Mining opinions using combination of bag of words and word sense disambiguation in the education domain
Opinions are essential for developing services, facilities, products, whereby the stakeholders in the market collect feedback from the customers and respond accordingly in various domains, such as politics, entertainment, health care, social networking and education. Researchers have filled big gaps...
Saved in:
| Main Author: | |
|---|---|
| Format: | Thesis |
| Published: |
2013
|
| Subjects: | |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| Summary: | Opinions are essential for developing services, facilities, products, whereby the stakeholders in the market collect feedback from the customers and respond accordingly in various domains, such as politics, entertainment, health care, social networking and education. Researchers have filled big gaps in the majority of the domains by constructing and developing data mining classifiers to classify the opinion polarity to positive, negative or neutral. The education domain has not received as much attention as the other domains in terms of collecting students’ opinions and building classifiers trained on dedicated education corpora. Therefore in this research, we propose an opinion mining framework using a combination of techniques from Web Crawling, Bag of Words (BoW) and Word Sense Disambiguation (WSD) to collect, process and classify opinions as positive, negative or neutral in the context of universities in Malaysia. A novel application of the "Inside-Outside-Begin" technique for WSD is included in the framework. |
|---|