Most dominant word sense achieves tremendous improvement in computation, now taking only milliseconds. Now only need to verify if this assumption preserves our hypothesis related to sentiment of draft.
Tue, Jun 20
I've added makefile rules and the features file but looks like we don't yet have language assets for Albanian in revscoring.
Mon, Jun 19
Removing stop words or restricting sentence to [-1,+1] during disambiguation did not give significant improvement.
Mon, Jun 12
Some dirty digging around performance, by profiling a script scoring long ( > 20) lines sentences:
Tue Jun 13 02:29:04 2017 stats
So I could setup a test with the library - https://github.com/kevincobain2000/sentiment_classifier/ that generates raw polarity scores for each document as an aggregate of positive and negative terms in the document. I used the https://github.com/wiki-ai/draftquality/blob/master/datasets/enwiki.draft_quality.75_not_OK_sample.censored.tsv and made the following observations:
Wed, Jun 7
Tue, May 30
May 19 2017
May 7 2017
removing assignee as this is the project task.
May 4 2017
Congrats to the selected participants for getting selected for GSoC 2017. This is a tracking task for your weekly project reports. Make sure each of you create a subtask of this task and also associate it as a subtask of your proposal, for weekly reporting as in the example.
Sorry for the confusion, restoring the previous state.
Apr 16 2017
Mar 22 2017
Mar 21 2017
@Yann, removing @amritsreekumar as task assignee. The parent task is open to participants to submit proposals to and the selected student can work on the project once the decision for the student of this project has been made. This is @SweatC 's proposal and has to be assigned to him.
I'm removing @amritsreekumar as the tasks's asignee. @amritsreekumar you need to submit a proposal for solving this project as mentioned above to be considered a candidate.
@Yann it'd be great if you could add one or two related microtasks that'd help you in judging the student/students for selection.