Using linkage information to improve the detection of relevant comment in social media
The vector space retrieval model relies on the notion of each comment is independence where the keywords influence to the document topic. However, such notion might not fit enough in the social media document called 'comment'. In social media comment, the occurrence of keywords does not gu...
Saved in:
| Main Authors: | , |
|---|---|
| Other Authors: | |
| Format: | Conference or Workshop Item |
| Published: |
2018
|
| Subjects: | |
| Online Access: | https://repository.li.mahidol.ac.th/handle/123456789/14003 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| Institution: | Mahidol University |
| Summary: | The vector space retrieval model relies on the notion of each comment is independence where the keywords influence to the document topic. However, such notion might not fit enough in the social media document called 'comment'. In social media comment, the occurrence of keywords does not guarantee the topic relevancy. Moreover, the absence of keywords does not guarantee the topic non-relevancy. These circumstances effect to the model accuracy because the social media language is relatively informal. Thus, people do not necessary to strict with the word usage in the proper meaning with respect to the conventional dictionary. We use the linkage information to create an augmented algorithm which improves the accuracy of the vector space retrieval model. Our experiment shows that our algorithm enhances the accuracy of the traditional vector space retrieval. © 2012 IEEE. |
|---|