Create your own gazetteer list

Question

I am new to Natural Language Processing and GATE.Currently I'm learning to use GATE / ANNIE . ANNIE's default gazetteer lists are great, but obviously they don't provide lists for everything. I need to create a list of characters in a story book.
Creating lists and adding entries to each and every list from GATE Gazetteer Editor (as mentioned in Gate manual 13.2.2) or using a text editor does not seem to be practicable .So anyone knows a method to create our own gazetteer lists other than,creating/editing directly through GATE or using a text editor?

Answer 1

As said in the GATE manual you can edit any of the existing lists in a text editor. Probably the most straight-forward way is to create these lists programatically. I.e. if you have them in a database, dump records in the gazetteer format (basically one word per line). If you have them in a csv or a web page, export them to the right format.

Another option is to use a more advanced gazetteer which uses an ontology or semantic repository. See the manual link above for different gazetteers and how to work with them.

Answer 2

I created a list using the contents of a column from a database table as suggested above. Simply saved it as a .lst file using Notepad++ in the same directory as all the other .lst files (I'm using the ANNIE gazetteer) and then added it using the gazetteer editor. One problem that I ran into was not having it saved in the correct encoding (UTF-8). GATE didn't like it and it showed in the messages when loading. Once I figured that out and corrected it, it worked fine.

If you need to create a list of entities from text maybe you could look into the gazetteer list collector - http://gate.ac.uk/sale/tao/splitch13.html - 13.7