COMPARING THE PERFORMANCE OF ONLINE FORUM VS ONLINE NEWS AS MARKET SENTIMENTS IN EXPLAINING THEIR RELATION WITH INDEX PERFORMANCES
Information is spreading widely online, where people from all around the world can access it. People can easily get information of others both from online news and also using social media, included the information of sector index performance. This research used large scale of study in sentiment anal...
Saved in:
| Main Author: | |
|---|---|
| Format: | Final Project |
| Language: | Indonesia |
| Online Access: | https://digilib.itb.ac.id/gdl/view/78746 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| Institution: | Institut Teknologi Bandung |
| Language: | Indonesia |
| Summary: | Information is spreading widely online, where people from all around the world can access it. People can easily get information of others both from online news and also using social media, included the information of sector index performance. This research used large scale of study in sentiment analysis to find the relationship of online forum and online news to consumer goods and mining sectors index performance in Indonesia, which are consist of 76 public traded firms. This research used 3 month weekdays data from January 1st 2015 – March 31st 2015.
Four thousand three hundred and fifty six (4.356) data were retrieved from online forum and online news for the sentiment analysis purpose. Bag –of-word and word weighted method were used in this study to determine the sentiment. The return of the sector index is measured for examining the sector performance in both sectors.
The result shows that online forum and online news have a positive relation to sector index return in both sectors. In addition, the sentiment from online forum has a bit stronger relation to sector index return compared to online news. The novelty of this research compared with previous research is using sector index data as the dependent variable. Furthermore, the sentiments used are from Indonesian language where the local content was analyzed as well in the bag-of-words. The limitation of this paper is only use online forum as the sources of social media data. For the future research, the author suggests to use more type of online forum media, such as Twitter or Facebook. Furthermore, the next study can use another financial aspect aside from sector index return to determine the relationship of sentiment analysis to the financial aspect. |
|---|