Bibliographic record
Abstract
This presentation examines the tools, approaches, collaboration, and findings of the Web Archives for Historical Research Group around the capture and analysis of Twitter for the 2015 Canadian Federal Election. \n \nWhile Twitter is not a representative sample of broader society - Pew Research notes that it skews young, college-educated, and affluent (above $50,000 household income) – Twitter still represents an exponential increase in the amount of information generated, retained, and preserved from non-elite people. Therefore, when historians study the 2015 federal election, Twitter will be a prime source. \n \nOn August 3, 2015, the team initiated both a search API and stream API collection with twarc using the hashtag #elxn42. Data collection ceased on November 5, 2015, the day after Justin Trudeau was sworn in as the 42nd Prime Minister of Canada. We collected for a total of 102 days, 13 hors and 50 minutes. \n \nTo analyze the data set, we took advantage of a number of utilities that are available within twarc and twarc-report, as well as jq, Mathematica, and Apache Spark Notebook. In accordance with the Twitter ToS, we also hosted the tweet ids in an institutional repository. \n \nOur analytics included: \n \n* breaking tweet text down by day to track change over time; \n* client analysis, allowing us to see how the scale of mobile devices affected medium interactions; \n* URL analysis, comparing both to Archive-It collections and the Wayback Availability API to add to our understanding of crawl completeness; \n * and image analysis, using an archive of extracted images. \n \nOur presentation introduces our collecting work, the analysis we have done, and provides a framework for other collecting institutions to do similar work with our off-the-shelf open-source tools. We hope that national libraries and other institutions will find our model useful as they consider how to archive ongoing events using Twtiter.
Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.
How this classification was reachedexpand
Full frame distilled prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. Learned from the 10,348 direct Codex labels and 10,348 direct Gemma labels. Candidate is the union of thresholded teacher heads; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels or direct frontier model labels.
Codex and Gemma teacher scores by category
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.000 | 0.000 |
| Meta-epidemiology (narrow) | 0.000 | 0.000 |
| Meta-epidemiology (broad) | 0.001 | 0.000 |
| Bibliometrics | 0.001 | 0.001 |
| Science and technology studies | 0.000 | 0.001 |
| Scholarly communication | 0.000 | 0.012 |
| Open science | 0.004 | 0.002 |
| Research integrity | 0.000 | 0.000 |
| Insufficient payload (model declined to judge) | 0.000 | 0.001 |
Machine scores (provisional)
The two teacher heads of the student model, read on this work. A score orders the frame for review; it never asserts a category, and the validation status ships verbatim with every row.
Baseline scores from an immature model (maturity gate not passed, 7 training rounds). Scores rank; they never assert a category.
score_only:v0-immature-baseline · verbatim from the scoring run: score_only means the number may rank works, and no category label ships from itClassification
machine, unvalidatedMachine predicted; a candidate call from one teacher head, not a consensus.
How this classification was reached, model by model and score by score, is at the end of the page under "How this classification was reached".