Self-training from labeled features for sentiment analysis

Sentiment analysis concerns about automatically identifying sentiment or opinion expressed in a given piece of text. Most prior work either use prior lexical knowledge defined as sentiment polarity of words or view the task as a text classification problem and rely on labeled corpora to train a sentiment classifier. While lexicon-based approaches do not adapt well to different domains, corpus-based approaches require expensive manual annotation effort. In this paper, we propose a novel framework where an initial classifier is learned by incorporating prior information extracted from an existing sentiment lexicon with preferences on expectations of sentiment labels of those lexicon words being expressed using generalized expectation criteria. Documents classified with high confidence are then used as pseudo-labeled examples for automatical domain-specific feature acquisition. The word-class distributions of such self-learned features are estimated from the pseudo-labeled examples and are used to train another classifier by constraining the model's predictions on unlabeled instances. Experiments on both the movie-review data and the multi-domain sentiment dataset show that our approach attains comparable or better performance than existing weakly-supervised sentiment classification methods despite using no labeled documents.

Semantic sentiment analysis of Twitter

Sentiment analysis over Twitter offer organisations a fast and effective way to monitor the publics' feelings towards their brand, business, directors, etc. A wide range of features and methods for training sentiment classifiers for Twitter datasets have been researched in recent years with varying results. In this paper, we introduce a novel approach of adding semantics as additional features into the training set for sentiment analysis. For each extracted entity (e.g. iPhone) from tweets, we add its semantic concept (e.g. Apple product) as an additional feature, and measure the correlation of the representative concept with negative/positive sentiment. We apply this approach to predict sentiment for three different Twitter datasets. Our results show an average increase of F harmonic accuracy score for identifying both negative and positive sentiment of around 6.5% and 4.8% over the baselines of unigrams and part-of-speech features respectively. We also compare against an approach based on sentiment-bearing topic analysis, and find that semantic features produce better Recall and F score when classifying negative sentiment, and better Precision with lower Recall and F score in positive sentiment classification.

Joint sentiment/topic model for sentiment analysis

Sentiment analysis or opinion mining aims to use automated tools to detect subjective information such as opinions, attitudes, and feelings expressed in text. This paper proposes a novel probabilistic modeling framework based on Latent Dirichlet Allocation (LDA), called joint sentiment/topic model (JST), which detects sentiment and topic simultaneously from text. Unlike other machine learning approaches to sentiment classification which often require labeled corpora for classifier training, the proposed JST model is fully unsupervised. The model has been evaluated on the movie review dataset to classify the review sentiment polarity and minimum prior information have also been explored to further improve the sentiment classification accuracy. Preliminary experiments have shown promising results achieved by JST.

Contextual semantics for sentiment analysis of Twitter

Sentiment analysis on Twitter has attracted much attention recently due to its wide applications in both, commercial and public sectors. In this paper we present SentiCircles, a lexicon-based approach for sentiment analysis on Twitter. Different from typical lexicon-based approaches, which offer a fixed and static prior sentiment polarities of words regardless of their context, SentiCircles takes into account the co-occurrence patterns of words in different contexts in tweets to capture their semantics and update their pre-assigned strength and polarity in sentiment lexicons accordingly. Our approach allows for the detection of sentiment at both entity-level and tweet-level. We evaluate our proposed approach on three Twitter datasets using three different sentiment lexicons to derive word prior sentiments. Results show that our approach significantly outperforms the baselines in accuracy and F-measure for entity-level subjectivity (neutral vs. polar) and polarity (positive vs. negative) detections. For tweet-level sentiment detection, our approach performs better than the state-of-the-art SentiStrength by 4-5% in accuracy in two datasets, but falls marginally behind by 1% in F-measure in the third dataset.

SentiCircles for contextual and conceptual semantic sentiment analysis of Twitter

Lexicon-based approaches to Twitter sentiment analysis are gaining much popularity due to their simplicity, domain independence, and relatively good performance. These approaches rely on sentiment lexicons, where a collection of words are marked with fixed sentiment polarities. However, words' sentiment orientation (positive, neural, negative) and/or sentiment strengths could change depending on context and targeted entities. In this paper we present SentiCircle; a novel lexicon-based approach that takes into account the contextual and conceptual semantics of words when calculating their sentiment orientation and strength in Twitter. We evaluate our approach on three Twitter datasets using three different sentiment lexicons. Results show that our approach significantly outperforms two lexicon baselines. Results are competitive but inconclusive when comparing to state-of-art SentiStrength, and vary from one dataset to another. SentiCircle outperforms SentiStrength in accuracy on average, but falls marginally behind in F-measure. © 2014 Springer International Publishing.

Improving the Lexicon of Positive/Negative Words and Bigrams for Sentiment Analysis

Report published in the Proceedings of the National Conference on "Education in the Information Society", Plovdiv, May, 2013

Social media and sentiment in bioenergy consultation

he push to widen participation in public consultation suggests social media as an additional mechanism through which to engage the public. Bioenergy companies need to build their capacity to communicate in these new media and to monitor the attitudes of the public and opposition organisations towards energy development projects. Design/methodology/approach This short paper outlines the planning issues bioenergy developments face and the main methods of communication used in the public consultation process in the UK. The potential role of social media in communication with stakeholders is identified. The capacity of sentiment analysis to mine opinions from social media is summarised, and illustrated using a sample of tweets containing the term ‘bioenergy’ Findings Social media have the potential to improve information flows between stakeholders and developers. Sentiment analysis is a viable Purpose The push to widen participation in public consultation suggests social media as an additional mechanism through which to engage the public. Bioenergy companies need to build their capacity to communicate in these new media and to monitor the attitudes of the public and opposition organisations towards energy development projects. Design/methodology/approach This short paper outlines the planning issues bioenergy developments face and the main methods of communication used in the public consultation process in the UK. The potential role of social media in communication with stakeholders is identified. The capacity of sentiment analysis to mine opinions from social media is summarised, and illustrated using a sample of tweets containing the term ‘bioenergy’ Findings Social media have the potential to improve information flows between stakeholders and developers. Sentiment analysis is a viable methodology, which bioenergy companies should be using to measure public opinion in the consultation process. Preliminary analysis shows promising results. Research limitations/implications Analysis is preliminary and based on a small dataset. It is intended only to illustrate the potential of sentiment analysis and not to draw general conclusions about the bioenergy sector. Originality/value Opinion mining, though established in marketing and political analysis, is not yet systematically applied as a planning consultation tool. This is a missed opportunity.

Learning user and product distributed representations using a sequence model for sentiment analysis

In product reviews, it is observed that the distribution of polarity ratings over reviews written by different users or evaluated based on different products are often skewed in the real world. As such, incorporating user and product information would be helpful for the task of sentiment classification of reviews. However, existing approaches ignored the temporal nature of reviews posted by the same user or evaluated on the same product. We argue that the temporal relations of reviews might be potentially useful for learning user and product embedding and thus propose employing a sequence model to embed these temporal relations into user and product representations so as to improve the performance of document-level sentiment analysis. Specifically, we first learn a distributed representation of each review by a one-dimensional convolutional neural network. Then, taking these representations as pretrained vectors, we use a recurrent neural network with gated recurrent units to learn distributed representations of users and products. Finally, we feed the user, product and review representations into a machine learning classifier for sentiment classification. Our approach has been evaluated on three large-scale review datasets from the IMDB and Yelp. Experimental results show that: (1) sequence modeling for the purposes of distributed user and product representation learning can improve the performance of document-level sentiment classification; (2) the proposed approach achieves state-of-The-Art results on these benchmark datasets.

Research on the influence of behavioral forces that motivate trader behavior and sentiment: A prospect theory exegesis

This study focuses on empirical investigations and seeks implications by utilizing three different methodologies to test various aspects of trader behavior. The first methodology utilizes Prospect Theory to determine trader behavior during periods of extreme wealth contracting periods. Secondly, a threshold model to examine the sentiment variable is formulated and thirdly a study is made of the contagion effect and trader behavior. ^ The connection between consumers' sense of financial well-being or sentiment and stock market performance has been studied at length. However, without data on actual versus experimental performance, implications based on this relationship are meaningless. The empirical agenda included examining a proprietary file of daily trader activities over a five-year period. Overall, during periods of extreme wealth altering conditions, traders "satisfice" rather than choose the "best" alternative. A trader's degree of loss aversion depends on his/her prior investment performance. A model that explains the behavior of traders during periods of turmoil is developed. Prospect Theory and the data file influenced the design of the model. ^ Additional research included testing a model that permitted the data to signal the crisis through a threshold model. The third empirical study sought to investigate the existence of contagion caused by declining global wealth effects using evidence from the mining industry in Canada. Contagion, where a financial crisis begins locally and subsequently spreads elsewhere, has been studied in terms of correlations among similar regions. The results provide support for Prospect Theory in two out of the three empirical studies. ^ The dissertation emphasizes the need for specifying precise, testable models of investors' expectations by providing tools to identify paradoxical behavior patterns. True enhancements in this field must include empirical research utilizing reliable data sources to mitigate data mining problems and allow researchers to distinguish between expectations-based and risk-based explanations of behavior. Through this type of research, it may be possible to systematically exploit "irrational" market behavior. ^

Essays on the impact of stakeholders' sentiment on the financial decision making process

The most important factor that affects the decision making process in finance is the risk which is usually measured by variance (total risk) or systematic risk (beta). Since investors’ sentiment (whether she is an optimist or pessimist) plays a very important role in the choice of beta measure, any decision made for the same asset within the same time horizon will be different for different individuals. In other words, there will neither be homogeneity of beliefs nor the rational expectation prevalent in the market due to behavioral traits. This dissertation consists of three essays. In the first essay, “ Investor Sentiment and Intrinsic Stock Prices”, a new technical trading strategy was developed using a firm specific individual sentiment measure. This behavioral based trading strategy forecasts a range within which a stock price moves in a particular period and can be used for stock trading. Results indicate that sample firms trade within a range and give signals as to when to buy or sell. In the second essay, “Managerial Sentiment and the Value of the Firm”, examined the effect of managerial sentiment on the project selection process using net present value criterion and also effect of managerial sentiment on the value of firm. Final analysis reported that high sentiment and low sentiment managers obtain different values for the same firm before and after the acceptance of a project. Changes in the cost of capital, weighted cost of average capital were found due to managerial sentiment. In the last essay, “Investor Sentiment and Optimal Portfolio Selection”, analyzed how the investor sentiment affects the nature and composition of the optimal portfolio as well as the portfolio performance. Results suggested that the choice of the investor sentiment completely changes the portfolio composition, i.e., the high sentiment investor will have a completely different choice of assets in the portfolio in comparison with the low sentiment investor. The results indicated the practical application of behavioral model based technical indicator for stock trading. Additional insights developed include the valuation of firms with a behavioral component and the importance of distinguishing portfolio performance based on sentiment factors.

Essays on the Impact of Stakeholders' Sentiment on the Financial Decision Making Process

The most important factor that affects the decision making process in finance is the risk which is usually measured by variance (total risk) or systematic risk (beta). Since investors' sentiment (whether she is an optimist or pessimist) plays a very important role in the choice of beta measure, any decision made for the same asset within the same time horizon will be different for different individuals. In other words, there will neither be homogeneity of beliefs nor the rational expectation prevalent in the market due to behavioral traits. This dissertation consists of three essays. In the first essay, Investor Sentiment and Intrinsic Stock Prices, a new technical trading strategy is developed using a firm specific individual sentiment measure. This behavioral based trading strategy forecasts a range within which a stock price moves in a particular period and can be used for stock trading. Results show that sample firms trade within a range and show signals as to when to buy or sell. The second essay, Managerial Sentiment and the Value of the Firm, examines the effect of managerial sentiment on the project selection process using net present value criterion and also effect of managerial sentiment on the value of firm. Findings show that high sentiment and low sentiment managers obtain different values for the same firm before and after the acceptance of a project. The last essay, Investor Sentiment and Optimal Portfolio Selection, analyzes how the investor sentiment affects the nature and composition of the optimal portfolio as well as the performance measures. Results suggest that the choice of the investor sentiment completely changes the portfolio composition, i.e., the high sentiment investor will have a completely different choice of assets in the portfolio in comparison with the low sentiment investor. The results indicate the practical application of behavioral model based technical indicators for stock trading. Additional insights developed include the valuation of firms with a behavioral component and the importance of distinguishing portfolio performance based on sentiment factors.

Research on the influence of behavioral forces that motivate trader behavior and sentiment- a prospect theory exegesis

This study focuses on empirical investigations and seeks implications by utilizing three different methodologies to test various aspects of trader behavior. The first methodology utilizes Prospect Theory to determine trader behavior during periods of extreme wealth contracting periods. Secondly, a threshold model to examine the sentiment variable is formulated and thirdly a study is made of the contagion effect and trader behavior. The connection between consumers' sense of financial well-being or sentiment and stock market performance has been studied at length. However, without data on actual versus experimental performance, implications based on this relationship are meaningless. The empirical agenda included examining a proprietary file of daily trader activities over a five-year period. Overall, during periods of extreme wealth altering conditions, traders "satisfice" rather than choose the "best" alternative. A trader's degree of loss aversion depends on his/her prior investment performance. A model that explains the behavior of traders during periods of turmoil is developed. Prospect Theory and the data file influenced the design of the model. Additional research included testing a model that permitted the data to signal the crisis through a threshold model. The third empirical study sought to investigate the existence of contagion caused by declining global wealth effects using evidence from the mining industry in Canada. Contagion, where a financial crisis begins locally and subsequently spreads elsewhere, has been studied in terms of correlations among similar regions. The results provide support for Prospect Theory in two out of the three empirical studies. The dissertation emphasizes the need for specifying precise, testable models of investors' expectations by providing tools to identify paradoxical behavior patterns. True enhancements in this field must include empirical research utilizing reliable data sources to mitigate data mining problems and allow researchers to distinguish between expectations-based and risk-based explanations of behavior. Through this type of research, it may be possible to systematically exploit "irrational" market behavior.

Text analytics of social media: Sentiment analysis, event detection and summarization

In the last decade, large numbers of social media services have emerged and been widely used in people's daily life as important information sharing and acquisition tools. With a substantial amount of user-contributed text data on social media, it becomes a necessity to develop methods and tools for text analysis for this emerging data, in order to better utilize it to deliver meaningful information to users. ^ Previous work on text analytics in last several decades is mainly focused on traditional types of text like emails, news and academic literatures, and several critical issues to text data on social media have not been well explored: 1) how to detect sentiment from text on social media; 2) how to make use of social media's real-time nature; 3) how to address information overload for flexible information needs. ^ In this dissertation, we focus on these three problems. First, to detect sentiment of text on social media, we propose a non-negative matrix tri-factorization (tri-NMF) based dual active supervision method to minimize human labeling efforts for the new type of data. Second, to make use of social media's real-time nature, we propose approaches to detect events from text streams on social media. Third, to address information overload for flexible information needs, we propose two summarization framework, dominating set based summarization framework and learning-to-rank based summarization framework. The dominating set based summarization framework can be applied for different types of summarization problems, while the learning-to-rank based summarization framework helps utilize the existing training data to guild the new summarization tasks. In addition, we integrate these techneques in an application study of event summarization for sports games as an example of how to better utilize social media data. ^

Text Analytics of Social Media: Sentiment Analysis, Event Detection and Summarization

In the last decade, large numbers of social media services have emerged and been widely used in people's daily life as important information sharing and acquisition tools. With a substantial amount of user-contributed text data on social media, it becomes a necessity to develop methods and tools for text analysis for this emerging data, in order to better utilize it to deliver meaningful information to users. Previous work on text analytics in last several decades is mainly focused on traditional types of text like emails, news and academic literatures, and several critical issues to text data on social media have not been well explored: 1) how to detect sentiment from text on social media; 2) how to make use of social media's real-time nature; 3) how to address information overload for flexible information needs. In this dissertation, we focus on these three problems. First, to detect sentiment of text on social media, we propose a non-negative matrix tri-factorization (tri-NMF) based dual active supervision method to minimize human labeling efforts for the new type of data. Second, to make use of social media's real-time nature, we propose approaches to detect events from text streams on social media. Third, to address information overload for flexible information needs, we propose two summarization framework, dominating set based summarization framework and learning-to-rank based summarization framework. The dominating set based summarization framework can be applied for different types of summarization problems, while the learning-to-rank based summarization framework helps utilize the existing training data to guild the new summarization tasks. In addition, we integrate these techneques in an application study of event summarization for sports games as an example of how to better utilize social media data.

A Machine Learning Analysis of Twitter Sentiment to the Sandy Hook Shootings

Gun related violence is a complex issue and accounts for a large proportion of violent incidents. In the research reported in this paper, we set out to investigate the pro-gun and anti-gun sentiments expressed on a social media platform, namely Twitter, in response to the 2012 Sandy Hook Elementary School shooting in Connecticut, USA. Machine learning techniques are applied to classify a data corpus of over 700,000 tweets. The sentiments are captured using a public sentiment score that considers the volume of tweets as well as population. A web-based interactive tool is developed to visualise the sentiments and is available at this http://www.gunsontwitter.com. The key findings from this research are: (i) There are elevated rates of both pro-gun and anti-gun sentiments on the day of the shooting. Surprisingly, the pro-gun sentiment remains high for a number of days following the event but the anti-gun sentiment quickly falls to pre-event levels. (ii) There is a different public response from each state, with the highest pro-gun sentiment not coming from those with highest gun ownership levels but rather from California, Texas and New York.