GiellaLT provides an infrastructure for rule-based language technology aimed at minority and indigenous languages, and streamlines building anything from keyboards to speech technology. Read more about Why. See also How to get started and our Privacy document.
NDS contains a few means of logging user interaction with the system:
The file is stored in the main NDS server folder, as user_log.txt.
The dictionary lookup log contains one line per entry for each lookup that the morpholexicon performs, from any source: the reader plugin and API, or from the web interface directly. The file is tab separated, and the format is the following, though the header isn’t included in the file:
user_input success source_lemma target_definitions source_lang target_lang
Some sample data:
frynse True frynse riesaldat; riessat, riessut nob sme
munnje False nob sme
munnje True munnje, mun til meg; jeg sme nob
It is meant for easy grepping.
The file morph_log.txt lists all lines that are not recognized by any
analyzer.
On the server, nginx logs are located in /var/log/nginx. All successful
responses (responses with status code 1xx, 2xx and 3xx) are in access.log,
while erroneous responses (status code 4xx and 5xx) are in error.log.
HTTP access logs allow for a little more detailed look at all of the resources transferred by each page view, and make it possible to look only at page-views to the API, or all page views. Anyone looking for statistics on usage of the reader plugin will need to take a look at this. It is greppable, but because of the file structure, it can be annoying to grep for very detailed things since so much data is contained in each line, without consistent delimiters between each piece of data. However, some good tools exist for browsing such log formats.
Google Analytics tracks demographic and geographic information about use, and tracks individual page views versus unique visitors. Many more detailed possibilities are available here: users’ devices, users’ languages, how users navigate the site, page views per visit, as well as live data for who is using the site currently.