Add time graph for annotator contributions

[xavierfav]

It would be nice to have a graph to visualize annotators contributions across time in the contribute page with "Annotators' ranking".
Not so useful, but could maybe motivate people in their progression...

[ffont]

I think this is in fact useful. The ranking is computed in the compute_annotators_ranking function. For every user, the total number of contributions is the sum of Annotation and Vote objects under her name (n_annotations = Annotation.objects.filter(created_by=user, sound_dataset__dataset=dataset).count() and n_votes = Vote.objects.filter(created_by=user, annotation__sound_dataset__dataset=dataset).count()).

To plot a graph showing, you should filter the above queries by the created_at property so you can get "total number of annotations by user X before date Y" and "total number of votes by user X before date Y" (to dot hat you need to understand how the Django ORM works, see this related example). If you can do that then you only need to define a range of dates (i.e. last 10 hours) and run the queries for each user and for each date. With that you get the data to build your "evolution" plot for the last 10 hours.

To actually plot it you can use any javascript graph plotting library. In Audio Commons mediator we use CanvasJS. See that example: http://canvasjs.com/editor/?id=http://canvasjs.com/example/gallery/line/site_traffic/ is quite easy.

[ffont]

I came across this project: https://github.com/wa0x6e/cal-heatmap and I thought that maybe we could use something like this in Freesound datasets to see recent contributions to a dataset.

[jordipons]

I also think these visualizations are useful in our case - IMO these can be seen as a gamification.

But just highlighting number of contributions, might be dangerous since some workers focus on going "fast" rather than getting "good" annotations. In order to acknowledge "good" workes, I propose including some sort of quality weighting to these curves which can be based on the quality estimate discussed in #31.

To this end, we could either weight the proposed curve, or show two rankings: i) # of contributions or ii) quality.

[xavierfav]

It is a good idea to use quality to weight the contribution graph.
One simple method done in crowdsourcing for calculating the reward, is to count the votes that are considered right (several worker voting with the same answer).
In our case we want two workers to give the same vote.
Thus, this weighted contribution graph cannot be done in "real time".
But we could calculate the total "valid" contributions of users each month/week.