Ranking Online Reviews Based on Consumer Preferences

To show a personalized review ranking list based on consumer preference is so necessary to the consumer that one can utilize online reviews effectively and efficiently to support his/her purchasing decisions. The paper formulates the ranking issue as an optimization problem and proposes a heuristic approximation algorithm to solve it stepwise and iteratively. In view of consumers' different reading behaviors, all possible review subsets that consumers would read from a ranking list are considered, and the objective of the algorithm is to maximize the expected matching degree between the consumer preference and the ranking list. Intensive experiments are conducted on real data, which reveal that ranking results from the proposed approach are better than those from the other related baseline methods.