<?xml version="1.0" standalone="yes"?>
<Paper uid="W06-1650">
  <Title>Automatically Assessing Review Helpfulness</Title>
  <Section position="4" start_page="0" end_page="0" type="intro">
    <SectionTitle>
1 Introduction
</SectionTitle>
    <Paragraph position="0"> Unbiased user-supplied reviews are solicited ubiquitously by online retailers like Amazon.com, Overstock.com, Apple.com and Epinions.com, movie sites like imdb.com, traveling sites like citysearch.com, open source software distributors like cpanratings.perl.org, and countless others. Because reviews can be numerous and varying in quality, it is important to rank them to enhance customer experience.</Paragraph>
    <Paragraph position="1"> In contrast with ranking search results, assessing relevance when ranking reviews is of little importance because reviews are directly associated with the relevant product or service. Instead, a key challenge when ranking reviews is to determine which reviews the customers will find helpful.</Paragraph>
    <Paragraph position="2"> Most websites currently rank reviews by their recency or product rating (e.g., number of stars in Amazon.com reviews). Recently, more sophisticated ranking schemes measure reviews by their helpfulness, which is typically estimated by having users manually assess it. For example, on Amazon.com, an interface allows customers to vote whether a particular review is helpful or not. Unfortunately, newly written reviews and reviews with few votes cannot be ranked as several assessments are required in order to properly estimate helpfulness. For example, for all MP3 player products on Amazon.com, 38% of the 20,919 reviews received three or fewer helpfulness votes. Another problem is that low-traffic items may never gather enough votes. Among the MP3 player reviews that were authored at least three months ago on Amazon.com, still only 31% had three or fewer helpfulness votes.</Paragraph>
    <Paragraph position="3"> It would be useful to assess review helpfulness automatically, as soon as the review is written.</Paragraph>
    <Paragraph position="4"> This would accelerate determining a review's ranking and allow a website to provide rapid feedback to review authors.</Paragraph>
    <Paragraph position="5"> In this paper, we investigate the task of automatically predicting review helpfulness using a machine learning approach. Our main contributions are: * A system for automatically ranking reviews according to helpfulness; using state of the art SVM regression, we empirically evaluate our system on a real world dataset collected from Amazon.com on the task of reconstructing the helpfulness ranking; and * An analysis of different classes of features most important to capture review helpfulness; including structural (e.g., html tags, punctuation, review length), lexical (e.g., ngrams), syntactic (e.g., percentage of verbs and nouns), semantic (e.g., product feature mentions), and meta-data (e.g., star rating).</Paragraph>
  </Section>
class="xml-element"></Paper>