Extracting Product Features and Opinions from Reviews

By Ana-Maria Popescu, Oren Etzioni. Human Language Technology - The Baltic Perspectiv, 2005

2,226 citations recorded by the source corpus.

Consumers are often forced to wade through many on-line reviews in order to make an informed product choice. This paper introduces Opine, an unsupervised information-extraction system which mines reviews in order to build a model of important product features, their evaluation by reviewers, and their relative quality across products.Compared to previous work, Opine achieves 22% higher precision (with only 3% lower recall) on the feature extraction task. Opine's novel use of relaxation labeling for finding the semantic orientation of words in context leads to strong performance on the tasks of finding opinion phrases and their polarity.

Subfields: Artificial Intelligence

View Paper: https://www.semanticscholar.org/paper/ff75055d4e47737702d3b550879d6128cec13233

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Artificial Intelligence

1. B. Liu
2. Bing Liu
3. Erik Cambria
4. Philip S. Yu
5. Soujanya Poria
6. Percy Liang
7. Yongfeng Zhang
8. Noah A. Smith

Citations in the corpus

• Sentiment Analysis and Subjectivity (2010, methodology): “See a more complex method based on relaxation labeling in [77].”
• Daily-Aware Personalized Recommendation based on Feature-Level Time Series Analysis (2015, methodology): “In this work, we tune the parameters in the method introduced in [22, 39] and record the precision and recall.”
• Machine Reading at the University of Washington (2010, methodology): “…al., 2005); Opine builds on KnowItAll and mines product reviews via self-supervision using joint inference over neighborhood features (Popescu and Etzioni, 2005); Kylin pop-ulates Wikipedia infoboxes via self-supervision bootstrapping from existing infoboxes (Wu and Weld, 2007); LEX conducts…”
• Automated Rule Selection for Aspect Extraction in Opinion Mining (2015, methodology): “Apart from choosing the right rules, some heuristics were also proposed in [Popescu and Etzioni, 2005; Qiu et al. , 2011] to prune the extracted aspects to reduce errors.”
• Opinion Word Expansion and Target Extraction through Double Propagation (2011, methodology): “…our work (and also the non-propagation version, i.e., extracting targets using only the opinion words) with those in Hu and Liu (2004, henceforth Hu) and Popescu and Etzioni (2005, hence-forth Popescu), which also considered only explicit noun targets and experimented with the same data sets.”
• Articles: Recognizing Contextual Polarity: An Exploration of Features for Phrase-Level Sentiment Analysis (2009, background): “…1997; Kamps and Marx 2002; Turney and Littman 2003; Hu and Liu 2004; Kim and Hovy 2004; Esuli and Sebastiani 2005; Takamura, Inui, and Okumura 2005; Popescu and Etzioni 2005; Andreevskaia and Bergler 2006; Esuli and Sebastiani 2006a; Kanayama and Nasukawa 2006) to characterizing the sentiment of…”
• Word Sense and Subjectivity (2006, methodology): “…as review classification (e.g., (Turney, 2002; Pang and Lee, 2004)), mining texts for product reviews (e.g., (Yi et al., 2003; Hu and Liu, 2004; Popescu and Etzioni, 2005)), summarization (e.g., (Kim and Hovy, 2004)), information extraction (e.g., (Riloff et al., 2005)), 1Note that sentiment,…”
• Identifying comparative sentences in text documents (2006, positive): “[26, 8] improve the work in [11].”
• Tempura: Query Analysis with Structural Templates (2020, background): “…approach helps ground various studies on extracting semantic phrases (drug entities from online medical form [12], product features from reviews [30], etc.) Later, Ratner et al. [32] recognized pattern-based heuristics as one of the primary source for writing labeling functions in their data…”
• A review of natural language processing techniques for opinion mining systems (2017, methodology): “classification [35, 38, 60, 61, 62, 92] and entity/aspect extraction [31, 63, 64].”
• Using Linguistic Cues for the Automatic Recognition of Personality in Conversation and Text (2007, neutral): “... Girand, Graciarena, Kathol, Michaelis, Pellom, Shriberg, & Stolcke, 2005), speaker charisma (Rosenberg & Hirschberg, 2005), mood (Mishne, 2005), dominance in meetings (Rienks & Heylen, 2006), point of view or subjectivity (Wilson, Wiebe, & Hwa, 2004; Wiebe, Wilson, Bruce, Bell, & Martin, 2004; W”
• Lifelong-RL: Lifelong Relaxation Labeling for Separating Entities and Aspects in Opinion Targets (2016, methodology): “Note that relaxation labeling was used for sentiment classification in (Popescu and Etzioni, 2007), but not for target classification.”
• Deriving the Pricing Power of Product Features by Mining Consumer Reviews (2011, methodology): “Etzioni [ 23 ] presented OPINE, an unsupervised information extraction system capable of identi-”
• Recursive Neural Conditional Random Fields for Aspect-based Sentiment Analysis (2016, background): “This task was first studied by Hu and Liu (2004a; 2004b), followed by Popescu and Etzioni (2005), Zhuang et al. (2006), Zhang et al. (2010), Qiu et al. (2011), Li et al. (2010).”
• Recognising realistic emotions and affect in speech: State of the art and lessons learnt from the first challenge (2011, background): “In addition, the large related fields of opinion mining (Wilson et al., 2004; Popescu and Etzioni, 2005; Kim and Hovy, 2005; Missen and Boughanem, 2009) and sentiment analysis (Yi et al., 2003; Fei et al., 2006; Godbole, 2007; Missen and Boughanem, 2009) in text bear interesting alternatives and…”