Forsta Plus Model Builder is an integral part of the Forsta Plus Genius text analytics application.
Model Builder is the application where you build the categorization model, which is used to analyze your text in Text Analytics. Model Builder houses Concept Miner, Forsta’s AI topic extraction tool, which helps you to understand your verbatim data, create topics to use in your categorization model, and speeds up the model building process.
The categorization model is built using a hierarchical tree structure and the nodes of the tree contain expressions, which are built up of keywords, Boolean operators and Concept Miner topics. These expressions determine the resulting categorization of your text.
For example, a shop could ask customers what they thought of the service they received while making their most recent purchase. A respondent's reply could read: "Your employees were very knowledgeable and helpful, but the floor was dirty and the shop was rather untidy." The meaning of this sentence and its clauses is obvious to a person who knows the language, and this sentence could result in a “staff knowledge” category (with keyword “knowledgeable”) and a “staff helpfulness” category (with keyword “helpful”), the “shop cleanliness” attribute (with keyword “dirty”) and “tidiness” attribute (with keyword “untidy”). However if you have several thousand such replies, all of them slightly different, extracting the categories by manual analysis becomes a laborious and inaccurate job. Analysis by computer then becomes the only feasible option. After automatic processing, the several thousand replies will provide the shop with a set of data they can use as a basis for changes such that the customer experience, and hopefully thereby customer retention, is improved.
Model Builder addresses the need to categorize and quantify topics and other information found in text. There is also, of course, the sentiment process, which adds the positive, neutral/mixed or negative sentiment analysis to the overall verbatim and also each of the categories, which further enriches your analysis. Note that Model Builder does not deal with the sentiment process – this happens in the Genius Text Analytics processing task from Forsta Plus.
During the analysis part of the process, each word in each response is compared against a database (the model) of words, groups of words and Concept Miner topics that could be used in responses to the question. The model must therefore contain all the terms and topics that are to be searched for in the responses. Model Builder is the application that assists you with creating and compiling the model.
As the vocabulary used in different situations can differ considerably, the model must be tuned for the specific customer domain. For example, the retail branch, call centers, consumer electronics branch, hotel branch and vehicle workshop branch etc. will all need different models. And even within branches, different specializations will need to be fine-tuned; for example a hardware store will need a different model than a chemist. Template building blocks are available from Forsta which can be used to speed-up the process.
Inactive Screen Lock
For security reasons, when the Model Builder application has been inactive for 30 minutes it will be locked and you will have to log back in to continue. If you have not logged back in within three hours, Model Builder will close and any work that has not been saved will be lost. If you log back in within the three-hour period, Model Builder will reopen at the place you were last working.
Warning
Save your work at regular intervals, and ensure you save before you leave Model Builder unattended.