How are lexicons created?
Mohamad's interest is in Programming (Mobile, Web, Database and Machine Learning). He is studying at the Center For Artificial Intelligence Technology (CAIT), Universiti Kebangsaan Malaysia (UKM).
There are different methods to create sentiment lexicons, which are collections of words or terms that are labeled as positive, negative, or neutral. Some of the common methods are:
• Manual annotation: This method involves human experts or crowdsourcing platforms to assign sentiment scores or labels to words based on their subjective judgment. This method can produce high-quality lexicons, but it is also time-consuming, costly, and prone to human bias or inconsistency. An example of a manually annotated lexicon is the AFINN lexicon (https://www.knime.com/blog/lexicon-based-sentiment-analysis/).
• Dictionary-based: This method relies on existing sentiment lexicons or general-purpose dictionaries to expand or adapt them to new domains or languages. This method can leverage the existing knowledge and resources, but it may also inherit the limitations or errors of the original lexicons or dictionaries. An example of a dictionary-based lexicon is SentiWordNet (https://www.knime.com/blog/lexicon-based-sentiment-analysis/).
• Corpus-based: This method uses statistical or machine learning techniques to extract sentiment words or phrases from large corpora of texts, such as reviews, tweets, or news articles. This method can discover new or domain-specific sentiment terms, but it may also require a lot of data and computational resources. An example of a corpus-based lexicon is VADER lexicon (https://www.knime.com/blog/lexicon-based-sentiment-analysis/).
