International Journal for Research in Applied Science & Engineering Technology (IJRASET) ISSN: 2321-9653; IC Value: 45.98; SJ Impact Factor: 7.538 Volume 10 Issue I Jan 2022- Available at www.ijraset.com The remaining 14-no attributes is prepared in an Attribute-Relation File Format (arff) for the REPTree machine learning phase described in section C. The description of the data attributes is as presented in Table 1 indicating the data type attributes. B. Data Preprocessing This study applies preprocessing tasks to remove irrelevant contents, as proposed in [21] through the following steps: 1) Transformations: Including conversion of SN 15 opinion to lower case 2) Noise Removal: Elimination of punctuations, white spaces etc. 3) Tokenization: Includes tokenization of texts with Regexp approach. A uni-gram approach of word-tokenization is implemented on the opinions. 4) Filtering: Exclusion of stop words including articles, conjunctions, and prepositions that do not carry enough discriminative content needed for the opinion-mining task.
Fig. 1. Proposed study framework
S/N 1.
2.
3.
4.
5.
6.
Table 1. Nature of dataset attribute Questions/Attribute Attribute_id Response options Staff Category s_cadre Teaching; Hostel porter; NonTeaching; Medical Staff How often do you use mask_wearing Always-Oftenyour nose/face mask? SometimesRarely-Never What is your salary staff_cat Below #120,000; level? #120,000 and above Have you had COVID- covid_test Yes; No 19 test and or vaccination before? How well is COVID- college_handling Not at all 19 precautions being satisfied; slightly handled in your satisfied; College? moderately satisfied; very satisfied; completely satisfied How well are your std_compl Not at all students complying compliant; slightly with COVID-19 safety compliant; measures? moderately compliant; very compliant; extremely
Arff response_id TS;HP;NT;MS
M-A;M-O;MS;M-R;M-N SS; JS
Yes; No
CH-1;CH-2;CH3;CH-4;CH-5
CC-1;CC-2;CC3;CC-4;CC-5
©IJRASET: All Rights are Reserved | SJ Impact Factor 7.538 | ISRA Journal Impact Factor 7.894 |
998
