Here is a quick video showing how to make a concept. 




Explorer Concept Modeler
Concept Modeler is a tool for additional analytics in Gavagai Explorer. You can define different concepts in Explorer and then analyze you data with respect to them. Same as models, concepts are independent from projects and one concept can be used in different projects. A concept is defined by its including terms. In contrast with topics, the including terms in a concept do not need to be synonyms or semantically similar terms, but they are elements that jointly define that concept. Let’s give an example.

In Hotel Reviews you might define a topic called breakfast, including the terms “breakfast” and “the breakfast”. Each of these terms that appear in a text we will know that the topic breakfast is included in the text. Now consider the two following sentences retrieving from texts:

They had excellent organic coffee available in the lobby.

I enjoyed starting the mornings with cinnamon latte and croissant.

None of the words “coffee”, “latte”, “croissant” or “mornings” can define the topic breakfast independently; as the case in the first sentence, but a combination of these words can tell us that the text might be related to the concept breakfast; as the second sentence.  

Therefore, the concept breakfast can be defined by set of words like {tea, coffee, latte, bread, cheese, jam, morning,…}. You can start defining the concept by a few most necessary words. Then Explorer will show you more suggestions to be added to your concept. When you save your project as Full Excel or Full CSV (see 9 “Save as” Button and Exporting the Result of your Analysis), Explorer analyzes your data with respect to your target concepts, and tells you the number of present terms relating to each concept for each text. It will also provide sentiments scores related to your target concept.