The text segmentation is described above.</p> <p>Embedded entities are excluded. Entities mentioned in direct speech (DS) or in comments (EK) can optionally be selected or deselected. These optional filters are indicated in the name of the files.</p> <p>To visualize the graph dynamically, one can use the text segmentation as timeline.</p> <p> </p> <p>release v1.0.0: data publication in the context of my dissertation. </p>
Text Message Treatment for Social Anxiety
ClinicalTrials.gov study NCT03562650. IPD Sharing: UNDECIDED. Countries: 1. Publications: 1.
Stay Well at Home: a Text-messaging Study Social Distancing
ClinicalTrials.gov study NCT04473599. IPD Sharing: YES. Countries: 1. Publications: 3.
The texts related to Postgraduate stress from mainstream social media in China.
Open the record for dataset details and reuse information.
Text Message Safety Behavior Fading for Social Anxiety
ClinicalTrials.gov study NCT06168253. IPD Sharing: NO. Countries: 1. Publications: 0.
PyTAIL Benchmark of Active Learning on Social Media Text Classification
<p>PyTAIL Benchmark of Active Learning on Social Media Text Classification</p><p>Read our paper for details: https://arxiv.org/abs/2211.13786</p><ul><li>ArXiv: https://arxiv.org/abs/2211.13786</li><li>Dataset: https://doi.org/10.5281/zenodo.7236430</li><li>Code: https://github.com/socialmediaie/pytail</li><li>Video: https://www.youtube.com/watch?v=AwDu64gN8t4 </li></ul>
SOCIAL NETWORK TEXT CLASSIFICATION ALGORITHM AND SOFTWARE TOOL
<p><em><span>In this article, today's information in social networks, global information flows and algorithms for their analysis, research in this regard and their effects are studied. Recent effective algorithms are presented and the procedure for their use is shown.</span></em></p>
Datasets from `Discovering and analysing lexical variation in social media text'
<p>This repository contains the datasets that were used in the following three papers, which are also included within P. Shoemark's PhD dissertation `Discovering and analysing lexical variation in social media text':</p> <ul> <li>P. Shoemark, D. Sur, L. Shrimpton, I. Murray, and S. Goldwater. <a href="https://www.aclweb.org/anthology/E17-1116/"><em>Aye or naw, whit dae ye hink? Scottish independence and linguistic identity on social media.</em></a> 15th Conference of the European Chapter of the Association for Computational Linguistics (EACL). 2017. </li> <li>P. Shoemark, J. Kirby, and S. Goldwater. <a href="https://www.aclweb.org/anthology/W17-4908/"><em>Topic and audience effects on distinctively Scottish vocabulary usage in Twitter data.</em> </a>Workshop on Stylistic Variation at EMNLP. 2017.</li> <li>P. Shoemark, J. Kirby, and S. Goldwater. <a href="https://www.aclweb.org/anthology/W18-6101/"><em>Inducing a lexicon of sociolinguistic variables from code-mixed text.</em> </a>Workshop on Noisy User Generated Text at EMNLP. 2018. </li> </ul> <p>Datasets consist of tab-separated-values files, in which rows correspond to tweets, with columns for user ID, tweet ID, and timestamp. </p> <p>The text of the tweets (and associated metadata) can be re-downloaded (in batches of 100 per request) using Twitter's <a href="https://developer.twitter.com/en/docs/tweets/post-and-engage/api-reference/get-statuses-lookup">GET Statuses/Lookup</a> API endpoint <em>(NB: Tweets which have been deleted or made private since the original datasets were collected can <strong>not </strong>be re-downloaded, so it may not be possible to reconstruct the original datasets in their entirety). </em></p> <p> </p> <p>Most of these datasets were originally drawn from <a href="https://developer.twitter.com/en/docs/tweets/sample-realtime/api-reference/get-statuses-sample">the Sample endpoint of Twitter’s Streaming API (</a>a.k.a. the ‘Spritzer’), which provides a random 1% sample of all public tweets in near real-time:</p> <p> </p> <p><strong>GU Dataset: <a href="https://zenodo.org/api/files/caf09b6a-e57a-4a01-b2f0-ce5ab15583d1/Geotagged-UK.zip?versionId=a9cc222e-1d07-4026-9b55-cf2189b58191">Geotagged-UK.zip</a></strong></p> <p><em>Tweets from Sept 2013 - Sept 2014 which are geotagged to locations within the UK.</em></p> <p>The file <strong>GU_pre-filtering.tsv </strong>contains the IDs for all tweets from the ‘Spritzer’ stream which were posted between September 1st 2013 and September 30th 2014, were classified as English by <a href="https://github.com/saffsd/langid.py">langid.py</a>, are not retweets or quotes, and are geotagged to locations within the UK.<strong> </strong><em><strong>• Tweets: </strong>1,768,334<strong> • Unique Users: </strong>455,075 <strong>• </strong></em></p> <p>The file <strong>GU.tsv </strong>contains the IDs for tweets in the <strong>final</strong> GU dataset that was used for the analyses in our <a href="http://www.aclweb.org/anthology/E17-1116/">EACL 2017</a> paper, after applying additional pre-processing heuristics to filter out tweets by bots and spammers. <em><strong>• Tweets: </strong>1,654,204<strong> • Unique Users: </strong>446,510 <strong>• </strong><sub>(the number of users in the GU dataset was slightly over-counted when reported in the paper; this is the actual number)</sub></em></p> <p> </p> <p><strong>GS Dataset: <a href="https://zenodo.org/api/files/caf09b6a-e57a-4a01-b2f0-ce5ab15583d1/Geotagged-Scotland.zip?versionId=146a0589-ddc0-4c42-86f6-7da79686ae56">Geotagged-Scotland.zip </a></strong></p> <p><em>The subset of Tweets in the GU dataset which are geo-tagged to locations within Scotland, specifically.</em></p> <p>The file <strong>GS_pre-filtering.tsv </strong>contains the IDs for all tweets from the ‘Spritzer’ stream which were posted between September 1st 2013 and September 30th 2014, were classified as English by <a href="https://github.com/saffsd/langid.py">langid.py</a>, are not retweets or quotes, and are geotagged to locations within Scotland. <em><strong>• Tweets: </strong>178,401<strong> • Unique Users: </strong>41,685 <strong>• </strong></em></p> <p>The file <strong>GS.tsv </strong>contains the IDs for tweets in the <strong>final</strong> GS dataset that was used for the analyses in our <a href="http://www.aclweb.org/anthology/E17-1116/">EACL 2017</a> paper, after applying additional pre-processing heuristics to filter out tweets by bots and spammers. <em><strong>• Tweets: </strong>166,992<strong> • Unique Users: </strong>40,837 <strong>• </strong><sub>(the number of users in the GS dataset was slightly over-counted when reported in the paper; this is the actual number)</sub></em></p> <p> </p> <p><strong>IT Dataset & Controls: <a href="https://zenodo.org/api/files/caf09b6a-e57a-4a01-b2f0-ce5ab15583d1/Indyref-Tweets.zip?versionId=41042a3e-f247-4493-9abe-29fb9ae81ee1">Indyref-Tweets.zip</a></strong></p> <p><em>Tweets from Sept 2013 - Sept 2014 which contain hashtags relating to the 2014 Scottish Independence Referendum (plus 'control' tweets which are by the same users but do not contain referendum-related hashtags)</em></p> <p>The file <strong>IT_</strong><strong>pre-filtering.tsv </strong>contains the IDs for all tweets from the ‘Spritzer’ stream which were posted between September 1st 2013 and September 30th 2014, were classified as English by <a href="https://github.com/saffsd/langid.py">langid.py</a>, are not retweets or quotes, and contain at least one of 47 hashtags we identified as relating to the 2014 Scottish Independence Referendum (see paper for hashtag list). <em><strong>• Tweets: </strong>77,708<strong> • Unique Users: </strong>26,019 <strong>• </strong></em></p> <p>The file <strong>IT</strong><strong>.tsv </strong>contains the IDs for tweets in the <strong>final</strong> IT dataset that was used for the analyses in our <a href="http://www.aclweb.org/anthology/E17-1116/">EACL 2017</a> paper, after applying additional pre-processing heuristics to filter out tweets by bots and spammers, and tweets which do not contain hashtags that we judged to <em>unambiguously</em> relate to the referendum. <em><strong>• Tweets: </strong>59,664 <strong>• Unique Users: </strong>18,589 <strong>• </strong></em></p> <p>The file <strong>IT_controls_pre-filtering.tsv </strong>contains the IDs for all tweets from the ‘Spritzer’ stream which were posted between September 1st 2013 and September 30th 2014, were classified as English by <a href="https://github.com/saffsd/langid.py">langid.py</a>, are not retweets or quotes, and do <em><strong>not</strong></em> contain any of the hashtags we identified as relating to the 2014 Scottish Independence Referendum, but are authored by a user who has <em><strong>also</strong> </em>authored a tweet in <strong>IT_</strong><strong>pre-filtering.tsv</strong>. <em><strong>• Tweets: </strong>1,354,701 <strong>• Unique Users: </strong>26,019 <strong>• </strong></em></p> <p>The file <strong>IT_controls</strong><strong>.tsv </strong>contains the IDs for tweets in the <strong>final</strong> set of Control tweets that was used for the analyses in our <a href="http://www.aclweb.org/anthology/E17-1116/">EACL 2017</a> paper, i.e. tweets which do not contain referendum-related hashtags but are authored by users who also have also authored tweets in <strong>IT</strong><strong>.tsv</strong>. <em><strong>• Tweets: </strong>881,679 <strong>• Unique Users: </strong>18,589 <strong>• </strong></em></p> <p> </p> <p><strong>SG-Users’ and IH-Users' Autumn 2014 Timeline Datasets: <a href="https://zenodo.org/api/files/caf09b6a-e57a-4a01-b2f0-ce5ab15583d1/Autumn-2014_Timelines.zip?versionId=c456b530-361e-4b88-9006-5bebd4a43c92">Autumn-2014_Timelines.zip</a></strong></p> <p><em>Complete tweet histories from Aug-Oct 2014 for users from the GS and IT datasets.</em></p> <p>The file <strong>SG-Users_Autumn_2014_timelines_pre-filtering.tsv </strong>contains the IDs for tweets which were posted in August, September, or October 2014 by users from the GS dataset, i.e. users we know to have used Scottish geotags. This dataset is not restricted to tweets which appear in the ‘Spritzer’ sample; instead it consists of complete User Timelines for the months concerned, retrieved using the <a href="https://developer.twitter.com/en/docs/tweets/timelines/api-reference/get-statuses-user_timeline">statuses/user timeline</a> endpoint of Twitter’s REST API in March 2017. Because there are limits on the number of tweets that can be retrieved using this endpoint, we were not able to retrieve complete Autumn 2014 tweet histories for <em>all</em> of the users in the GS dataset. <em><strong>• Tweets: </strong>3,014,029 </em> <em><strong>• Unique Users: </strong>18,274 <strong>• </strong></em></p> <p>The file <strong>SG-Users_Autumn_2014_timelines.tsv </strong>contains the IDs for tweets in the <strong>final</strong> SG-Users dataset that was used for the analyses in our <a href="http://www.aclweb.org/anthology/W17-4908/">StyleVar 2017</a> paper. This dataset consists only of tweets which contain at least one instance of one of 50 lexical variables that were the focus of the study, and has also undergone various other filtering steps; see the paper for full details. <em><strong>• Tweets: </strong>1,112,931</em> <em><strong>• Unique Users: </strong>10,103 <strong>• </strong></em></p> <p>The file <strong>IH-Users_Autumn_2014_timelines_pre-filtering.tsv </strong>contains the IDs for tweets which were posted in August, September, or October 2014 by users from the IT dataset, i.e. users we know to have used Indyref-related hashtags. This dataset was collected in the same manner as SG-Users_Autumn_2014_timelines_pre-filtering.tsv; however, due to an error in this process, <strong>the IDs of most of the tweets in this dataset were not recorded</strong>. For such tweets the tweet ID column instead contains a placeholder tweet ID of the form _<user_ID>_<month>_<integer>, where the integer denotes the tweet's position in the reverse-chronological list of tweets that were retrieved for that user from that month (e.g. _147527441_09_286 is the placeholder tweet ID we assigned to the 286th September tweet we retrieved from the user whose Account ID is 147527441). Unfortunately, therefore, the tweets in this file whose 'IDs' begin with an underscore cannot be straightforwardly re-downloaded using Twitter's free <a href="https://developer.twitter.com/en/docs/tweets/post-and-engage/api-reference/get-statuses-lookup">GET Statuses/Lookup</a> API endpoint; but since their user IDs and timestamps are intact, it would still be possible to retrieve them using the (paid-for) <a href="http:// https://developer.twitter.com/en/docs/tutorials/choosing-historical-api">Historical APIs</a>. <em><strong>• Tweets: </strong>6,997,858 <strong>• Tweets whose IDs were recorded: </strong>288,394</em><em><strong> </strong> <strong>• Unique Users: </strong>14,645</em><em> <strong>• </strong></em></p> <p>The file <strong>IH-Users_Autumn_2014_timelines.tsv </strong>contains the IDs for tweets in the <strong>final</strong> IH-Users dataset that was used for the analyses in our <a href="http://www.aclweb.org/anthology/W17-4908/">StyleVar 2017</a> paper. Like the SG-Users dataset, this dataset consists only of tweets which contain at least one instance of one of 50 lexical variables that were the focus of the study, and has also undergone various other filtering steps; see the paper for full details. As with the pre-filtered version, most of the tweet IDs are unfortunately missing in this dataset. <em><strong>• Tweets: </strong>2,165,320 <strong>• Tweets whose IDs were recorded: </strong>115,366</em><em><strong> </strong> <strong>• Unique Users: </strong>10,784 <strong>• </strong></em></p> <p> </p> <p> </p> <p><strong>US Geotags: <a href="https://zenodo.org/api/files/caf09b6a-e57a-4a01-b2f0-ce5ab15583d1/Geotagged-USA.zip">Geotagged-USA.zip</a></strong></p> <p><em>Tweets from June 2013 - July 2016 which are geotagged to locations within the USA.</em></p> <p>The file <strong>GUSA.tsv</strong> contains all tweets from the ‘Spritzer’ sample which were posted between June 30th 2013 to July 1st 2016, are classified as English by <a href="https://github.com/saffsd/langid.py">langid.py</a>, are not retweets, do not contain urls or embedded media, are not by users with more than 1000 friends or followers, and are geotagged to locations within the USA. This dataset (along with the GU Dataset) was used in our <a href="http://www.aclweb.org/anthology/W18-6101/">WNUT 2018</a> paper. <em><strong>• Tweets: </strong></em> <em>8,375,573 </em> <em><strong>• Unique Users: </strong>1</em>,<em>826,260</em><em> <strong>• </strong></em></p> <p> </p>
ScienceDex guides
Understand access before you commit
These curated guides explain access requirements, typical timelines, costs, and reuse considerations for widely used research datasets.
Allen Brain Atlas
Allen Brain Atlas is an Allen Institute collection of brain map atlases, datasets, APIs, and analysis tools covering mouse, human, and non-human primate brain resources.
Annotated Behaviour and Observability Dataset (ABODe)
ABODe is a University of Edinburgh DataShare dataset for behavior classification in group-housed mice using home-cage video, identities, bounding boxes, ground-plate positions, and annotator labels.
DANDI Archive for NWB datasets
DANDI is a BRAIN Initiative archive for publishing and sharing neurophysiology data, including electrophysiology, optophysiology, and behavioral data packaged as NWB and related standards.
International Brain Laboratory public data
The International Brain Laboratory public data releases expose standardized mouse decision-making experiments, including Neuropixels recordings, widefield calcium imaging, behavior, and session metadata accessed through the ONE API.
OpenNeuro
OpenNeuro is a free, open platform for sharing neuroimaging datasets, with public search, dataset pages, and download paths for web, S3, DataLad, and the OpenNeuro CLI.