Graph Database Setup - Convert XML Data to Neoj4

This project involves writing software to establish a Neoj4 database on a Windows server running in Amazon AWS. Code should be written in PHP to import the MEDLINE database which is currently in XML format and transfer that data to a Neoj4 database. Details regarding the MEDLINE database are below. A test set of 30,00 items is available online. The full dataset contains about 23 million records and is about 120 Gigabytes in size. The PHP software should automatically download the MEDLINE data which is available over ftp as a collection of multiple files. This data set should be downloaded once annually when the full data set is released and then interim updates during the year should be downloaded and added to the database. The programmer should consult with the employer on the structure of the Neoj4 database; in general, each XML data element type will be a node and each record will represent a property of that node. Employer will also provide access to a hosted Wordpress site at Wpengine. The programmer should develop a basic Wordpress site which allows for access of the Neo4j database through 3 different graphical visualization tools (built-in Neo4j browser, [login to view URL] library, and Linkurious) as described at [login to view URL] . Sample NLM® Data INSTRUCTIONS FOR FTP OF SAMPLE RECORDS FTP to the NLM anonymous FTP server: ftp://[login to view URL] (login as a non-fee/anonymous user; use your e-mail address as password) You will see a directory for each NLM database. Go to the directory you want and get the desired files. MEDLINE®/PubMed® (includes approximately 98% of all records in PubMed) NLM distributes MEDLINE/PubMed data in XML format. 2015 Production Year Data Sample data using the current NLMMedlineCitationSet DTD are available: A small sample file of representative records covering each of the five status categories of records distributed to MEDLINE/PubMed licensees (i.e., MEDLINE, In-Data-Review, In-process, PubMed-not-MEDLINE, and OLDMEDLINE). Eight large sample files, each in .gz and .zip format, each containing 30,000 records (see access instructions at the top of this page). These files contain records in MEDLINE, PubMed-not-MEDLINE, and OLDMEDLINE statuses. Note that maintained versions of all sample records may be exported to licensees during the year. Documentation A document describing the MEDLINE/PubMed XML data elements, including definitions of the record status categories is available at [login to view URL]