elifesciences / refract

Convert NLM XML files into the Lens JSON
Other
9 stars 3 forks source link

Click tracking service #14

Open michael opened 11 years ago

michael commented 11 years ago

Just some initial ideas. Maybe that could be a task for Graham. You could build a simple API for using using your favorite language and Amazon Infrastructure.

Here's some initial thoughts on the API we'd need.

PUT http://clicktracker.elifesciences.org/00011/click_stats

Body:

{
  "node": "figure_25"
}

Then we should be able to query a document for most interesting content.

GET http://clicktracker.elifesciences.org/00011/click_stats

->

[
{node: "figure_5", clicks: 6017}},
{node: "figure_8", clicks: 4212}},
...
]
ivangrub commented 11 years ago

One good thing to include instead of just the number of clicks per figure, to also include where they came from. This means if it came from the paragraphs or the resource panel. I think getting real usage stats on if people are using the focus mode would be useful. Also provides some insight into the reading style of the article.

Additionally, it would be interesting to track the amount of time that is spent on articles, the time spent on different areas of the article. You could envision providing that information to authors who would want to know how people are interacting with their publications.

On Thu, Aug 29, 2013 at 3:23 AM, Michael Aufreiter <notifications@github.com

wrote:

Just some initial ideas. Maybe that could be a task for Graham. You could build a simple API for using using your favorite language and Amazon Infrastructure.

Here's some initial thoughts on the API we'd need.

PUT http://clicktracker.elifesciences.org/00011/click_stats

Body:

{ "node": "figure_25" }

Then we should be able to query a document for most interesting content.

GET http://clicktracker.elifesciences.org/00011/click_stats

->

[ {node: "figure_5", clicks: 6017}}, {node: "figure_8", clicks: 4212}}, ... ]

— Reply to this email directly or view it on GitHubhttps://github.com/elifesciences/refract/issues/14 .

gnott commented 11 years ago

A simpler first step to collecting intra-article stats may be to implement the Google Analytics as suggested in

https://github.com/elifesciences/lens/issues/31

This would alleviate having to manage the service collecting the clicks and probably has some visualizations already built. This could be on the client side of lens, and more a more custom method could be introduced into refract later to improve on them.

michael commented 11 years ago

you're probably right.. i'll have a look at the api once i get some time. so hold it! ;)