The Topic Overview tab contains a visual representation of the topic model used for this study. A topic model is a statistical model that uses machine learning to determine the probability of correspondence between specified topics in a collection of documents. The model generates these topics on the basis of word co-occurrence. After running the model, we categorized topics into four categories: general research topics (using NSF categories), specific subfield topics, water budget topics, and method topics. These topics allow us to interpret which fields of water science have the most and the least comprehensive research—or in other words, which are the “bright” and “blind” spots of water science in Latin America.
Important note: the Spanish and Portuguese topic models rely on far smaller corpora than the English topic model below. Because of this, the other two models are not as comprehensive; other visualizations on the platform thus rely on the topics generated from the English corpus.
×
Topic Model: How-To
This text provides instructions for analyzing the topic model to interpret the intertopic distance and top 30 most relevant terms of each topic. Refer to the page FAQs for more tips on interpreting the information and reasoning behind the model.
The left panel projects the relative distance between each topic and the predicted frequency the term will appear in relevant topics. The right panel depicts the corpus-wide (blue) and topic-specific frequency (red) of each word.
Determine topic numbers relevant to your query using the Topic Labels at the bottom of the page.
Enter corresponding numbers into the Selected Topic search bar to display the Top-30 Most Salient Terms for each topic.
Hover over a specific word in the term list to view the predicted frequency it will appear in relevant topics.
To reset the model visualization, select Clear Topic
This searchable table contains labeling information for individual topics in the model. Each topic label corresponds to a specific topic shown above. In total, there are 5 NSF general research topics (e.g. physical sciences), 43 NSF specific topics representing subfields of research (e.g. geochemistry). Each topic is labeled by a number, theme, National Science Foundation (NSF) Specific topic, NSF General topic, and contains a brief description based on spatial scale, water budget (e.g. reservoirs), or methods (e.g. remote sensing). Irrelevant topics (“noise”) are unlabeled.
×
Topic Labels: How-To
To use this table, follow these steps:
Click on Search Options to apply filters to the table or use the search bar on the right to narrow down the number of topics. To order the rows alphabetically or numerically, click on the arrows next to each table header.
Click Download CSV or Download Excel to download the table as a comma-separated value or Excel file, respectively. Users can also click on specific rows to highlight them and download just those rows.
For more information on the articles used in the model, check the Document to Topic tab above.
Topic Number
Topic Label
Theme
NSF Specific
NSF General
Description
Topic Number
Topic Label
Theme
NSF Specific
NSF General
Description
Document to Topic
This tab displays the estimated number of articles in the corpus related to a topic and the country. Select a topic from the drop down menu below to see a predicted distribution of articles about a given country.
×
Document to Topic: How-To
To use this visualization, select a topic from the drop down menu below to view the predicted distribution of articles for the given countries. The graph displays the estimated number of articles in the corpus related to a topic and the country.
Article Listings
This searchable table contains information about the articles used in the model. With it, users can construct queries to find information about authors, publishers, and, in some cases, specific geographic features or areas. Irrelevant topics (“noise”) are unlabeled.
×
Article Listings: How-To
The table lists all the articles that make up the corpus and any relevant information including the author(s), title, year, source, DOI (digital object identifier), predicted country, top topic, and topic label. To use the table, follow these steps:
Click on Search Options to apply specific filters to the table or use the search bar to narrow down the number of papers.
To increase the number of rows in the table, click on the Show 10 Rows button and change the number of rows.
To order the rows alphabetically or numerically, click on the arrows next to each table header.
Click Download CSV or Download Excel to download the table as a comma-separated value or Excel file, respectively. Users can also click on specific rows to highlight them and download just those rows.
