Here are some notes on Named Entity Recognition (NER). This is the task of recognizing when a noun phrase in natural language text refers to a person, organization or place that is named - like President Obama, Denver CO, or the United Nations. 

There have been several conferences organized around this computer task.  They have included recognition problems that a number of teams have tackled.  The summaries of those results and the descriptions from the teams of their algorithms is a good way to survey the fields.  Here are a few of the papers. 

CoNLLnum1.pdf CoNLLnum2.pdf CoNLLnum3.pdf CoNLLnum4.pdf isoquest.pdf ltg_muc7.pdf marsh_slides.pdf

Here's a description of several algorithms used at the Stanford NLP group.  jenny-ner-Stanford-2007.pdf

We'll cover Bikel's paper describing an HMM model that he and his colleagues developed. 

algthatlearns-Bikel.doc.pdf