Use sklearn TfidfVectorizer instead of gensim dictionary and tf-idf model. Fixes #81

[osma]

No description provided.

[codecov]

Codecov Report

Merging #84 into master will increase coverage by 0.16%.
The diff coverage is 90.62%.

@@            Coverage Diff             @@
##           master      #84      +/-   ##
==========================================
+ Coverage   95.17%   95.34%   +0.16%     
==========================================
  Files          29       29              
  Lines        1161     1138      -23     
==========================================
- Hits         1105     1085      -20     
+ Misses         56       53       -3

Impacted Files	Coverage Δ
annif/corpus/subject.py	83.33% <ø> (-2.67%)	⬇️
tests/test_cli.py	100% <100%> (ø)	⬆️
tests/test_backend_tfidf.py	100% <100%> (ø)	⬆️
annif/util.py	100% <100%> (ø)	⬆️
annif/corpus/__init__.py	100% <100%> (ø)	⬆️
annif/backend/tfidf.py	100% <100%> (ø)	⬆️
annif/project.py	93.38% <76.92%> (+1.65%)	⬆️

---

Continue to review full report at Codecov.

Legend - Click here to learn more
Δ = absolute <relative> (impact), ø = not affected, ? = missing data
Powered by Codecov. Last update c272acf...bdeec43. Read the comment docs.

[coveralls]

Coverage increased (+0.2%) to 95.347% when pulling 49ca399 on sklearn-vectorizer into c272acf on master.

[coveralls]

Coverage increased (+0.2%) to 95.343% when pulling bdeec43 on sklearn-vectorizer into c272acf on master.