Impact Visualizer

Impact Visualizer is a web application maintained by Wiki Education, which can show how a set of articles — a Topic — has evolved over time. It can also highlight the contributions of a set of users (such as participants in an outreach program) within a Topic.

The instance of Impact Visualizer on Wikimedia Cloud — https://impact-visualizer.wmcloud.org/ — is available for use by the Wikimedia community. The code is managed on GitHub.

This example topic shows off the features of the Impact Visualizer, including the optional "Classification" feature that provides variant visualizations based on the Wikidata properties of articles within the Topic.

Overview

Impact Visualizer analyzes data based on a user-defined "Topic", which is a list of article titles from a single language version of Wikipedia, along with a list of users whose contributions within that Topic will be highlighted. Each Topic covers a date range, and shows how the articles within that Topic have evolved over time: when they were created, how much content was added over time, and how their estimated Article Quality scores changed over time.

How to add a topic

To add a new topic, you will need a CSV file of article titles, and a CSV file of usernames. The "Tools" menu provides some tools for generating CSVs from Wikidata properties, pre-existing PetScan queries or PagePile IDs, or courses or campaigns from Programs & Events Dashboard or Wiki Education Dashboard.

1. Log in to https://impact-visualizer.wmcloud.org/ using your Wikimedia account (via OAuth).
2. Click "Manage Your Topics", then "Create a New Topic"
3. Enter a title, slug (unique identifier that may form part of a URL) and description
4. Select the Wiki (Wikipedia language version). The system relies on data from the WikiWho API, so only languages available from that service are supported.
5. Enter a start and end date for the analysis. Starting from the beginning of Wikipedia (2001, in the case of English Wikipedia) is usually a good choice, and ending at the present is usually a good choice.
6. Optionally, set the "Timepoint Day Interval". If you are tracking a Topic over many years, a value of 365 days is a good idea.
7. Optionally, set the "Chart Time Unit". If you are tracking a Topic over many years, "Year" is the best option to keep the charts easily readable.
8. Optionally, check the "Convert Tokens to Words" and set a conversion value of "Tokens per Word". Some of the analysis will be based on "tokens" within the wikitext of the articles, as determined by the tokenization analysis from the WikiWho algorithm, the results are easier to understand if presented in terms of words. 3.25 tokens per word of readable prose is a rough estimate for English Wikipedia articles circa 2025.
9. Optionally, select one or more Classifications to apply to the Topic. Classifications provide additional visualizations based on data from the Wikidata items corresponding to the articles within the Topic.
10. Upload a CSV of article titles (one titles per line), and a CSV of usernames (one username per line).
11. Click "Submit". As soon as server resources are available, the system will import the list of articles and users from the CSV files; the status will be shown in the "Management Actions" section of your Topic page as these tasks progress.
12. Once the articles and users have been imported, click "Generate Timepoints". As soon as server resources are available, the system will begin gather data. For large topics (tens of thousands of articles), it may take several days to gather the required data.
13. Once the "Generate Timepoints" tasks are complete, the Topic page will show a variety of data visualizations.