Hotel Reviews: Sentiment Analysis & Topic Summarization
Click the Poster to View Full Screen, Right click to save image
Grant: Freshman Research Initiative
Olivia Tirso
College:
The Dorothy and George Hennings College of Science, Mathematics, and Technology
Major:
Computer Information Systems
Faculty Research Advisor(s):
Daehan Kwak
Abstract:
This research addresses the challenge of summarizing hotel reviews effectively by categorizing
them into key aspects. The goal is to comprehensively analyze customer sentiments and
hotel-related topics, providing valuable insights for businesses and travelers. By aggregating and
summarizing reviews, patterns, and trends can be identified, helping hotels understand strengths
and areas for improvement. The research aims to bridge the gap between review data and
actionable insights, enhancing the overall hotel experience.
The objective is to deliver an extensive analysis of customer sentiments and hotel-related topics,
simplifying decision-making for individuals overwhelmed by numerous reviews. This research
categorizes hotel reviews into cleanliness, service quality, cost, and location, offering nuanced
insights for informed decision-making. Using nltk and pandas, the study sets up libraries for
sentiment analysis with Vader, ensuring lexicon and rules are downloaded. Comprehensive
datasets from GitHub and Kaggle are analyzed, exhibiting diversity in content and format. The
research facilitates hotels in optimizing offerings and deepens the connection between
operational strategies and evolving customer preferences. Concise summaries also serve as
valuable resources for travelers.
Hotel reviews are crucial in shaping perceptions and decision-making, often overwhelming
consumers with information. Summarizing reviews helps identify patterns, assisting hotels in
understanding strengths and addressing faults. Sentiment analysis using Vader is performed on
review data, and word frequencies are calculated and visualized using Pandas, WordCloud, and
Matplotlib. The results include Vader scores, word cloud visualizations, and a rating
classification for each review.
This research significantly contributes to sentiment analysis and topic modeling, focusing on
hotel reviews but providing a blueprint for broader applications. It explores practical applications
of advanced techniques, enhancing understanding and offering insights for successful
implementation in various industries. The knowledge extends beyond the hospitality sector,
empowering businesses in diverse sectors to leverage customer reviews for informed,
data-driven decision-making.