Shelby Bice (Research Plan)

From edegan.com
Jump to navigation Jump to search

Overview

Overall goals:

  • Create better database that includes all the patent data to which the McNair Center has access.
  • More importantly, create documentation of process so it can improved upon/replicated in the future.

General Outline - updated 2/21/2017

  • Familiarize myself with SQL, Perl, and database design
  • Familiarize myself with existing scripts and schema for existing database
  • Design a better representation for database
  • Fix scripts if necessary
  • Start moving data into new database by querying existing databases (using SQL)
  • Use scripts to query new data
  • Test database
  • Remove extraneous information from database (copies, patents that we're not interested in, etc.)

Documentation I need to include:

  • Schema of new database (with justification of design), would like to include a visual representation
  • SQL commands that were used to fill database with explanation of what they do
  • Clear instructions on where to find scripts in bulk drive and an explanation of what each script does
  • Visual representation of example table entries that isn't just copied and pasted from a CSV file

Project Pages: Shelby Bice (Redesigning Patent Database)

Log

2/16/2017 - Talked over project with Ed, began reading existing wiki pages related to patent data and databases

2/21/2017 - Brushed up on SQL, Entity - Relationship model of designing databases

  • In the documentation, I want to briefly explain what the entity-relationship model is before including

the diagram so that readers have a little bit of background

  • Found a tool for creating a visual representation called ERDPlus.com - create a standalone instead of an account, can download

Learning commands from Patent Data - SQL Steps

  • copy command is PostgreSQL that copies a SQL table to a text file
    • DELIMITER set what will separate columns in text file
    • HEADER specifies that there will be a header in the text file with the names of the columns
  • Definitely need to include more detail about what these do in the documentation
  • insert into command inserts a new entry into the table

2/23/2017 - Read great database design article, dug through some more wiki articles, started reviewing Perl

  • What client do we use to interact with the current patent database?
  • Will need to determine all the fields that need to be included in the database before finishing the design and ER diagram, will need Ed's input

3/2/2017 - Started compiling a list of what fields to include and how they would be related.

  • Created an Excel spreadsheet that records the each table, their current attributes in the existing patentdata table, what I think the attributes should be in the new table, their relationship to a patent (i.e. one-to-many, many-to-many, etc.), their primary key, questions I have relating to the table, future steps for cleaning up the data in the table (i.e., once all the data has been move to the new database, removing patents that are not US-based), and current problems that have been recorded with the existing table for that information (if an existing table exists)
  • Once Excel spreadsheet is completed (and questions in the Questions column are answered or removed from the spreadsheet entirely) I will look into trying to embed it on my "Redesigning Patent Database" wiki page so that future users can sort of follow my thought process. I will also create separate wiki pages to explain each table once the new database is created
  • Making spreadsheet led me to realize there is some data that is repeated (filedate in fee table when it is also located in patent table, and the fee table includes the patent