Academia.eduAcademia.edu

Figure 19 – uploaded by Hamed Rubaiee

See full PDFdownloadDownload figure
TABLE. XXII. NUMBER OF TWEETS UTILIZED FOR DATA MODEL COMPONENT THREE

Table 19 XXII. NUMBER OF TWEETS UTILIZED FOR DATA MODEL COMPONENT THREE

Related Figures (37)

Fig. 1 illustrates the proposed framework for integrating CRM with CEM. The framework consists of three processes: (1) sentiment analysis for CEM, (2) classification using CRM, and (3) data tuning of CRM via CEM. In terms of data modelling, the main components of the three processes are very similar. The difference between process one and process two is in the input: one takes unstructured social media input, while the other takes a structured customer database. In process three, CRM data can be labelled automatically by CEM or vice versa. In this work, we will focus on the fine-tuning of CRM through CEM.
Fig. 1 illustrates the proposed framework for integrating CRM with CEM. The framework consists of three processes: (1) sentiment analysis for CEM, (2) classification using CRM, and (3) data tuning of CRM via CEM. In terms of data modelling, the main components of the three processes are very similar. The difference between process one and process two is in the input: one takes unstructured social media input, while the other takes a structured customer database. In process three, CRM data can be labelled automatically by CEM or vice versa. In this work, we will focus on the fine-tuning of CRM through CEM.
TABLE. II. © COMMON ARABIC WORDS AND STANDARD SAUDI DIALECT REPRESENTING THE CLASSES Two distance education employees who have experience in e-learning labelled the data manually. Positive tweets were given the label "1", while negative tweets were given the label "-1". Neutral tweets were given the label "0", and irrelevant tweets were deleted from the database. A survey with four questions for e-learning criteria 2 was created and made available on Twitter via hashtags (see Table IID). Three questions asked a specific question on the criteria, while the last one was an open question asking students about their opinions in general about e-learning. Table 4 shows the four questions asked of participants.
TABLE. II. © COMMON ARABIC WORDS AND STANDARD SAUDI DIALECT REPRESENTING THE CLASSES Two distance education employees who have experience in e-learning labelled the data manually. Positive tweets were given the label "1", while negative tweets were given the label "-1". Neutral tweets were given the label "0", and irrelevant tweets were deleted from the database. A survey with four questions for e-learning criteria 2 was created and made available on Twitter via hashtags (see Table IID). Three questions asked a specific question on the criteria, while the last one was an open question asking students about their opinions in general about e-learning. Table 4 shows the four questions asked of participants.
TABLE. III. E-LEARNING CRITERIA
TABLE. III. E-LEARNING CRITERIA
Then, the same pre-processing steps were performed: tokenization, removal of stop words, light stemming, filtering tokens by length, and application of the N-gram feature. Next, the 'Process Documents from Data’ operator generated a word vector from the dataset after pre-processing and represented the text data as a matrix to show the frequency of occurrence of each term; then, relevant data mining approaches, i.e., NB and SVM, were explored for building models to classify tweets according to their sentiment polarity into positive, negative, and neutral. Finally, an evaluation was carried out using precision and recall methods. TABLE. IV. A SURVEY WITH FOUR QUESTIONS FOR E-LEARNING CRITERIA
Then, the same pre-processing steps were performed: tokenization, removal of stop words, light stemming, filtering tokens by length, and application of the N-gram feature. Next, the 'Process Documents from Data’ operator generated a word vector from the dataset after pre-processing and represented the text data as a matrix to show the frequency of occurrence of each term; then, relevant data mining approaches, i.e., NB and SVM, were explored for building models to classify tweets according to their sentiment polarity into positive, negative, and neutral. Finally, an evaluation was carried out using precision and recall methods. TABLE. IV. A SURVEY WITH FOUR QUESTIONS FOR E-LEARNING CRITERIA
\BLE. V. ACCURACY, PRECISION, AND RECALL FOR ALL CLASSES USING SVM AND NB CLASSIFIERS WITHOUT N-GRAM FOR QUESTION | TABLE. VI. ACCURACY, PRECISION, AND RECALL FOR ALL CLASSES USING SVM AND NB CLASSIFIERS WITH N-GRAM FOR QUESTION |
\BLE. V. ACCURACY, PRECISION, AND RECALL FOR ALL CLASSES USING SVM AND NB CLASSIFIERS WITHOUT N-GRAM FOR QUESTION | TABLE. VI. ACCURACY, PRECISION, AND RECALL FOR ALL CLASSES USING SVM AND NB CLASSIFIERS WITH N-GRAM FOR QUESTION |
TABLE. VII. ACCURACY, PRECISION, AND RECALL FOR ALL CLASSES USING FOR SVM AND NB CLASSIFIERS WITHOUT N-GRAM FOR QUESTION 2
TABLE. VII. ACCURACY, PRECISION, AND RECALL FOR ALL CLASSES USING FOR SVM AND NB CLASSIFIERS WITHOUT N-GRAM FOR QUESTION 2
TABLE. VIII. ACCURACY, PRECISION, AND RECALL FOR ALL CLASSES USING FOR SVM AND NB CLASSIFIERS WITH N-GRAM FOR QUESTION 2
TABLE. VIII. ACCURACY, PRECISION, AND RECALL FOR ALL CLASSES USING FOR SVM AND NB CLASSIFIERS WITH N-GRAM FOR QUESTION 2
TABLE. IX. ACCURACY, PRECISION AND RECALL FOR ALL CLASSES USING FOR SVM AND NB CLASSIFIERS WITHOUT N-GRAM FOR QUESTION 2
TABLE. IX. ACCURACY, PRECISION AND RECALL FOR ALL CLASSES USING FOR SVM AND NB CLASSIFIERS WITHOUT N-GRAM FOR QUESTION 2
\BLE. X. ACCURACY, PRECISION AND RECALL FOR ALL CLASSES USING FOR SVM AND NB CLASSIFIERS WITH N-GRAM FOR QUESTION 3
\BLE. X. ACCURACY, PRECISION AND RECALL FOR ALL CLASSES USING FOR SVM AND NB CLASSIFIERS WITH N-GRAM FOR QUESTION 3
ABLE. XI. ACCURACY, PRECISION, AND RECALL FOR ALL CLASSES USING FOR SVM AND NB CLASSIFIERS WITHOUT N-GRAM FOR QUESTION 4
ABLE. XI. ACCURACY, PRECISION, AND RECALL FOR ALL CLASSES USING FOR SVM AND NB CLASSIFIERS WITHOUT N-GRAM FOR QUESTION 4
TABLE. XII. ACCURACY, PRECISION, AND RECALL FOR ALL CLASSES USING FOR SVM AND NB CLASSIFIERS WITH N-GRAM FOR QUESTION 4
TABLE. XII. ACCURACY, PRECISION, AND RECALL FOR ALL CLASSES USING FOR SVM AND NB CLASSIFIERS WITH N-GRAM FOR QUESTION 4
TABLE. XIV. QUESTION | CRITERIA Table XV shows the criteria utilised to label record as positive, negative, or neutral for question |.
TABLE. XIV. QUESTION | CRITERIA Table XV shows the criteria utilised to label record as positive, negative, or neutral for question |.
TABLE. XV. QUESTION 2 CRITERIA Table XVI shows the criteria utilised to label record as Ositive, negative, or neutral for question 1.
TABLE. XV. QUESTION 2 CRITERIA Table XVI shows the criteria utilised to label record as Ositive, negative, or neutral for question 1.
TABLE. XVIII. ACCURACY, PRECISION, AND RECALL WITH ALL CLASSES FOR SVM AND NB CLASSIFIERS The following table shows the type of tuning to provide the best accuracy in King Abdul-Aziz University’s CRM classification.
TABLE. XVIII. ACCURACY, PRECISION, AND RECALL WITH ALL CLASSES FOR SVM AND NB CLASSIFIERS The following table shows the type of tuning to provide the best accuracy in King Abdul-Aziz University’s CRM classification.
The following table shows the type of tuning to provide the best accuracy in King Abdul-Aziz University’s CRM classification. XX shows the classification accuracy, precision, and recall for the NB and SVM classifiers: Crosse-validation=10, sampling type=stratified sampling. The best accuracy was achieved by NB due to the advantages of NB, such as its simplicity, ease of implementation, and combination of efficiency with acceptable accuracy.
The following table shows the type of tuning to provide the best accuracy in King Abdul-Aziz University’s CRM classification. XX shows the classification accuracy, precision, and recall for the NB and SVM classifiers: Crosse-validation=10, sampling type=stratified sampling. The best accuracy was achieved by NB due to the advantages of NB, such as its simplicity, ease of implementation, and combination of efficiency with acceptable accuracy.
TABLE. XX. ACCURACY, PRECISION, AND RECALL WITH ALL CLASSES FOR SVM AND NB CLASSIFIERS The following table shows the type of tuning to provide the best accuracy in King Abdul-Aziz University’s CRM classification. Table XXI shows the classification accuracy, precision, and recall for the NB and SVM classifiers: Crosse-
TABLE. XX. ACCURACY, PRECISION, AND RECALL WITH ALL CLASSES FOR SVM AND NB CLASSIFIERS The following table shows the type of tuning to provide the best accuracy in King Abdul-Aziz University’s CRM classification. Table XXI shows the classification accuracy, precision, and recall for the NB and SVM classifiers: Crosse-
XIX shows the classification accuracy, precision, and recall for the NB and SVM classifiers: Crosse-validation=10, sampling type=stratified sampling. The best accuracy was achieved by NB due to the advantages of NB, such as its simplicity, ease of implementation, and combination of efficiency with acceptable accuracy.
XIX shows the classification accuracy, precision, and recall for the NB and SVM classifiers: Crosse-validation=10, sampling type=stratified sampling. The best accuracy was achieved by NB due to the advantages of NB, such as its simplicity, ease of implementation, and combination of efficiency with acceptable accuracy.
TABLE. XIX. ACCURACY, PRECISION, AND RECALL WITH ALL CLASSES FOR SVM AND NB CLASSIFIERS
TABLE. XIX. ACCURACY, PRECISION, AND RECALL WITH ALL CLASSES FOR SVM AND NB CLASSIFIERS
Fig. 2 shows a comparison between CRM labelling and CEM labelling. Thirty tweets were labelled similarly by CRM and CEM. On the other hand, 48 tweets were labelled dissimilarly by CEM and CRM. In other words, 62% of collected tweets were labelled differently by CRM and CEM, and only 38% were similarly labelled by CEM and CRM. Fig. 2. Comparing CEM labelling with CRM labelling for Question 1.
Fig. 2 shows a comparison between CRM labelling and CEM labelling. Thirty tweets were labelled similarly by CRM and CEM. On the other hand, 48 tweets were labelled dissimilarly by CEM and CRM. In other words, 62% of collected tweets were labelled differently by CRM and CEM, and only 38% were similarly labelled by CEM and CRM. Fig. 2. Comparing CEM labelling with CRM labelling for Question 1.
From Fig. 3, there are 46 positive tweets in CEM, compared to 31 tweets in CRM. In contrast, there are only five negative tweets in CEM, compared to 23 negative tweets in CRM. For the neutral class, there are 28 tweets in CEM and 25 in CRM. Fig. 3. Comparing CEM labelling class with CRM labelling class for Question 1.
From Fig. 3, there are 46 positive tweets in CEM, compared to 31 tweets in CRM. In contrast, there are only five negative tweets in CEM, compared to 23 negative tweets in CRM. For the neutral class, there are 28 tweets in CEM and 25 in CRM. Fig. 3. Comparing CEM labelling class with CRM labelling class for Question 1.
TABLE. XXI. ACCURACY, PRECISION AND RECALL WITH ALL CLASSES FOR SVM AND NB CLASSIFIERS.
TABLE. XXI. ACCURACY, PRECISION AND RECALL WITH ALL CLASSES FOR SVM AND NB CLASSIFIERS.
Fig. 4. Comparing CEM labelling with CRM labelling for Question 2. Fig. 5 shows the number of tweets with similar labelling between the CRM positive class and to the CEM (positive, negative, and neutral) labelling for Question 2. In CEM, 29 tweets were labelled as positive. By contrast, CRM labelled 11 of these tweets as negative, 10 as positive, and eight as neutral.
Fig. 4. Comparing CEM labelling with CRM labelling for Question 2. Fig. 5 shows the number of tweets with similar labelling between the CRM positive class and to the CEM (positive, negative, and neutral) labelling for Question 2. In CEM, 29 tweets were labelled as positive. By contrast, CRM labelled 11 of these tweets as negative, 10 as positive, and eight as neutral.
Fig. 5. Comparing CEM labelling with CRM labelling for Question 2.
Fig. 5. Comparing CEM labelling with CRM labelling for Question 2.
Fig. 7. Comparing CEM labelling with CRM labelling for Question 3. From Fig. 7, there were 28 positive tweets in CEM, compared to 36 tweets in CRM. In contrast, CEM had only 13 negative tweets, compared to 26 in CRM. The neutral class contained 25 tweets in CEM and 30 in CRM, a difference of only five tweets.
Fig. 7. Comparing CEM labelling with CRM labelling for Question 3. From Fig. 7, there were 28 positive tweets in CEM, compared to 36 tweets in CRM. In contrast, CEM had only 13 negative tweets, compared to 26 in CRM. The neutral class contained 25 tweets in CEM and 30 in CRM, a difference of only five tweets.
Fig.6. Comparing CEM labelling with CRM labelling for Question 3
Fig.6. Comparing CEM labelling with CRM labelling for Question 3
Fig. 8. Comparing CEM labelling with CRM labelling for Question 4.
Fig. 8. Comparing CEM labelling with CRM labelling for Question 4.
From Fig. 9, there were 18 positive tweets in CEM, compared to 31 tweets in CRM. In contrast, there were 26 negative tweets in CEM, compared to 20 negative tweets in CRM. For the neutral class, there were 35 tweets in CEM, compared to 28 in CRM. Fig. 9. Comparing CEM labelling with CRM labelling for Question 4.
From Fig. 9, there were 18 positive tweets in CEM, compared to 31 tweets in CRM. In contrast, there were 26 negative tweets in CEM, compared to 20 negative tweets in CRM. For the neutral class, there were 35 tweets in CEM, compared to 28 in CRM. Fig. 9. Comparing CEM labelling with CRM labelling for Question 4.
Values of the criteria were changed for aspects in Q2: Changes occurred in student attendance in CRM from seven days to 14 days. Table XXIV shows the comparison between the evaluation of the classification before and after the tuning. The results show that the accuracy after the tuning is higher than before for the aspects in Q2. TABLE. XXIII. EVALUATION FOR ACCURACY, PRECISION, AND RECALL WITH ALL CLASSES FOR SVM AND NB CLASSIFIERS FOR QUESTION.
Values of the criteria were changed for aspects in Q2: Changes occurred in student attendance in CRM from seven days to 14 days. Table XXIV shows the comparison between the evaluation of the classification before and after the tuning. The results show that the accuracy after the tuning is higher than before for the aspects in Q2. TABLE. XXIII. EVALUATION FOR ACCURACY, PRECISION, AND RECALL WITH ALL CLASSES FOR SVM AND NB CLASSIFIERS FOR QUESTION.
TABLE. XXIV. EVALUATION FOR ACCURACY, PRECISION, AND RECALL WITH ALL CLASSES FOR SVM AND NB CLASSIFIERS FOR QUESTION 2 Values of the criteria were changed for aspects in Q3: Changes occurred in evaluating the total number of posts, including discussions and tests. Table XXV_ shows the comparison between the evaluation of the classification before and after the tuning. The results show that the accuracy after the tuning is almost the same.
TABLE. XXIV. EVALUATION FOR ACCURACY, PRECISION, AND RECALL WITH ALL CLASSES FOR SVM AND NB CLASSIFIERS FOR QUESTION 2 Values of the criteria were changed for aspects in Q3: Changes occurred in evaluating the total number of posts, including discussions and tests. Table XXV_ shows the comparison between the evaluation of the classification before and after the tuning. The results show that the accuracy after the tuning is almost the same.
TABLE. XXV. EVALUATION FOR ACCURACY, PRECISION, AND RECALL WITH ALL CLASSES FOR SVM AND NB CLASSIFIERS FOR QUESTION 3 Q4 combines the aspects in QI, Q2, and Q3. Table XX VI shows the comparison between the evaluation of the classification before and after the tuning. The results show that the accuracy after the tuning is higher than before. This indicates that changing some aspects of the criteria can help to enhance the classification accuracy in the CRM data.
TABLE. XXV. EVALUATION FOR ACCURACY, PRECISION, AND RECALL WITH ALL CLASSES FOR SVM AND NB CLASSIFIERS FOR QUESTION 3 Q4 combines the aspects in QI, Q2, and Q3. Table XX VI shows the comparison between the evaluation of the classification before and after the tuning. The results show that the accuracy after the tuning is higher than before. This indicates that changing some aspects of the criteria can help to enhance the classification accuracy in the CRM data.
Fig. 10 shows a comparison between the CEM labelling and the distance education criteria with suggestion one for Question |. The number of positive tweets increases, as does the number of negative tweets, while the number of neutral tweets sharply decreases. Fig. 10. Comparing CEM labelling with CRM with the old and new criteria for Question 1.
Fig. 10 shows a comparison between the CEM labelling and the distance education criteria with suggestion one for Question |. The number of positive tweets increases, as does the number of negative tweets, while the number of neutral tweets sharply decreases. Fig. 10. Comparing CEM labelling with CRM with the old and new criteria for Question 1.
Fig. 11. Comparing CEM labelling with CRM with the old and new criteria for Question 2.
Fig. 11. Comparing CEM labelling with CRM with the old and new criteria for Question 2.
Fig. 13 shows a comparison between the CEM labelling and the distance education criteria with suggestion one for Question 1. The number of positive and neutral tweets increases, and the number of negative tweets decreases. Fig. 13. Comparing CEM labelling with CRM with the old and new criteria for Question 4.
Fig. 13 shows a comparison between the CEM labelling and the distance education criteria with suggestion one for Question 1. The number of positive and neutral tweets increases, and the number of negative tweets decreases. Fig. 13. Comparing CEM labelling with CRM with the old and new criteria for Question 4.
Fig. 12. Comparing CEM labelling with CRM with the old and new criteria for Question 3.
Fig. 12. Comparing CEM labelling with CRM with the old and new criteria for Question 3.
To sum up King Abdul-Aziz University’s validation CRM experiments, the best accuracy was achieved by NB due to the advantages of NB such as its simplicity, ease of implementation, and combination of efficiency and acceptable accuracy.
To sum up King Abdul-Aziz University’s validation CRM experiments, the best accuracy was achieved by NB due to the advantages of NB such as its simplicity, ease of implementation, and combination of efficiency and acceptable accuracy.
580 California St., Suite 400
San Francisco, CA, 94104
© 2025 Academia. All rights reserved