[ CoBib ]
Collaborative indexing and annotation of bibliographic databases

 

Summer 2006

This eight-week term gave the team the opportunity to implement the fundamental functionality of the CoBib system, as well as solidify the underlying structure and our objectives regarding the user experience. We endured various tribulations, including days trying to "fix" the zebra server and disagreements regarding the future direction of CoBib. Nevertheless, working full-time during the summer allowed us to experience research in a concentrated manner.

Flow Charts

One of our goals for the summer was to iron out the details concerning the interaction between the databases and the user, as well as how particular sequences of events will take place along a timeline. These charts are a summary thus far of how we anticipate our server to act.

Graph Description
Uploading a Citation Shows the various sources from which bibliographic data can originate and the process towards creating new citation records. This includes citation matching and adding entries to both databases if necessary.
Uploading a Paper Similar to above, except with the addition of the data extraction step. This process will be the origination of the bibliographic data.
Searching for Citations Searching for a citation using bibliographic search terms interacts directly with the zebra server while, showing listings of citations from a tag or recommendation query deals with the MySQL server. After displaying a citation or tag, it's a cyclical process to find more stemming from that.
Annotating a Citation Demonstrates which database is used to allow the user to write both long annotations and word/phrase tags. Here, interaction with the zebra server is not necessary.
MySQL Tables How the individual tables in the MySQL database will relate to each other. Graphically displaying these connections will aid in our understanding of how to use the data to best accomplish both simple and complex tasks, especially those involving joins.

Functionality

Most of the basic functionality was implemented. In particular, the following features:

  • upload citation(s)
  • upload paper
  • tag citation
  • annotate citation
  • modify citation
  • search using bibliographic fields
  • edit user profile

Citation Matching

An integral aspect of removing database redundancies is developing an algorithm to determine whether a newly submitted citation refers to the same paper as a citation already in the database. We have done research into the field of string-matching, and implemented several tests using edit-distance, token-based, and hybrid algorithms.

 

Spring 2006 Summer 2006 Fall 2006 Spring 2007 Home