DefinePK

Research Objective

To uncover consistent patterns and explain variations in online travel ratings, comments, and reviews by analyzing millions of reviews from Pakistan's top online travel companies, Uber and Careem, using semantic affiliation analysis.

Methodology

The study employed a semantic association analysis approach. Data was collected from online reviews of Uber and Careem in Pakistan. Preprocessing involved data cleaning, tokenization, and removal of stop words and duplicates. Thematic words were extracted and classified by experts. Statistical analysis of bigram co-occurrence phrases was performed. K-Means clustering was used to analyze rating reviews and segment users based on their preferences.

Methodology Flowchart

                        graph TD
    A["Data Collection: Uber & Careem Reviews"] --> B["Data Preprocessing: Cleaning, Tokenization, Stop Word/Duplicate Removal"];
    B --> C["Semantic Association Analysis: Thematic Word Extraction & Classification"];
    C --> D["Bigram Co-occurrence Phrase Analysis"];
    B --> E["Rating Review Analysis: K-Means Clustering"];
    D --> F["Analysis of Thematical Words & Phrases"];
    E --> G["Analysis of User Segments & Ratings"];
    F --> H["Discussion & Interpretation"];
    G --> H;
    H --> I["Conclusion & Implications"];                    

Discussion

The research highlights the importance of online travel reviews in influencing consumer decision-making within the tourism sector. By analyzing content, sentiment, and patterns, valuable insights can be gained to help businesses understand customer preferences and improve services. The study also introduces a theoretical framework exploring user-generated content and social network influences in online travel reviews, suggesting that reviews reflect social connections and interpersonal relationships.

Key Findings

Significant differences were observed among channels concerning topical vocabulary, subject distribution, structural traits, and community links. Network visualization illustrated connections between key concepts and words. Passengers highly prioritize well-prepared trips and dependable transportation services, as indicated by common bigram co-occurrence terms like "horrible accident," "ride explanation," and "ride satisfaction." K-Means clustering identified four user segments based on their activity preferences and associated median ratings.

Conclusion

The study provides a method to better understand strategic issues in the travel sector and gain insights into popular assessments to better serve tourists, lodging establishments, and trade groups. The proposed methodology aids in pinpointing the source of issues and examining the effort needed for rectification, aiming to create a model that prioritizes impactful adjustments.

Fact Check

1. The study collected millions of reviews from Uber and Careem in Pakistan.
2. Bigram co-occurrence phrases like "horrible accident," "ride explanation," and "ride satisfaction" were found to be common.
3. K-Means clustering was used to segment users into four distinct groups based on their preferences.

Mind Map