Introduction
Twitter has become an inevitable channel for brand management. It has compelled brands to become more responsive to their customers. On the other hand, the damage it would cause can’t be undone. The 140 character tweets has now become a powerful tool for customers / users to directly convey messages to brands.
For companies, these tweets carry a lot of information like sentiment, engagement, reviews and features of its products and what not. However, mining these tweets isn’t easy. Why? Because, before you mine this data, you need to perform a lot of cleaning. These tweets, once extracted can come with unwanted html characters, bad grammar and poor spellings – making the mining very difficult.
Below is the infographic, which displays the steps of cleaning this data related to tweets before mining them. While the example in use is of Twitter, you can of course apply these methods to any text mining problem. We’ve used Python to execute these cleaning steps.
Download the PDF Version of this infographic and refer the python codes to perform Text Mining and follow your ‘Next Steps…’ -> Download Here
To view the complete article on effective steps to perform data cleaning using python -> visit here
If you like what you just read & want to continue your analytics learning, subscribe to our emails, follow us on twitter or like our facebook page.
Stemming is also an important step in text mining. You could include that too.
remember to deal with character encodings.
Thou r awwssomee
This is great. I’m doing a project on social media data mining for my capstone class for my MS in MIS, and this is very helpful. Thanks so much!
This is awesome, can you check the urls in the last step? They seem not working 🙂
Great resource. Thanks.