Blog Corpus Research Articles

Appraisal PageRank for Arabic and English Blogs: Comparative Study

comparative study is made between Arabic and English Blogs. The difference between Arabic and English blogs starts with their Pagerank. The question to answer in this paper is Do the Arab blogger benefits from the search engine optimizer properties or not? In other words does the user cares for Pagerank? This question rises after the fame that the social media in general and blogs in specific have played during the political changes in Middle East region. Using a corpus of Arabic & English blogs, their content & link properties that effect the Pagerank score to answer the question (to prove or deny whether our assumption is true or not). The results obtained showed that there is a huge difference between Arabic & English blogs in terms of Page rank and content.

Proactive screening for depression through metaphorical and automatic text analysis

Proactive and automatic screening for depression is a challenge facing the public health system. This paper describes a system for addressing the above challenge. The system implementing the methodology--Pedesis--harvests the Web for metaphorical relations in which depression is embedded and extracts the relevant conceptual domains describing it. This information is used by human experts for the construction of a "depression lexicon". The lexicon is used to automatically evaluate the level of depression in texts or whether the text is dealing with depression as a topic. Tested on three corpora of questions addressed to a mental health site the system provides 9% improvement in prediction whether the question is dealing with depression. Tested on a corpus of Blogs, the system provides 84.2% correct classification rate (p<.001) whether a post includes signs of depression. By comparing the system's prediction to the judgment of human experts we achieved an average 78% precision and 76% recall. Depression can be automatically screened in texts and the mental health system may benefit from this screening ability.

Feature Analyses for Catchphrases Using Multiple Corpora

Catchphrases attract attention in a form of concise expressions; however, there are few researches which investigate them using the knowledge of natural language processing. First, for the analysis based on language engineering, we employed a catchphrase corpus containing 6,466 catchphrases, an encyclopedia corpus, a blog corpus and a dialog corpus. Moreover, we conducted an analysis on each theme for revealing the difference between themes on the catchphrase corpus. The result from the language engineering shows that forward parts of speeches tagging in the catchphrase corpus are similar to those of both the encyclopedia and blog corpora. In terms of backward parts of speeches tagging, a lot of catchphrases end with nouns, which differentiates the other corpora. In addition, for each theme in catchphrase corpus, we found that specific words such as “taste” and “skin” frequently appear.

A hierarchical approach to mood classification in blogs

AbstractIn this article, we explore the task of mood classification for blog postings. We propose a novel approach that uses the hierarchy of possible moods to achieve better results than a standard machine learning approach. We also show that using sentiment orientation features improves the performance of classification. We used the Livejournal blog corpus as a data set to train and evaluate our method. We present extensive error analysis and discuss the difficulty of the task.

Clustering of Blogs with Enhanced Semantics

Blogs are among the fastest growing space among the user generated content over the internet. It is fast becoming the tool for information dissemination, and communication. Blogs provide a platform for information sharing, discussions, and expression of reader‘s reactions to the blog post. Clustering of blogs greatly simplify blog searching and browsing by organizing them into similar groups. The Blogs are generally organized using tags. In this paper, we have studied the effect of considering other relevant neighborhood contexts and adding the extracted information to the original tag set carried by the blog. The added semantics is extracted by disambiguating all the synsets for the important terms/ or key phrases within the blog. This work reports the study of measuring similarity, on enhanced blog features and subsequently grouping of all blog articles based on the semantics of the tags they carry. We propose to include the semantics extracted from the title, body, and comments of a blog post to its original tagset in clustering blog documents and evaluate the hypothesis that adding extracted semantics from these blog constituents improves the cluster quality. For clustering k-means algorithm is used. The experimental results obtained confirm our hypothesis that adding the semantics improves better clusters. The approach first extracts the relevant features from the target blog corpus, title and comments. The other senses represented by the relevant keywords are discovered by using a general purpose semantics extractor. All the synsets of the relevant keywords are extracted from the WORDNET. The extracted keyword senses are then appended to the base tagsets. A semantic similarity measure is used for computing the semantic similarity among the documents. Clusters are obtained based on it. The two clusters output are compared. General Terms Clustering, blog mining, blog, text mining, semantic similarity

Stance-taking and public discussion in blogs

Blogs, which can be written and read by anyone with a computer and an internet connection, would seem to expand the possibilities for engagement in public sphere debates. Indeed, blogs are full of the kind of vocabulary that suggests intense discussion. However, a closer look at the way this vocabulary is used in context suggests that the main concern of writers is self-presentation, positioning themselves in a crowded forum, in what has been called stance-taking. When writers mark their stances, for instance by saying I think, they enact different ways of signalling a relation to others, marking disagreement, enacting surprise, and ironicising previous contributions. All these moves are ways of presenting one's own contribution as distinctive, showing one's entitlement to a position. In this paper, I use concordance tools to identify strings that are very frequent in a corpus of blogs, relative to a general corpus of written texts, focus on those relatively frequent words that mark stance and analyse these markers in context. I argue that the prominence of stance-taking indicates the priority of individual positioning over collective and deliberative discussion.

Efficient Cross-Domain Classification of Weblogs

Text classification is one of the core applications in data mining due to the huge amount of uncategorized textual data available. Training a text classifier results in a classification model that reflects the characteristics of the domain it was learned on. However, if no training data is available, labeled data from a related but different domain might be exploited to perform cross-domain classification. In our work, we aim to accurately classify unlabeled weblogs into commonly agreed upon newspaper categories using labeled data from the news domain. The labeled news and the unlabeled blog corpus are highly dynamic and hourly growing with a topic drift, so the classification needs to be efficient. Our approach is to apply a fast novel centroid-based text classification algorithm, the Class-Feature-Centroid Classifier (CFC), to perform efficient cross-domain classification. Experiments showed that this algorithm achieves a comparable accuracy than k-Nearest Neighbour (k-NN) and Support Vector Machines (SVM), yet at linear time cost for training and classification. We investigate the classifier performance and generalization ability using a special visualization of classifiers. The benefit of our approach is that the linear time complexity enables us to efficiently generate an accurate classifier, reflecting the topic drift, several times per day on a huge dataset.

Birds of a feather: How personality influences blog writing and reading

The rapid development of blogs as a social networking tool has greatly increased the number of people who are expressing themselves in text published online. As yet little is known about the social psychology of online interaction using recently developed Web 2.0 functionalities. To what extent do principles of social psychology carry over into the online domain and how can appropriate use of those principles assist in activities such as community building, e-commerce, marketing, and personalization of services? There would seem to be a large number of novel research questions that can be posed with respect to online interaction. One pertinent question concerns the extent to which compatibility of personality influences online interaction. We investigated this question in terms of hypotheses derived from the personality research literature, using a two-part experiment. In the first study, eight participants wrote blogs in two distinct genres (personal diaries and commentaries) and rated their own personalities. In the second study 12 different participants judged the personality of authors as implied in the blog texts created in the first study. Those participants also completed a personality questionnaire and rated their attraction to the blog text author. Readers of the blog corpus were able to consistently judge the personality of the writers based solely on the text that they wrote. Moreover, they followed a well-established social rule regarding interpersonal attraction in real-life interaction. Blog readers were significantly more attracted to blog writers with more similar personalities—offering support that in a blog environment, “birds of a feather flock together.” Emotion word use in the blog corpus correlated with writer’s personality, suggesting that online personality may be signalled by linguistic cues. The two most popular genres of blog writing, personal journal and commentary, differed in how accurately readers judged author personality traits. Based on the results obtained it is suggested that personality is an important determinant and that further research on how people communicate with blogs will be relevant to social network analysis and to marketing.

Competing narratives, gender and threaded identity in cyberspace

Using a blog corpus, the current study of weblog writing explores changing cultural norms surrounding technology-mediated formation of interpersonal relationships and aims to gain insight into important aspects of online female behaviour. More specifically, this article analyses the way in which British and Spanish female adolescents disclose personal information about their love-lives. In doing so, I argue that these bloggers create competing narratives which consist of either crediting or puncturing traditional love and relationship myths. The study of these competing narratives, in turn, reveals that these female adolescents live out different gendered discourses when they are entering, and when they are leaving, love relationships. Furthermore, this study examines the contemporary association between laddish attitudes in female adolescents and the involvement of verbal aggression in their personal weblogs.

TEENAGE GIRLS' PERSONAL WEBLOG WRITING

This paper presents an analysis of the construction of female gender identity in British and Spanish personal weblog writing. More precisely, this study offers an insight into the way white female adolescents conceptualize gender roles in twenty-first century society, especially in terms of self-expression. Inspection of this British and Spanish blog corpus suggests that British and Spanish teenage females' gain of self-esteem is still at an early stage as far as interpersonal relationships are concerned. The results highlight that both cultures seem to head towards a particular schema of androgyny which throws further light on gender role parameters. In addition, this culturally based redefinition reveals two different social phenomena resulting from the fight for gender equality in both societies.

Constructing a blog corpus for Japanese learners of English

Researchers have directed increased attention to the building and analysing of written learner corpora – databases of written language produced by language learners – to address issues such as the words that non-native learners of English use in their writing, and how their word use differs from that of native speakers. This paper offers an initial look at a new written learner corpus, currently under construction, which is composed of lower/intermediate-level learner blogs. Preliminary data from the corpus regarding high-frequency vocabulary use is compared to frequency lists from the British National Corpus in order to illustrate basic usage differences.

Stylometric Analysis of Bloggers’ Age and Gender

We report results of stylometric differences in blogging for gender and age group variation. The results are based on two mutually independent features. The first feature is the use of slang words which is a new concept proposed by us for Stylometric study of bloggers. Slang is a non-dictionary word that has evolved with time due to its frequent and popular usage. For the second feature, we have analysed the variation in average length of sentences across various age groups and gender. These two features are then augmented with previous study results reported in literature for stylometric analysis of age and gender. The combined feature list enhances the accuracy by a remarkable extent in predicting age and gender. These experiments were done on a 20,000 blog corpus. Experimental results show that these features work well in detection of bloggers’ demography. However, gender determination is more accurate than age group detection over a data spread across all ages but the accuracy of age prediction increases if we sample data with remarkable age difference.

What Are They Blogging About? Personality, Topic and Motivation in Blogs

Personal weblogs (blogs), provide individuals with the opportunity to write freely and express themselves online in the presence of others. In such situations, what do bloggers write about, and what are their motivations for blogging? Using a large blog corpus annotated with the LIWC text analysis program, we examine the content of blogs to provide insight into the role of personality in motivation for blogging. As predicted, we find that highly Neurotic authors use blogs to serve a cathartic or auto-therapeutic function, and reflect mainly upon themselves and negative emotions. Highly Extraverted blog authors, as expected, use blogs to document their lives at a high level, and uniquely interact directly with the reader. Additionally Extraverts use blogs to vent both positive and negative emotions. Bloggers who are high scorers on the Openness trait are more concerned with leisure activities, although they are more evaluative than intellectual, whereas high Conscientiousness bloggers tend to report daily life – and work – around them. As in other contexts, expressing positive rather than negative emotions is associated with high Agreeableness, but that trait is associated with self reference to a greater degree in blogs than elsewhere. In general, findings are consistent with other contexts indicating that bloggers tend to adapt to the possibilities of the medium, rather than try to present themselves differently.

Sentiment Analysis in Weblog Using Contextual Information: A Machine Learning Approach

This paper investigates sentiments in blog data using machine learning techniques such as support vector machines (SVM) and conditional random fields (CRF). Bloggers collaboratively contribute to the formation of blog corpora, which contain a rich set of icon-tagged text expressions. First, a collocation model on the expressions is used to construct a lexicon which provides keyword features for subsequent classification models. SVM and CRF classifiers use these keywords as features to assign a sentence sentiment bipolarity or arousal-valence space categories. Our methods determine a sentiment category by taking contexts into account. Experiments show that CRF classifiers outperform SVM classifiers and the performance benefits from increasing the training size of blog materials.

Semi-Supervised Learning to Classify Evaluative Expressions from Labeled and Unlabeled Texts

In this paper, we present a method to automatically acquire a large-scale vocabulary of evaluative expressions from a large corpus of blogs. For the purpose, this paper presents a semi-supervised method for classifying evaluative expressions, that is, tuples of subjects, their attributes, and evaluative words, that indicate either favorable or unfavorable opinions towards a specific subject. Due to its characteristics, our semi-supervised method can classify evaluative expressions in a corpus by their polarities, starting from a very small set of seed training examples and using contextual information in the sentences the expressions belong to. Our experimental results with real Weblog data as our corpus show that this bootstrapping approach can improve the accuracy of methods for classifying favorable and unfavorable opinions. We also show that a reasonable amount of evaluative expressions can be really acquired.