Topic modeling of short texts
In this recipe, we will be using Yelp reviews. These are from the same dataset that we used in Chapter 3, Representing Text: Capturing Semantics. We will break the reviews down into sentences and cluster them using the gsdmm
package. The resulting clusters should be about similar aspects and experience, and while many reviews are about restaurants, there are also other reviews, such as those concerning nail salon ratings.
Getting ready
To install the gsdmm
package, you will need to create a new folder and then either download the zipped code from GitHub (https://github.com/rwalk/gsdmm) or clone it into the created folder using the following command:
git clone https://github.com/rwalk/gsdmm.git
Then, run the setup script in the folder you installed the package in:
python setup.py install
How to do it…
In this recipe, we will load the data, divide it into sentences, preprocess it, and then use the gsdmm
model to cluster the sentences...