4

I tried to search a paper using semanticscholar with exact titles. But it seems that it just treated the title as a list of keywords and gave only irrelevant results. Is that possible to have semanticscholar to treat the query as a phrase rather than keywords? I tried to put enclose the query with quotes but semanticscholar seems to ignore it.

Take the example search below,

https://api.semanticscholar.org/graph/v1/paper/search?query=%22semi-supervised%20classification%20with%20graph%20convolutional%20networks%22

It returns ten papers and none of them is the original one. And the paper above is not an obscure one but a popular paper with more than 10K citations.

2
  • 2
    Not an answer, but removing the "-" from the query in "semi-supervised" gives the paper you are looking for. Perhaps it sees the minus sign as asking to exclude the term right after it, as search engines commonly do?
    – GoodDeeds
    Commented Jul 24, 2022 at 0:17
  • Thanks! It is a nice catch. At least it returns the paper as one of the candidates.
    – user559678
    Commented Jul 24, 2022 at 4:26

2 Answers 2

2

The API documentation does not mention anything about specific title searches.

Depending on your needs, I would suggest the following approach: https://api.semanticscholar.org/graph/v1/paper/search?query=title:(%22semi-supervised%20classification%20with%20graph%20convolutional%20networks%22)&limit=1

(Note that I added the title-tag, and a further limit=1 to ensure that only the most relevant title is visible.)

Another possibility would be for you to download the bulk dataset and search for exact titles there.

0

In Semantic Scholar API, endpoint /graph/v1/paper/search/match is for paper title search:

is intended for retrieval of a single paper based on closest title match to given query.

Reference documentation: https://api.semanticscholar.org/api-docs/#tag/Paper-Data/operation/get_graph_paper_title_search

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .