Research Interest

My Current Research:

A huge amount text data generated by various types of social networking sites (e.g., Facebook, Twitter, Instagram or Snapchat) are often used by various research by leveraging many data-driven techniques such as text mining, machine learning, or natural language processing. Particularly, in text mining, one of the key steps is to construct a text representation model. A well-known model for the text representation is the Vector Space Model (VSM). However, due to shortness and informality in tweets, using VSM to represent the tweets is limited by sparsity and scalability problems. To solve this issue, a graph-based representation model has been proposed based on certain network features such as graph-based pagerank, HITS, density (GB-PHD).

In my research, we propose a new graph-based tweet representation model based on Centrality, Weight features (GB-CW). We use the heart attack dataset as a use case to assess our model. The main aim in the use case to classify the tweets into either informational or non-Informational in relation to the context of heart attack.