Choose the most suitable sentiment analysis tool

street signs

With the growing capabilities of machine learning and natural language processing, bunches of sentiment analysis tools are popping up. There are various sentiment analysis for different types of text, including social media sentiment analysis tools, text analysis software and tools for voice sentiment analysis, like tone analyzers.

Online sentiment tools

To choose a sentiment analysis tool for your project, the first thing you need to think about is whether you prefer using existing online tools or writing code to conduct the analysis.

If you hope to avoid coding, you can check these online sentiment analysis tools: MonkeyLearn, Tone Analyzer, MeaningCloud, these tools all have freemium versions of their plans. However, these may have limited features. 

Here are some sentiment analysis tools with more advanced features: Qualarro, Lexalytics, Critical Mention, Social Mention, Sentiment Analyzer. These tools all have premium plans, You can choose them purposefully for your text. 

If you are a student/staff of the University of Edinburgh you can access a ready-to-go tool to do Sentiment Analysis via the GALE Scholar Lab. NB you will need to log in via EASE to access the pages 

 

 

Sentiment analysis within programme languages

If you’re comfortable with coding, you can try using programme languages like Python or R to conduct the sentiment analysis.

If you prefer to use Python, you can use the NLTK Library

NLTK is a platform for work with human language data. It provides easy-to-use interfaces to over 50 corpora and lexical resources such as WordNet, along with a suite of text processing libraries for classification, tokenization, stemming, tagging, parsing, and semantic reasoning, wrappers for industrial-strength NLP libraries.

 

If you prefer to use R, you can use the Quanteda or the Tidytext libraries

The Quanteda package is an R package for managing and analysing textual data

Sentiment analysis using dictionaries can be applied to any text, tokens, or dfm using textstat_polarity() or textstat_valence(). This function takes the quanteda object as an input, along with a dictionary whose valence or polarity has been set. The two ways of setting dictionary values allow a user to weigh each key with a polarity weight, or each value within keys with a valence weight.

The Tidytext package provides access to several sentiment lexicons. Three general-purpose lexicons are

All three of these lexicons are based on unigrams, i.e., single words.

 

Thumbnail with title of the training

Comparing Sentiment Analysis Models in R

Go Back to the Sentiment Analysis Pathway