So you did your research by looking at different books, magazines and Wikipedia articles about your domain and you managed to get a lot of information; so you proceed to interview some of the stakeholders you outlined in your previous work. You probably got long narratives about what they do and how they do it. Having that information is good but, there will probably be a lot of text that is not necessary for your understanding of the domain. Plus, having a body of text the length of a thesis won't help you comprehend the domain if it's not organized! This is where Domain Description Units and Indexing comes in!
When it comes to Domain Description Units, they are basically excerpts of text that you have taken out from narratives. These detail different phenomenon or concepts that aid you in gaining understanding of your domain. Usually, getting them from articles and websites are pretty straight forward because they are usually compiled in such a way that they are divided into discernible description units. It's more complicated with narratives because they might not be as easy divided up into units. You might have a description unit made of various paragraphs, each having one or two relevant sentences about the same phenomenon or concept that you are trying to model.
Once you have gotten them into description units, there is still some organization you might want to do. For instance, you might have different phenomenon that pertain to security or maybe various stakeholders describe the same phenomena but with different properties. This is where Indexing comes in.
Indexing is, in laymen terms, 'tagging' your information in such a way that makes it easier to organize your description units in order to use it for your understanding of the domain. This helps Domain Engineers take information about something and make it easier to understand.