Determining the polarity of a sentimentbearing expression requires more than a simple bag-of-words approach. In particular, words or constituents within the expression can interact with each other to yield a particular overall polarity. In this paper, we view such subsentential interactions in light of compositional semantics, and present a novel learningbased approach that incorporates structural inference motivated by compositional semantics into the learning procedure. Our experiments show that (1) simple heuristics based on compositional semantics can perform better than learning-based methods that do not incorporate compositional semantics (accuracy of 89.7 % vs. 89.1%), but (2) a method that integrates compositional semantics into learning performs better than all other alternatives (90.7%). We also find that “contentword negators”, not widely employed in previous work, play an important role in determining expression-level polarity. Finally, in contrast to conventional wisdom, we find that expression-level classification accuracy uniformly decreases as additional, potentially disambiguating, context is considered. 1

Textual information in the world can be broadly categorized into two main types: facts and opinions. Facts are objective expressions about entities, events and their properties. Opinions are usually subjective expressions that describe people’s sentiments, appraisals or feelings toward entities, events and their properties. The concept of opinion is very broad. In this chapter, we only focus on opinion expressions that convey people’s positive or negative sentiments. Much of the existing research on textual information processing has been focused on mining and retrieval of factual information, e.g., information retrieval, Web search, text classification, text clustering and many other text mining and natural language processing tasks. Little work had been done on the processing of opinions until only recently. Yet, opinions are so important that whenever we need to make a decision we want to hear others ’ opinions. This is not only true for individuals but also true for organizations. One of the main reasons for the lack of study on opinions is the fact that there was little opinionated text available before the World Wide Web. Before the Web, when an individual needed to make a decision, he/she typically asked for opinions from friends and families. When an organization wanted to find the opinions or sentiments of the general public about its products and services, it conducted opinion polls, surveys, and focus groups. However, with the Web, especially with the explosive growth of the usergenerated

We examine the viability of building large polarity lexicons semi-automatically from the web. We begin by describing a graph propagation framework inspired by previous work on constructing polarity lexicons from lexical

It is a challenging task to identify sentiment polarity of Chinese reviews because the resources for Chinese sentiment analysis are limited. Instead of leveraging only monolingual Chinese knowledge, this study proposes a novel approach to leverage reliable English resources to improve Chinese sentiment analysis. Rather than simply projecting English resources onto Chinese resources, our approach first translates Chinese reviews into English reviews by machine translation services, and then identifies the sentiment polarity of English reviews by directly leveraging English resources. Furthermore, our approach performs sentiment analysis for both Chinese reviews and English reviews, and then uses ensemble methods to combine the individual analysis results. Experimental results on a dataset of 886 Chinese product reviews demonstrate the effectiveness of the proposed approach. The individual analysis of the translated English reviews outperforms the individual analysis of the original Chinese reviews, and the combination of the individual analysis results further improves the performance. 1

In this paper, we investigate structured models for document-level sentiment classification. When predicting the sentiment of a subjective document (e.g., as positive or negative), it is well known that not all sentences are equally discriminative or informative. But identifying the useful sentences automatically is itself a difficult learning problem. This paper proposes a joint two-level approach for document-level sentiment classification that simultaneously extracts useful (i.e., subjective) sentences and predicts document-level sentiment based on the extracted sentences. Unlike previous joint learning methods for the task, our approach (1) does not rely on gold standard sentence-level subjectivity annotations (which may be expensive to obtain), and (2) optimizes directly for document-level performance. Empirical evaluations on movie reviews and U.S. Congressional floor debates show improved performance over previous approaches. 1

Online user reviews are increasingly becoming the de-facto standard for measuring the quality of electronics, restaurants, merchants, etc. The sheer volume of online reviews makes it difficult for a human to process and extract all meaningful information in order to make an educated purchase. As a result, there has been a trend toward systems that can automatically summarize opinions from a set of reviews and display them in an easy to process manner [1, 9]. In this paper, we present a system that summarizes the sentiment of reviews for a local service such as a restaurant or hotel. In particular we focus on aspect-based summarization models [8], where a summary is built by extracting relevant aspects of a service, such as service or value, aggregating the sentiment per aspect, and selecting aspect-relevant text. We describe the details of both the aspect extraction and sentiment detection modules of our system. A novel aspect of these models is that they exploit user provided labels and domain specific characteristics of service reviews to increase quality. 1.

Automated identification of diverse sentiment types can be beneficial for many NLP systems such as review summarization and public media analysis. In some of these systems there is an option of assigning a sentiment value to a single sentence or a very short text. In this paper we propose a supervised sentiment classification framework which is based on data from Twitter, a popular microblogging service. By utilizing 50 Twitter tags and 15 smileys as sentiment labels, this framework avoids the need for labor intensive manual annotation, allowing identification and classification of diverse sentiment types of short texts. We evaluate the contribution of different feature types for sentiment classification and show that our framework successfully identifies sentiment types of untagged sentences. The quality of the sentiment identification was also confirmed by human judges. We also explore dependencies and overlap between different sentiment types represented by smileys and Twitter hashtags. 1

Supervised polarity classification systems are typically domain-specific. Building these systems involves the expensive process of annotating a large amount of data for each domain. A potential solution to this corpus annotation bottleneck is to build unsupervised polarity classification systems. However, unsupervised learning of polarity is difficult, owing in part to the prevalence of sentimentally ambiguous reviews, where reviewers discuss both the positive and negative aspects of a product. To address this problem, we propose a semi-supervised approach to sentiment classification where we first mine the unambiguous reviews using spectral techniques and then exploit them to classify the ambiguous reviews via a novel combination of active learning, transductive learning, and ensemble learning. 1

The lack of Chinese sentiment corpora limits the research progress on Chinese sentiment classification. However, there are many freely available English sentiment corpora on the Web. This paper focuses on the problem of cross-lingual sentiment classification, which leverages an available English corpus for Chinese sentiment classification by using the English corpus as training data. Machine translation services are used for eliminating the language gap between the training set and test set, and English features and Chinese features are considered as two independent views of the classification problem. We propose a cotraining approach to making use of unlabeled Chinese data. Experimental results show the effectiveness of the proposed approach, which can outperform the standard inductive classifiers and the transductive classifiers. 1

We explore how features based on syntactic dependency relations can be utilized to improve performance on opinion mining. Using a transformation of dependency relation triples, we convert them into “composite back-off features ” that generalize better than the regular lexicalized dependency relation features. Experiments comparing our approach with several other approaches that generalize dependency features or ngrams demonstrate the utility of composite back-off features. 1