b'@online{Mukjherjee2017e,'b'\nTITLE = {Exploring Latent Semantic Factors to Find Useful Product Reviews},\nAUTHOR = {Mukherjee, Subhabrata and Popat, Kashyap and Weikum, Gerhard},\nLANGUAGE = {eng},\nURL = {http://arxiv.org/abs/1705.02518},\nEPRINT = {1705.02518},\nEPRINTTYPE = {arXiv},\nYEAR = {2017},\nABSTRACT = {Online reviews provided by consumers are a valuable asset for e-Commerce<br>platforms, influencing potential consumers in making purchasing decisions.<br>However, these reviews are of varying quality, with the useful ones buried deep<br>within a heap of non-informative reviews. In this work, we attempt to<br>automatically identify review quality in terms of its helpfulness to the end<br>consumers. In contrast to previous works in this domain exploiting a variety of<br>syntactic and community-level features, we delve deep into the semantics of<br>reviews as to what makes them useful, providing interpretable explanation for<br>the same. We identify a set of consistency and semantic factors, all from the<br>text, ratings, and timestamps of user-generated reviews, making our approach<br>generalizable across all communities and domains. We explore review semantics<br>in terms of several latent factors like the expertise of its author, his<br>judgment about the fine-grained facets of the underlying product, and his<br>writing style. These are cast into a Hidden Markov Model -- Latent Dirichlet<br>Allocation (HMM-LDA) based model to jointly infer: (i) reviewer expertise, (ii)<br>item facets, and (iii) review helpfulness. Large-scale experiments on five<br>real-world datasets from Amazon show significant improvement over<br>state-of-the-art baselines in predicting and ranking useful reviews.<br>},\n}\n'