Mining Opinion Words and Opinion Targets in a Two-Stage Framework
Xu, Liheng and Liu, Kang and Lai, Siwei and Chen, Yubo and Zhao, Jun

Article Structure

Abstract

This paper proposes a novel two-stage method for mining opinion words and opinion targets.

Introduction

Opinion mining not only assists users to make informed purchase decisions, but also helps business organizations understand and act upon customer feedbacks on their products or services in real-time.

Related Work

In opinion words/targets mining task, most unsupervised methods rely on identifying opinion relations between opinion words and opinion targets.

The First Stage: Sentiment Graph Walking Algorithm

In the first stage, we propose a graph-based algorithm called Sentiment Graph Walking to mine opinion words and opinion targets from reviews.

Topics

confidence score

Appears in 6 sentences as: confidence score (4) confidence scores (2)
In Mining Opinion Words and Opinion Targets in a Two-Stage Framework
  1. We speculate that it may be helpful to introduce a confidence score for each pattern.
    Page 2, “Introduction”
  2. For the second key, we utilize opinion words and opinion patterns with their confidence scores to represent an opinion target.
    Page 4, “The First Stage: Sentiment Graph Walking Algorithm”
  3. where conf denotes confidence score estimated by RWR, f req(-) has the same meaning as in Section 3.2.
    Page 5, “The First Stage: Sentiment Graph Walking Algorithm”
  4. where T is the opinion target set in which each element is classified as positive during opinion target refinement, s(ti) denotes confidence score exported by the target refining classifier.
    Page 5, “The First Stage: Sentiment Graph Walking Algorithm”
  5. Hence the confidence score conf in Equations (4) and (5) have no values and they are set to l. The initial labeled examples are exactly the same as Ours-Full.
    Page 6, “The First Stage: Sentiment Graph Walking Algorithm”
  6. This shows the effectiveness of Sentiment Graph Walking algorithm since the confidence scores estimated in the first stage are indispensable and indeed key to the learning of the second stage.
    Page 7, “The First Stage: Sentiment Graph Walking Algorithm”

See all papers in Proc. ACL 2013 that mention confidence score.

See all papers in Proc. ACL that mention confidence score.

Back to top.

dependency tree

Appears in 4 sentences as: dependency tree (4)
In Mining Opinion Words and Opinion Targets in a Two-Stage Framework
  1. For a given sentence, we first obtain its dependency tree .
    Page 3, “The First Stage: Sentiment Graph Walking Algorithm”
  2. Figure 1 gives a dependency tree example generated by Minipar (Lin, 1998).
    Page 3, “The First Stage: Sentiment Graph Walking Algorithm”
  3. Figure l: The dependency tree of the sentence “The style of the screen is gorgeous”.
    Page 3, “The First Stage: Sentiment Graph Walking Algorithm”
  4. The “OC-TC” pattern is the shortest path between an OC wildcard and a TC wildcard in dependency tree , which captures opinion relation between an opinion word candidate and an opinion target candidate.
    Page 3, “The First Stage: Sentiment Graph Walking Algorithm”

See all papers in Proc. ACL 2013 that mention dependency tree.

See all papers in Proc. ACL that mention dependency tree.

Back to top.

graph-based

Appears in 3 sentences as: graph-based (3)
In Mining Opinion Words and Opinion Targets in a Two-Stage Framework
  1. There were also many works employed graph-based method (Li et al., 2012; Zhang et al., 2010; Hassan and Radev, 2010; Liu et al., 2012), but none of previous works considered confidence of patterns in the graph.
    Page 2, “Related Work”
  2. In the first stage, we propose a graph-based algorithm called Sentiment Graph Walking to mine opinion words and opinion targets from reviews.
    Page 3, “The First Stage: Sentiment Graph Walking Algorithm”
  3. We can see that our graph-based methods (Ours-Bigraph and 0urs-Stage1 ) achieve higher recall than Zhang.
    Page 7, “The First Stage: Sentiment Graph Walking Algorithm”

See all papers in Proc. ACL 2013 that mention graph-based.

See all papers in Proc. ACL that mention graph-based.

Back to top.