• Digital Humanities and Cultural Heritage
• Research data close
• Dataset close
• European Commission close
• EC|H2020 close
• EU close
• Digital Humanities and Cultural Her... close

The Observatory Knowledge Graph (OKG) is a knowledge graph with tweets on inequality in terms of the OBIO ontology (https://w3id.org/okg/obio-ontology/), which integrates social media metadata with various types of linguistic knowledge. The OKG can be used as the backbone of a social media observatory, to facilitate a deeper understanding of social media discourse on inequality. We retrieved tweets and retweets published from the end (30th) of May 2020 to the beginning (1st) of May 2023. In this version of the OKG, we use a sample of 85,247 tweets, published from May 30th to August 27th, 2020. To be compliant with Twitter's policies, we remove usernames and id's, as well as the tweet texts and sentences. We also replace user IRIs with skolem IRIs through skolemization. Access to the OKG as well as the SPARQL endpoint can be requested by sending a mail to the contact person with the following information: A description of the use case Affiliation of the researchers involved How their work is in line with Twitter's policies: https://developer.twitter.com/en/developer-terms/policy#4-d

The corpus contains 668,529 tweets (tweet IDs) relevant to "impact investing", accompanied by sentiment labels given by an automated sentiment classifier. Impact investing involves investments made into companies, organizations, and funds with the intention to generate social and environmental impact alongside a financial return. The tweets relevant to impact investing were collected in the period from March 28, 2017, to January 28, 2018, through the Twitter Search API, and annotated for sentiment labels "Negative", "Neutral" or "Positive" by a general-purpose English language sentiment classifier. The tweets were collected based on a list of known impact investing Twitter users, relevant keywords and impact investing related events. In particular, the queries include relevant users (@YF_Academy, @esmeefairbairn, @resonanceltd, @Big PotentialSI, etc.), single hashtags (#socfin, #impinv #socialfinance, #impactinvestment, etc.), combined hashtags (#social & #finance, #social & #investment, #impact & #assessment, etc.), and hashtags of major impact investing events (#impact2, #socap17, #OxfordIIP, #skollwf, etc.).

The resource consists of two datasets related to Members of the 8th European Parliament (MEPs). The first one is a dataset of 2,535 roll-call votes of MEPs until 2016-03-01. The second one is a dataset of 26,133 retweets between MEPs in the period between 2014-10-01 and 2016-03-01. The data can be used to examine the patterns of covoting and retweeting of MEPs and analyze the extent to which they are similar. The resource is presented and used in the paper: Darko Cherepnalkoski, Andreas Karpf, Igor Mozetič, Miha Grčar "Cohesion and coalition formation in the European Parliament: Roll-call votes and Twitter activities". PLoS ONE 11(11): e0166586, 2016. http://dx.doi.org/10.1371/journal.pone.0166586 The dataset contains 5 files, of which 3 contain metadata and 2 data. The metadata comprises information about the Members of 8th European Parliament (MEPs) until 2016-03-01, about roll-call votes (RCV) and possible actions during a RCV. The first data file contains a matrix with the votes of all MEPs during all RCVs while the second contains the retweets between the MEPs.

The corpus contains over 4.5 million tweets (tweet IDs) automatically labeled by a machine learning program with stance regarding Brexit: Positive (supporting Brexit), Negative (opposing Brexit), or Neutral (uncommitted). The Brexit referendum was held on June 23, 2016, to decide whether the UK should leave or remain in the EU. In the weeks before the referendum, starting on May 12, the UK geo-located Brexit-related tweets were continuously collected resulting in a dataset of around 4.5 million (4,508,440) tweets from almost one million (998,054) users. A large sample of the collected tweets (35,000) was manually labeled for the stance of their authors regarding Brexit: Positive (supporting Brexit), Negative (opposing Brexit), or Neutral (uncommitted). The labeled tweets were used to train a classifier which then automatically labeled all the remaining tweets. The corpus contains tweet ids and stance labels. The tweets are grouped into files one hour per file. In each file, one row represents one entry (twitter_id, sentiment_label). Lines are ordered by the tweet time. The data collection, annotation, model training and performance estimation is described in detail in: Miha Grčar, Darko Cherepnalkoski, Igor Mozetič, Petra Kralj Novak: Stance and influence of Twitter users regarding the Brexit referendum. Computational Social Networks 4/6. 2017. http://dx.doi.org/10.1186/s40649-017-0042-6