What is it about?
Short text data is widely disseminated and retrievable on numerous platforms, including social media and news article titles, resulting in a large amount of information being distributed. The process of extracting topics from text data enables faster retrieval of information. Topic extraction, also known as topic modeling, has a number of popular algorithms that are frequently used for both long and short text data. In this study, we used LDA, NMF, and GDSMM to see which one performed better in terms of information extraction.
Featured Image
Photo by Jason Goodman on Unsplash
Why is it important?
It is critical to understand topic modeling methods that work with short text data. Because we can then select the quickest and best topic modeling method to extract information from various scattered data. This can make it easier for us to follow trends, learn about the latest topics being hotly debated in society, and innovate to solve problems in areas that are currently in demand.
Perspectives
This article may appear simple, but I'm confident it contains vital information. We are surrounded by a large amount of short text data that disperse so quickly that it would be a shame if the data were not explored to find the information contained within it. Because data movement is so rapid, an easy, cheap, and quick information extraction process is also required. That is why the topic modelling algorithm is so important in accomplishing this. Choosing the correct algorithm will assist in finding relevant and accessible information for humans to interpret, allowing us to dig up the most recent information.
Nuraisa Novia Hidayati
Badan riset dan inovasi nasional
Read the Original
This page is a summary of: Performance Comparison of Topic Modeling Algorithms on Indonesian Short Texts, November 2022, ACM (Association for Computing Machinery),
DOI: 10.1145/3575882.3575905.
You can read the full text:
Contributors
The following have contributed to this page
