Generating Recommendations Based on Robust Term Extraction from Users' Reviews

Abstract

In this paper, we propose a technique to automatically describe items based on users' reviews in order to be used by recommender systems. For that, we extract items' features using a robust term extraction method that applies transductive semi-supervised learning to automatically identify aspects that represent the different subjects of the reviews. Then, we apply sentiment analysis in a sentence level to indicate the polarities, yielding a consensus of users regarding the features of items. Our approach is evaluated using a collaborative filtering method, and comparisons using structured metadata as baselines show promising results.

Multi-Criteria Review-Based Recommender System–The State of the Art

“…• The vast size of the items is considered a major problem because when the recommendation is made, the content of every item has to be examined to discover items that are most likely relatable to the user's interest [19]. This task is error-prone and time-consuming [20].…”

“…• The over-specialization problem occurs in this type of approaches because users do not receive diverse or new items because of the restriction in his profile regarding the description of similar items [20].…”

“…In this case, the items that are recommended to a user are based on his term-based profile. Researches that use this type of element prove its usefulness in improving the RS performance such as in [20], [61].…”

Exploiting feature extraction techniques on users’ reviews for movies recommendation

“…In a previous work, we tested two different filters [13]: (1) filter_DF, which removes the words that occur only in one document in the database and (2) filter_DF_N that also deletes those words that are not nouns. In our experiments, we found out that the filter_DF_N performed better since it provides a significantly smaller set of candidate words to be classified by the transductive learning step, and hence the overall outcome contains a smaller but more descriptive terms set.…”

“…In this article, we propose two new feature extraction techniques: classification terms and hierarchy aspects. The classification terms technique, which was briefly explored in [13], extracts terms by using transductive semi-supervised learning, while the hierarchy aspects technique uses an hierarchical clustering solution to identify topic-related document clusters and elect the most important words of each group to form aspects. With these techniques, we aggregate machine learning into the terms/aspect extraction, differing from the previous techniques, which only relied on simple heuristics.…”