Navigation auf uzh.ch
Navigation auf uzh.ch
Places are usually ambiguous and context dependent. Place functionality as a context in place descriptions is one of the prominent and distinguishing features of place.
Mina Karimi; Mohammad Saadi Mesgari; Ross Purves
A comparative assessment of machine learning methods in extracting place functionality from textual content
In this paper, we try to extract and place functionality from online reviews shared on Tripadvisor. We tested two strategies: the whole review and only action verbs. Then, we compare the efficiency of three NLP methods in selecting and extracting features, as well as different machine learning text classifier methods to predict place functionality. Evaluation results show that using the whole text of the review and Bag-of-Words as feature selection method can extract place functionality better than the others. Among classifiers, Logistic Regression leads to higher performance.
