New techniques and framework for sentiment analysis and tuning of CRM structure in the context of Arabic language
Authors
AL-Rubaiee, Hamed SaadIssue Date
2018-01Subjects
opinion miningcustomer relationship management
customer experience management
sentiment analysis
Arabic language
N550 International Marketing
Metadata
Show full item recordAbstract
Knowing customers’ opinions regarding services received has always been important for businesses. It has been acknowledged that both Customer Experience Management (CEM) and Customer Relationship Management (CRM) can help companies take informed decisions to improve their performance in the decision-making process. However, real-word applications are not so straightforward. A company may face hard decisions over the differences between the opinions predicted by CRM and actual opinions collected in CEM via social media platforms. Until recently, how to integrate the unstructured feedback from CEM directly into CRM, especially for the Arabic language, was still an open question. Furthermore, an accurate labelling of unstructured feedback is essential for the quality of CEM. Finally, CRM needs to be tuned and revised based on the feedback from social media to realise its full potential. However, the tuning mechanism for CEM of different levels has not yet been clarified. Facing these challenges, in this thesis, key techniques and a framework are presented to integrate Arabic sentiment analysis into CRM. First, as text pre-processing and classification are considered crucial to sentiment classification, an investigation is carried out to find the optimal techniques for the pre-processing and classification of Arabic sentiment analysis. Recommendations for using sentiment analysis classification in MSA as well as Saudi dialects are proposed. Second, to deal with the complexities of the Arabic language and to help operators identify possible conflicts in their original labelling, this study proposes techniques to improve the labelling process of Arabic sentiment analysis with the introduction of neural classes and relabelling. Finally, a framework for adjusting CRM via CEM for both the structure of the CRM system (on the sentence level) and the inaccuracy of the criteria or weights employed in the CRM system (on the aspect level) are proposed. To ensure the robustness and the repeatability of the proposed techniques and framework, the results of the study are further validated with real-word applications from different domains.Citation
AL-Rubaiee, H.S. (2018) 'New techniques and framework for sentiment analysis and tuning of CRM structure in the context of Arabic language'. PhD thesis. University of Bedfordshire.Publisher
University of BedfordshireType
Thesis or dissertationLanguage
enDescription
A thesis submitted to the University of Bedfordshire in partial fulfilment of the requirements for the degree of Doctor of PhilosophyCollections
The following license files are associated with this item:
- Creative Commons
Except where otherwise noted, this item's license is described as http://creativecommons.org/licenses/by-nc-nd/4.0/