Stance detection and mobile app recommendation discourse on tweets

Web‐based social networking such as microblogging administrations and long‐range informal communication locales are changing the way in which individuals collaborate on the web and search for data and opinions. An essential parameter of online networking discourse is searchability. A key semiotic asset supporting this capacity is the hashtag, a type of social label that enables microbloggers to insert metadata in online networking posts. In this paper, an attempt is made to analyze stance detection and app recommendation discourse on tweets in view of hashtag techniques, which is in the territory of etymology, and to spotlight the structure of dialect at the provision level. With a revival of enthusiasm for topics identified by modeling language at the discourse level, a graphical model of conversational structure (ie, the structural topic model) has been constructed by means of utilizing three methods: displaying words connected with topics or documents highly connected with topics, calculating topic correlations, and assessing associations between metadata and topical content, its capture of latent topics, and topical structures inside documents on a benchmark dataset (ie, SemEval 2016) has been scrutinized for stance detection, and data have been crawled from Twitter, using the hashtag #App for app recommendations.     

流行病模型在微博转发预测中的应用

Microblogs currently play an important role in social communication. Hot topics currently being tweeted can quickly become popular within a very short time as a result of re-tweeting. Gaining an understanding of the retweeting behavior is desirable for a number of tasks such as topic detection, personalized message recommendation, and fake information monitoring and prevention. Inter-estingly, the propagation of tweets bears some similarity to the spread of infectious diseases. We present a method to model the tweets’ spread behavior in microblogs based on the classic Susceptible-Infectious-Susceptible (SIS) epidemic model that was developed in the medical field for the spread of infectious dis-eases. On the basis of this model, future re-tweeting trends can be predicted. Our experi-ments on data obtained from the Chinese micro-blog?ging website Sina Weibo show that the proposed model has lower predictive error compared to the four commonly used predic-tion methods.     

Twitter推文与情感词典SentiWordNet匹配算法研究

在Twitter情感分类研究中,经常会采用将推文中的单词匹配情感词典中的同义词条查找相应情感值的方法. 但推文书写比较随意,包含许多俚语、缩写和特殊符号,导致许多词汇与情感词典中的词条无法匹配,匹配率不高直接影响推文的情感分类性能. 针对Twitter的语言特征,提出了一套Twitter推文与情感词典SentiWordNet的匹配算法. 该算法首先通过对推文内容进行数据清洗、替代处理、词性标注和词形还原等预处理,增加了命名实体识别、对hashtags内容的断词处理、基于Word Clusters的否定句处理和词组匹配等方法. 实验结果表明,采用此方法的匹配率可达90%以上.     

基于相似消息的流行度预测方法

社交网络中消息的流行度预测问题在很多应用领域都有着重要意义。传统的流行度预测方法包括基于特征的方法和基于点过程的方法。基于点过程的方法无法利用历史消息的信息,而基于特征的方法则使用一个统一的模型来对所有的消息进行预测,没有考虑消息的特异性。因此,该文提出了一种基于相似消息的流行度预测方法。对于待预测微博,我们从历史消息选取出与之最相似的前K条消息来进行预测。在计算消息相似度时,我们借助了文档建模领域的LDA模型来学习消息的表示。在数据集上的实验结果表明,该方法可以有效发现在传播模式上与待预测消息相似的历史消息,并在流行度预测任务上取得了比对比模型更好的预测效果。     

一种相关话题微博信息的筛选规则学习算法

微博作为一种新型的社会媒体,以其信息的高实时性、话题动态关注、传播速度快的特点,逐渐被人们所接受和使用。筛选出相关话题的微博信息,帮助用户关注话题的动态发展,成为迫切需要解决的问题。由于微博信息篇幅极短、包含的信息和特征少等特点,为相关话题微博信息的筛选带来了新的挑战,而传统的文本分类技术已不再适用。该文提出了基于信息熵的筛选规则学习算法,利用学习得到的规则对微博信息进行有效的筛选。算法利用信息熵来评价规则的好坏,同时基于模拟退火的随机策略使算法中的规则选择避免了过于贪心。分别通过来自新浪微博的约九万条标注数据和TREC2011中约三千条特定话题的标注数据进行实验,该文算法相比于CPAR和SVM算法,学习得到的规则在筛选时取得了较高的F值。     

基于推文与属性的社交网络用户重识别方法

大数据隐私安全正成为各界关注的热点. 攻击者通过识别用户不同网站的账户,可以构建用户的完整画像,对用户隐私形成威胁. 模拟评估攻击者的重识别能力是进行用户隐私保护的前提. 因此,本文提出一种高相似同天同行为算法. 该算法通过检测账户在不同网站是否存在多次同天发表相近或相同内容的行为,判断账户是否属于同一用户,并通过为用户属性构建一种权重计算模型,进一步提高用户重识别的准确率. 经过对两个国内主流社交网站的一万多用户进行实验,本文算法表现出良好的效果. 实验表明,即使不考虑用户社交关系,用户的推文与属性依然提供了足够的信息使攻击者将用户不同网站的账户相关联,从而导致更多的隐私被泄露.     

Mining government tweets to identify and predict citizens engagement

The rise of social media offered new channels of communication between a government and its citizens. The social media channels are interactive, inclusive, low-cost, and unconstrained by time or place. This two-way communication between governments and citizens is referred to as electronic citizen participation, or e-participation. E-participation in the age of technology is considered as a mean for citizens to express their opinions and as a new input to be integrated by policy makers to take decisions. Governments and policy makers always aim to increase such participation not only to utilize public expertise and experience, but also to increase the transparency, trust, and acceptability of government decisions. In this research we investigate how governments can increase citizens e-participation on social media. We collected 55,809 tweets over a period of one year from Twitter accounts of a progressive government in the Arab world. This was followed by statistical analysis of posts characteristics (Type, Day, Time) and their impact on citizens' engagement. Then, we evaluated how well can different machine learning techniques predict user engagement. Results of the statistical analysis confirmed that post type (video, image, link, and status) impacted citizens' engagement, with videos and images having the highest positive impact on engagement. Furthermore, posting government tweets on weekdays obtained higher citizens’ engagement than weekends. Conversely, time of post had a weak effect on engagement. The results from the machine learning experiments show that two techniques (Random Forest and Adaboost) produced more accurate predictions, particularly when tweet textual contents were also used in the prediction. These results can help governments increase the engagement of their citizens.     

Using Hashtags to Capture Fine Emotion Categories from Tweets

Detecting emotions in microblogs and social media posts has applications for industry, health, and security. Statistical, supervised automatic methods for emotion detection rely on text that is labeled for emotions, but such data are rare and available for only a handful of basic emotions. In this article, we show that emotion‐word hashtags are good manual labels of emotions in tweets. We also propose a method to generate a large lexicon of word–emotion associations from this emotion‐labeled tweet corpus. This is the first lexicon with real‐valued word–emotion association scores. We begin with experiments for six basic emotions and show that the hashtag annotations are consistent and match with the annotations of trained judges. We also show how the extracted tweet corpus and word–emotion associations can be used to improve emotion classification accuracy in a different nontweet domain. Eminent psychologist Robert Plutchik had proposed that emotions have a relationship with personality traits. However, empirical experiments to establish this relationship have been stymied by the lack of comprehensive emotion resources. Because personality may be associated with any of the hundreds of emotions and because our hashtag approach scales easily to a large number of emotions, we extend our corpus by collecting tweets with hashtags pertaining to 585 fine emotions. Then, for the first time, we present experiments to show that fine emotion categories such as those of excitement, guilt, yearning, and admiration are useful in automatically detecting personality from text. Stream‐of‐consciousness essays and collections of Facebook posts marked with personality traits of the author are used as test sets.