Analyzing Geo-Text Data with NLP Tools
Author(s): Zhaonan Wang
In this lab, you will go through the data wrangling process with two types of geo-text data, with explicit geo-tags or implicit location mentions within text.
- Explicit geo-text dataset: 2145 business located in Illinois side of St. Louis, derived from Yelp Academic Dataset. Each data record also comes with a user review, in plain text. Refer here for data format (https://www.yelp.com/dataset/documentation/main) of utilized business and review data. Your task is to perform sentiment analysis on each review and map the polarity score onto a map.
- Implicit geo-text dataset: news reports usually mention various locations, like countries, states, and even local toponyms (place names). In this notebook, you will play with a toy corpus containing three chunks of online news about some dam failure events. Your task to extract location mentions buried in the unstructured text.
Keywords: Named Entity Recognition, Geo-Text Data, Sentiment Analysis