Kevan Buckley

Sentiment strength detection for the social web

Sentiment analysis is concerned with the automatic extraction of sentiment-related information from text. Although most sentiment analysis addresses commercial tasks, such as extracting opinions from product reviews, there is increasing interest in the affective dimension of the social web, and Twitter in particular. Most sentiment analysis algorithms are not ideally suited to this task because they exploit indirect indicators of sentiment that can reflect genre or topic instead. Hence, such algorithms used to process social web texts can identify spurious sentiment patterns caused by topics rather than affective phenomena. This article assesses an improved version of the algorithm SentiStrength for sentiment strength detection across the social web that primarily uses direct indications of sentiment. The results from six diverse social web data sets (MySpace, Twitter, YouTube, Digg, RunnersWorld, BBCForums) indicate that SentiStrength 2 is successful in the sense of performing better than a baseline approach for all data sets in both supervised and unsupervised cases. SentiStrength is not always better than machine-learning approaches that exploit indirect indicators of sentiment, however, and is particularly weaker for positive sentiment in news-related discussions. Overall, the results suggest that, even unsupervised, SentiStrength is robust enough to be applied to a wide variety of different social web contexts.

Sentiment in short strength detection informal text

A huge number of informal messages are posted every day in social network sites, blogs, and discussion forums. Emotions seem to be frequently important in these texts for expressing friendship, showing social support or as part of online arguments. Algorithms to identify sentiment and sentiment strength are needed to help understand the role of emotion in this informal communication and also to identify inappropriate or anomalous affective utterances, potentially associated with threatening behavior to the self or others. Nevertheless, existing sentiment detection algorithms tend to be commercially oriented, designed to identify opinions about products rather than user behaviors. This article partly fills this gap with a new algorithm, SentiStrength, to extract sentiment strength from informal English text, using new methods to exploit the de facto grammars and spelling styles of cyberspace. Applied to MySpace comments and with a lookup table of term sentiment strengths optimized by machine learning, SentiStrength is able to predict positive emotion with 60.6p accuracy and negative emotion with 72.8p accuracy, both based upon strength scales of 1–5. The former, but not the latter, is better than baseline and a wide range of general machine learning approaches.

Negative emotions boost user activity at BBC forum

We present an empirical study of user activity in online BBC discussion forums, measured by the number of posts written by individual debaters and the average sentiment of these posts. Nearly 2.5 million posts from over 18 thousand users were investigated. Scale-free distributions were observed for activity in individual discussion threads as well as for overall activity. The number of unique users in a thread normalized by the thread length decays with thread length, suggesting that thread life is sustained by mutual discussions rather than by independent comments. Automatic sentiment analysis shows that most posts contain negative emotions and the most active users in individual threads express predominantly negative sentiments. It follows that the average emotion of longer threads is more negative and that threads can be sustained by negative comments. An agent-based computer simulation model has been used to reproduce several essential characteristics of the analyzed system. The model stresses the role of discussions between users, especially emotionally laden quarrels between supporters of opposite opinions, and represents many observed statistics of the forum.

Topic-based sentiment analysis for the social web: The role of mood and issue-related words

General sentiment analysis for the social web has become increasingly useful for shedding light on the role of emotion in online communication and offline events in both academic research and data journalism. Nevertheless, existing general‐purpose social web sentiment analysis algorithms may not be optimal for texts focussed around specific topics. This article introduces 2 new methods, mood setting and lexicon extension, to improve the accuracy of topic‐specific lexical sentiment strength detection for the social web. Mood setting allows the topic mood to determine the default polarity for ostensibly neutral expressive text. Topic‐specific lexicon extension involves adding topic‐specific words to the default general sentiment lexicon. Experiments with 8 data sets show that both methods can improve sentiment analysis performance in corpora and are recommended when the topic focus is tightest.

Damping sentiment analysis in online communication: discussions, monologs and dialogs

Sentiment analysis programs are now sometimes used to detect patterns of sentiment use over time in online communication and to help automated systems interact better with users. Nevertheless, it seems that no previous published study has assessed whether the position of individual texts within on-going communication can be exploited to help detect their sentiments. This article assesses apparent sentiment anomalies in on-going communication --- texts assigned significantly different sentiment strength to the average of previous texts --- to see whether their classification can be improved. The results suggest that a damping procedure to reduce sudden large changes in sentiment can improve classification accuracy but that the optimal procedure will depend on the type of texts processed.

Lognormal distributions of user post lengths in Internet discussions - a consequence of the Weber-Fechner law?

The paper presents an analysis of the length of comments posted in Internet discussion fora, based on a collection of large datasets from several sources. We found that despite differences in the forum language, the discussed topics and user emotions, the comment length distributions are very regular and described by the lognormal form with a very high precision. We discuss possible origins of this regularity and the existence of a universal mechanism deciding the length of the user posts. We suggest that the observed lognormal dependence may be due to an entropy maximizing combination of two psychological factors which are perceived on a non-linear, logarithmic scale in accordance with the Weber-Fechner law, namely the time spent on post related considerations and the comment length itself. This hypothesis is supported by an experimental check of text length recognition capacity, confirming proportionality of the ‘just noticeable differences’ for text lengths - the basis of the Weber-Fechner law.

Automated flexion crease identification using internal image seams

Palmar flexion crease recognition is a palmprint identification method for verifying biometric identity. This paper proposes a method of automated flexion crease recognition that can be used to identify palmar flexion creases in online palmprint images. A modified image seams algorithm is used to extract the flexion creases, and a matching algorithm, based on kd-tree nearest neighbour searching, is used to calculate the similarity between them. Experimental results show that in 1000 images from 100 palms, when compared to manually identified flexion creases, a genuine acceptance rate of 100% can be achieved, with a false acceptance rate of 0.0045%.

On desideratum for B2C E-commerce reputation systems

This paper reviews existing approaches to reputation systems, their constraints as well as available solutions. Furthermore, it presents and evaluates a novel and comprehensive reputation model devoted to the distributed reputation system for Business-to-Consumer (B2C) E-commerce applications that overcomes the discussed drawbacks. The algorithm offers a comprehensive approach as it considers a number of issues that have a bearing on trust and reputation such as age of ratings, transaction value, credibility of referees, number of malicious incidents, collusion and unfair ratings. Moreover, it also extends the existing frameworks based on information about past behaviour, with other aspects affecting online trading decisions which relate to the characteristic of the providers, such as existence of trustmark seals, payment intermediaries, privacy statements, security/privacy strategies, purchase protection/insurance, alternative dispute resolutions as well as the existence of first party information.

Latent fingermark pore area reproducibility.

The study of the reproducibility of friction ridge pore detail in fingermarks is a measure of their usefulness in personal identification. Pore area in latent prints developed using cyanoacrylate and ninhydrin were examined and measured by photomicrography using appropriate software tools. The data were analysed statistically and the results showed that pore area is not reproducible in developed latent prints, using either of the development techniques. The results add further support to the lack of reliability of pore area in personal identification.

Subjectivity annotation of the microblog 2011 realtime adhoc relevance judgments

In this work, we extend the Microblog dataset with subjectivity annotations. Our aim is twofold; first, we want to provide a high-quality, multiply-annotated gold standard of subjectivity annotations for the relevance assessments of the real-time adhoc task. Second, we randomly sample the rest of the dataset and annotate it for subjectivity once, in order to create a complementary annotated dataset that is at least an order of magnitude larger than the gold standard. As a result we have 2,389 tweets that have been annotated by multiple humans and 75,761 tweets that have been annotated by one annotator. We discuss issues like inter-annotator agreement, the time that it took annotators to classify tweets in correlation to their subjective content and lastly, the distribution of subjective tweets in relation to topic categorization. The annotated datasets and all relevant anonymised information are freely available for research purposes.