----------------------- REVIEW 1 ---------------------
SUBMISSION: 19
TITLE: Politextpy: A Toolkit for Analyzing Media Political Discourse
AUTHORS: Yekai Xu, Mingqi Xie and Tinghao Han

----------- Overall evaluation -----------
SCORE: 0 (borderline paper)
----- TEXT:
It is a nice end-to-end pipeline but has very limited contribution. for example, if the author has tried multiple topic extraction, maybe it made it a preferable package, than sklearn.
also, it would be nice if authors had an ipython notebook, to showcase with a sample dataset.
I checked the git page, codes are clean, but I am not convinced why I should use your lda rather than just using directly the code inside.

from gensim import corpora
Lda = gensim.models.ldamodel.LdaModel

----------------------- REVIEW 2 ---------------------
SUBMISSION: 19
TITLE: Politextpy: A Toolkit for Analyzing Media Political Discourse
AUTHORS: Yekai Xu, Mingqi Xie and Tinghao Han

----------- Overall evaluation -----------
SCORE: -2 (reject)
----- TEXT:
The submission proposes a toolkit to analyze text from political communications. I felt that the submission lacks a concrete problem definition and its motivation, as well as some substantial technical novelty which would justify its presence in the conference. Also, I am not clear on what exactly would the authors present as a demonstration, as there is no clear use case presented.

Additional comments:
- The choice of Sections is not as clear to me; are Sections such as "Reading Files" informative enough for the reader to be included in the submission?
- Posting a screenshot of the GitHub page corresponding to the toolkit is not informative and could be omitted by the submission.