Using empythy to score likability based on sentiment analysis of recent tweets about a given person
Project description
Using empythy to score likability based on sentiment analysis of recent tweets about a given person
Purpose
To piggyback off of the empythy natural languare classifier package to analyze average sentiment of tweets related to a particular person to calculate a ‘likability score’ for that person. Useful in tracking sentiment changes across a certain period of time, i.e. the likability score of a celebrity before and after a concert.
Instructions
- Open terminal. Make sure you have python3 and pip downloaded.
- pip install likability
- Create a csv file with the names of the people you’d like to analyze for likability. Name this file name.csv in the current directory.
- Determine how many recent tweets you’d like to query for each person. This will be used in the script below as num_tweets.
- Make sure you have Twitter API keys and access tokens. If you do not, go to [Twitter Apps](https://apps.twitter.com/), create an app, and find the required keys and tokens under Applications Settings -> Consumer Key (API Key) -> manage keys and access tokens.
- Run Python 3 by typing python into the terminal.
- Enter script below to run the LikabilityAnalyzer module.
- from likability import LikabilityAnalyzer
- filepath = 'name.csv'
- num_tweets = 100
- sentimentScore = LikabilityAnalyzer.analyzer(filepath,num_tweets)
- When prompted, enter in your Twitter API keys. This will allow likability to access the Twitter API to query the tweets needed to complete the sentiment analysis.
- Wait for script to run to completion. Please note, due to Twitter API Rate Limiting, querying more than 15 names will lead to longer wait times. Please allow 1 minute per name for lists greater than 15 names.
- Upon completion, open the newly created Sentiment.csv in the current directory to access the likability scores for each person.
Possible Usage
- Score top fantasy football players to see what the Twittersphere thinks about each player pre-draft
- Instead of names of people, use product names to track customer sentiment in real-time
- Solve the question: who is more likable, Justin Timberlake or Jimmy Fallon
Project details
Release history Release notifications | RSS feed
Download files
Download the file for your platform. If you're not sure which to choose, learn more about installing packages.
Source Distribution
lykability-1.0.0.tar.gz
(6.6 kB
view hashes)
Built Distribution
Close
Hashes for lykability-1.0.0-py2.py3-none-any.whl
Algorithm | Hash digest | |
---|---|---|
SHA256 | 5f3402deb36b370c2b9bf4b08673c340f8e6b1774167a5c4cb6a0e8bc9962873 |
|
MD5 | 9d5a225776f2de9478d16d645eeb787f |
|
BLAKE2-256 | a69142d6fa164eaf0423f47f0f4ae7f62ea44fd4e1bd7b9352bb6451bdf6e073 |