Semantic Scholar - First test

I quickly had a look at Semantic Scholar and the first results did not impress me.
Of course, I tried searching using my name:

• The "improved" key phrases describing my papers seem to be not quite descriptive. Things like "peer review", "information overload", "implicit culture" or "implicit knowledge", "recommender systems" do not appear in the list. Instead, the list contains "SIC" - well, sometimes you cannot just drop the last S, "Systems for Implicit Culture Support" I mentioned in the papers were abbreviated as "SICS":)

Then I tried to search for "implicit culture" - the concept introduced by Blanzieri and Giorgini in early 2000, which I then extended in my PhD.

• Among 6400 results, the paper by Blanzieri and Giorgini is indeed the first - bravo! However, there is no my name in the list of authors with more than 2 papers. Moreover, the facets contains a lot of authors who worked just on "culture" (and not even on tacit/implicit knowledge), publication venues not related to "implicit culture" and "SIC" again.

OK, let's go for sth more mainstream. "Linked Open Data":

Datasets used (PubChem, DBPedia, Yago) indeed seem relevant

SScholar does not know that LOD and Linked Open Data is the same thing

In the list of authors I do not see any ranking by importance or influence - only ranking by the number of publications. Sorry, how is this different from Microsoft Academic Search, Google Scholar, etc?

Publication venues do not contain any journal, with most popular place to publish on LOD (then, on almost anything?:)) being Arxiv:

• Keyphrases (Ontology, Semantic Web, Data Source, Large Scale, Sparql, Dbpedia) seem relevant, but no so insightful. There are no specific keywords related to LOD-related problems or techniques. 

Overall, so far, Semantic Scholar did not impress me, as I do not see any better filtering by using "semantic concepts" - just selection of multiple not-so-nicely-mined keywords - but may be I am missing sth...