During the curation phase of a text exploration, it would be beneficial to be able to recode recoded words or phrases. The process of curation typically involves cleaning up the data, exploring the results, and then determining if another round of curation is necessary. Often times there are cumulative recodes that are necessary to get the context included into a term or phrase. For example, first I need to recode and group the terms Smelled and Smelling into Smells. Then I can find phrases of Smells Good or Smells Bad and so I would Add phrases to my term list. However, there are multiple variations of these phrases and if any of the phrases include a term that is already recoded, then the phrase can no longer be modified (even though the newly recoded phrase will show up in the Recoded term list). So I would no longer be able to modify Smelled Good and Smelling Good into a new phrase, "All Smells good phrases". An example can be seen with the sample Pet Survey. Steps: Analyze > Text Explorer Select all terms > Recode Group similar values > OK Select a few phrases such as Dogs Bark and Dogs Food Add phrase Term list Select the 2 added phrases Recode >Group these 2 phrases and give them a new Name (to make it easy to find) such as "Dog stuff" Not all phrases will be recoded (although a peek at the Managed Recodes will show that this change has been made). In this example it might not make sense to make "Dog stuff" but I am trying to show the issue. An actual example might include perfume comments. Imagine I first need to recode all related synonyms (Scent, scents, smell, smells, smelled, odor, fragrance....ect). Next I want to add phrases that provide some context such as Good scent, Great smell, Amazing perfume and recode the phrases into a Positive Scent or a Negative Scent group. This is where the recode evolution would be useful.
... View more
Scott Reese, Senior Scientist, Procter & Gamble A. Narayanan, Principal Scientist, Procter & Gamble
Higher quality text curation will result in higher quality insights from unstructured text (the same as with all other analysis). This talk will focus on a few examples of how to be more efficient with your text cleanup. Examples will include: recoding demonstration, using Genreg to identify stop words, finding the right number of topics and using Genreg approaches to focus on key topics.
... View more
Level: Beginner Scott Reese, Scientist, Procter & Gamble Amy Phillips, Principal Scientist, Procter & Gamble To improve and support P&G’s data-driven decision making, our team has been assessing the quantitative analytical needs of our internal JMP users. We use the analogy that providing internal support is similar to managing a consumer goods product. Instead of a typical product and package that we would market and sell to consumers, we translated our support efforts into a combination of the product (necessary tools), package (delivery of training) and communication (awareness of benefits). As part of this effort, we conducted internal surveys to understand which JMP platforms were being used, the existing skill levels of users and what areas of future learning were desired. This talk focuses on demonstrating basic consumer survey analysis using the Categorical platform with the results of the internal JMP usage survey. We will also discuss how we approached designing and placing the survey to encourage hearing from as many of our users as possible.
... View more