Write some Software

# The goal is to build a simple publication database for scientific articles. The database is originally empty, but should be filled by accessing publication information from google scholar. For that purpose we need to: (1) Create an appropriate database, managed by MySQL. Importantly, the database should not only contain the standard bibliographical information for each paper (title, authors, journal reference, year) but also the citation graph between the papers. (2) Write a script (preferably in Python, but PHP or Perl would be considered if good reasons are given why they are preferable) that can selectively access data from google scholar and store the data in the database. (3) Importantly, that script should not only read the publication data, but also the citation graph (which publication cites a given publication) from google scholar, that is, for a given publication one needs to also create publication entries in the database for all publications citing that publication (but stop there, not for the second level of citations of citations). (4) Write a simple web-interface that can be used to fill the database by entering certain search queries (e.g. author name) for which the corresponding google scholar results are then stored in the database. (5) Write a simple web-interface that can be used to access the database itself by also entering a search query (e.g. author name) for which then the corresponding database entries are shown. # We have a linux server running ubuntu and abache2 web server. The above described database software and interface should be delivered with clear instruction for installation on our system, and with excellent documentation of the programming code to allow later modification. # There appears to be no official google scholar API. The script that accesses google scholar should be careful to conform with reasonable terms of service conventions, e.g. not too many search queries for google per minute. Speed is not our primary concern. It is no problem if a given query takes multiple minutes to get all required data. However, we certainly do not want to get blocked by google. # Further details of the database structure and interfaces can be discussed. However, at this point the main goal is to get a running system (as a proof of concept), which we can afterwards still optimize and add features to it.