Closed Bug 1400183 Opened 8 years ago Closed 8 years ago

TAAR: Compare existing recommenders

Categories

(Data Platform and Tools :: General, enhancement)

Product:
Data Platform and Tools
Component:
General
Type:
enhancement
Priority:
Not set
Severity:
normal

Tracking

(Not tracked)

Status:
RESOLVED FIXED

People

(Reporter: fhartmann, Assigned: fhartmann, Mentored)

References

Details

The individual addon recommenders we have so far require different data to make recommendations. The goal is to use the longitudinal dataset to find out how often they can make recommendations and whether they complement each other nicely. This should be done using a Jupyter notebook or using re:dash.
I have started with the analysis using a Jupyter notebook. The repository for this can be found on GitHub [1]. The pull request for the notebook is not merged yet. [1] https://github.com/florian/taar-prototyping
There's now a merged Jupyter notebook [1] for the analysis. As planned, the longitudinal dataset is used, but we also use other an addon data source as an additional filter. The results are summarized in the notebook but the general take-away is that the locale recommender can be used for nearly all users, while the collaborative and similarity recommenders are still useful for a good portion of users. [1] https://github.com/florian/taar-prototyping/blob/master/evaluation.ipynb
Status: NEW → RESOLVED
Closed: 8 years ago
Resolution: --- → FIXED
Moved to new component, per bug 1425844.
Component: General → Add-on Recommender
Component: Add-on Recommender → General
You need to log in before you can comment on or make changes to this bug.