Social sentiment sensor: a visualization system for topic detection and topic sentiment analysis on microblog

As a new form of social media, microblogging provides platform sharing, wherein users can share their feelings and ideas on certain topics. Bursty topics from microblogs are the results of the emerging issues that instantly attract more followers and more attention online, which provide a unique opportunity to gauge the relation between expressed public sentiment and hot topics. This paper presents a Social Sentiment Sensor (SSS) system on Sina Weibo to detect daily hot topics and analyze the sentiment distributions toward these topics. SSS includes two main techniques, namely, hot topic detection and topic-oriented sentiment analysis. Hot topic detection aims to detect the most popular topics online based on the following steps, topic detection, topic clustering, and topic popularity ranking. We extracted topics from the hashtags using a hashtag filtering model because they can cover almost all the topics. Then, we cluster the topics that describe the same issue, and rank the topic clusters via their popularity to exploit the final hot topics. Topic-oriented sentiment analysis aims to analyze public opinions toward the hot topics. After retrieving the topic-related messages, we recognize sentiment for each message using a state-of-the-art SVM (Support Vector Machine) sentiment classifier. Then, we summarize the sentiments for the hot topic to achieve topic sentiment distribution. Based on the above framework and algorithms, SSS produces a real-time visualization system to monitor social sentiments, which is offering the public a new and timely perspective on the dynamics of the social topics.     

Phrase-based hashtag recommendation for microblog posts

In microblogs, authors use hashtags to mark keywords or topics. These manually labeled tags can be used to benefit various live social media applications (e.g., microblog retrieval, classification). However, because only a small portion of microblogs contain hashtags, recommending hashtags for use in microblogs are a worthwhile exercise. In addition, human inference often relies on the intrinsic grouping of words into phrases. However, existing work uses only unigrams to model corpora. In this work, we propose a novel phrase-based topical translation model to address this problem. We use the bag-of-phrases model to better capture the underlying topics of posted microblogs. We regard the phrases and hashtags in a microblog as two different languages that are talking about the same thing. Thus, the hashtag recommendation task can be viewed as a translation process from phrases to hashtags. To handle the topical information of microblogs, the proposed model regards translation probability as being topic specific. We test the methods on data collected from realworld microblogging services. The results demonstrate that the proposed method outperforms state-of-the-art methods that use the unigram model.     

科技文献中作者研究兴趣动态发现

针对挖掘大规模科技文献中作者、主题和时间及其关系的问题,考虑科技文献的内外部特征,提出了一个作者主题演化(AToT)模型。模型中文档表示为一定概率比例的主题混合体,每个主题对应一个词项上的多项分布和一个随时间变化的贝塔分布,主题词项分布不仅由文档中单词共现决定,同时受文档时间戳影响,每个作者也对应一个主题上的多项分布。主题词项分布与作者主题分布分别用来描述主题随时间变化的规律和作者研究兴趣的变化规律。采用吉布斯采样的方法,通过学习文档集可以获得模型的参数。在1700篇NIPS会议论文集上的实验结果显示,作者主题演化模型可以描述文档集中潜在的主题演化规律,动态发现作者研究兴趣的变化,可以预测与主题相关的作者,与作者主题模型相比计算困惑度更低。     

Culture and social media: the relationship between cultural values and hashtagging styles

ABSTRACT

Many social media users include #-signs before particular terms on social media – which is termed hashtagging. Recent research indicates that people tend to use the pound key for uncommon words, including ‘artistic’ words that are unlikely to serve functional purposes, and that cultural differences in hashtagging styles exist. The current study examines characteristics of hashtags and the impact of individual cultural values on hashtagging behaviour. Findings reveal four dimensions of hashtags, concluding that hashtags can be inspirational, structural, entertaining, and artistic. Second, findings show that hashtags are used to structure content equally independent of cultural values. However, inspirational hashtags are common among users with collectivistic, uncertainty avoidant, and masculine cultural values. Moreover, collectivistic and masculine values are also associated with artistic hashtags – whereas uncertainty avoidance is related to entertaining hashtags. In addition, findings show that cultural values associated with power distance relate to a higher hashtagging intensity.     

融合主题与语言模型的蒙古文信息检索方法研究*

摘 要: 为了从日益丰富的蒙古文信息中快速准确地检索用户需求的主题信息,提出了一种融合主题模型LDA与语言模型的方法。该方法首先对蒙古文文本建立一元和二元语言模型,得到文本的语言概率分布;然后基于LDA建立主题模型,利用吉普斯抽样方法计算模型的参数,挖掘得到文档隐含的主题概率分布;最后,计算出文档主题分布与语言分布的线性组合概率分布,以此分布来计算文档主题与查询关键词之间的相似度,返回与查询关键词主题最相关的文档。语言模型充分利用蒙古文语法特征,而主题模型LDA又具有良好的潜在语义挖掘及主题发现的泛化学习能力,从而结合两种方法更好的实现蒙古文文档的主题语义检索,提高检索准确性。实验结果表明,融合LDA模型与语言模型的方法相比单一模型体现主题语义方面取得了较好的效果。     

主题方面共享的领域主题层次模型

层次主题模型是构建主题层次的重要工具. 现有的层次主题模型大多通过在主题模型中引入nCRP构造方法, 为文档主题提供树形结构的先验分布, 但无法生成具有明确领域涵义的主题层次结构, 即领域主题层次. 同时, 领域主题不仅存在层次关系, 而且不同父主题下的子主题之间还存在子领域方面共享的关联关系, 在现有主题关系研究中没有合适的模型来生成这种领域主题层次. 为了从领域文本中自动、有效地挖掘出领域主题的层次关系和关联关系, 在4个方面进行创新研究. 首先, 通过主题共享机制改进nCRP构造方法, 提出nCRP+层次构造方法, 为主题模型中的主题提供具有分层主题方面共享的树形先验分布; 其次, 结合nCRP+和HDP模型构建重分层的Dirichlet过程, 提出rHDP (reallocated hierarchical Dirichlet processes)层次主题模型; 第三, 结合领域分类信息、词语语义和主题词的领域代表性, 定义领域知识, 包括基于投票机制的领域隶属度、词语与领域主题的语义相关度和层次化的主题-词语贡献度; 最后, 通过领域知识改进rHDP主题模型中领域主题和主题词的分配过程, 提出结合领域知识的层次主题模型rHDP_DK (rHDP with domain knowledge), 并改进采样过程. 实验结果表明, 基于nCRP+的层次主题模型在评价指标方面均优于基于nCRP的层次主题模型(hLDA, nHDP)和神经主题模型(TSNTM); 通过rHDP_DK模型生成的主题层次结构具有领域主题层次清晰、关联子主题的主题词领域差异明确的特点. 此外, 该模型将为领域主题层次提供一个通用的自动挖掘框架.     

时间加权的TF-LDA学术文献摘要主题分析

随着网络的发展,主题提取的应用越来越广泛,尤其是学术文献的主题提取。尽管学术文献摘要是短文本,但其具有高维性的特点导致文本主题模型难以处理,其时效性的特点致使主题挖掘时容易忽略时间因素,造成主题分布不均、不明确。针对此类问题,提出一种基于TTF-LDA(time+tf-idf+latent Dirichlet allocation)的学术文献摘要主题聚类模型。通过引入TF-IDF特征提取的方法,对摘要进行特征词的提取,能有效降低LDA模型的输入文本维度,融合学术文献的发表时间因素,建立时间窗口,限定学术文献主题分析的时间,并通过文献的发表时间增加特征词的时间权重,使用特征词的时间权重之和协同主题引导特征词词库作为LDA的影响因子。通过在爬虫爬取的数据集上进行实验,与标准的LDA和MVC-LDA相比,在选取相同的主题数的情况下,模型的混乱程度更低,主题与主题之间的区分度更高,更符合学术文献本身的特点。     

基于聚类语言模型的生物文献检索技术研究

近年来研究表明使用主题语言模型增强了信息检索的性能,但是仍然不能解决信息检索存在的一些难点问题,如数据稀疏问题,同义词问题,多义词问题,对文档中不可见项和可见项的平滑问题。这些问题在一些领域相关文献检索中显得尤其重要,比如大规模的生物文献检索。本文提出了一种新的基于聚类的主题语言模型方法进行生物文献检索,这主要包括两个方面工作,一是采用本体库中的概念表示文档,并在此基础上进行模糊聚类,把聚类的结果作为数据集中的主题,文档属于某个主题的概率由文档与聚类的模糊相似度决定。二是采用EM算法来估计主题产生项的概率。把上述方法集成到语言模型中就得到本文的语言模型。本文的语言模型能够准确描述项在不同主题中的分布概率,以及文档属于某个主题的概率,并且利用本体中概念部分地解决了同义词问题,而且项可以由不同的主题产生,这也能够部分解决词的多义问题。本文的方法在TREC 2004/05 Genomics Track数据集上进行了测试,与简单语言模型以及现有主题语言模型相比,检索性能得到一定的提高。     

Topic detection and tracking on heterogeneous information

Given the proliferation of social media and the abundance of news feeds, a substantial amount of real-time content is distributed through disparate sources, which makes it increasingly difficult to glean and distill useful information. Although combining heterogeneous sources for topic detection has gained attention from several research communities, most of them fail to consider the interaction among different sources and their intertwined temporal dynamics. To address this concern, we studied the dynamics of topics from heterogeneous sources by exploiting both their individual properties (including temporal features) and their inter-relationships. We first implemented a heterogeneous topic model that enables topic–topic correspondence between the sources by iteratively updating its topic–word distribution. To capture temporal dynamics, the topics are then correlated with a time-dependent function that can characterise its social response and popularity over time. We extensively evaluate the proposed approach and compare to the state-of-the-art techniques on heterogeneous collection. Experimental results demonstrate that our approach can significantly outperform the existing ones.     

融合PAM和主题偏好TextRank的历史沿革信息抽取

针对从自然标注大数据中抽取历史沿革主题信息的问题,提出了一种融合PAM主题模型与主题偏好TextRank的方法。该方法利用PAM主题模型获取历史沿革主题基于其它相关主题的分布,和不同主题基于词的分布;主题偏好TextRank算法则根据PAM所获得的主题和词的分布,在随机游走的过程中更加偏好于与历史沿革主题相关度大的结点,从而更有利于抽取历史沿革主题信息。因历史沿革主题特征复杂,与其它主题关联度大,词项本身是否表达历史沿革主题信息也并不明确,因此PAM即可以获取基于词空间的分布,又可以获取基于主题分布,对解决这类问题有很大的帮助。利用已获取的主题信息,主题偏好TextRank算法偏向于与历史沿革主题相关的结点进行随机游走,使得抽取结果趋向于历史沿革主题,从而提高了抽取的准确性。实验结果表明,该方法在抽取历史沿革主题信息上更有效。     

QuestionHolic: Hot topic discovery and trend analysis in community question answering systems

Community question answering (CQA) has recently become a popular social media where users can post questions on any topic of interest and get answers from enthusiasts. The variation of topics in questions and answers indicate the change of users’ interests over time. It can help users focus on the most popular products or events and track their changes by exploiting hot topics and analyzing the trend of a specific topic. In this paper, we present a hot topic detection and trend analysis system to capture hot topics in a CQA system and track their evolutions over time. Our system consists of hot term extraction, question clustering and trend analysis. Experimental results using datasets from Yahoo! Answers show that our system can discover meaningful hot topics. We also show that the evolution of topics over time can be accurately exploited by trend graphing.     

基于DMA与特征划分的多源文本主题模型

针对传统主题模型在挖掘多源文本数据集信息时存在主题发现效果不佳的问题,设计一种基于狄利克雷多项式分配（DMA）与特征划分的多源文本主题模型。以DMA模型为基础,放宽对预先输入的主题数量的限制,为每个数据源分配专有的主题分布参数,使用Gibbs采样算法估计每个数据源的主题数量。同时,对每个数据源分配专有的噪音词分布参数以及主题-词分布参数,采用特征划分方法区分每个数据源的特征词和噪音词,并学习每个数据源的用词特征,避免噪音词集对模型聚类的干扰。实验结果表明,与传统主题模型相比,该模型能够保留每个数据源特有的词特征,具有更好的主题发现效果及鲁棒性。     

主题模型LDA的多文档自动文摘

近年来使用概率主题模型表示多文档文摘问题受到研究者的关注.LDA (latent dirichlet allocation)是主题模型中具有代表性的概率生成性模型之一.提出了一种基于LDA的文摘方法,该方法以混乱度确定LDA模型的主题数目,以Gibbs抽样获得模型中句子的主题概率分布和主题的词汇概率分布,以句子中主题权重的加和确定各个主题的重要程度,并根据LDA模型中主题的概率分布和句子的概率分布提出了2种不同的句子权重计算模型.实验中使用ROUGE评测标准,与代表最新水平的SumBasic方法和其他2种基于LDA的多文档自动文摘方法在通用型多文档摘要测试集DUC2002上的评测数据进行比较,结果表明提出的基于LDA的多文档自动文摘方法在ROUGE的各个评测标准上均优于SumBasic方法,与其他基于LDA模型的文摘相比也具有优势.     

基于语义增强的短文本主题模型

传统主题模型方法很大程度上依赖于词共现模式生成文档主题, 短文本由于缺乏足够的上下文信息导致的数据稀疏性成为传统主题模型在短文本上取得良好效果的瓶颈. 基于此, 本文提出一种基于语义增强的短文本主题模型, 算法将DMM (Dirichlet Multinomial Mixture)与词嵌入模型相结合, 通过训练全局词嵌入与局部词嵌入获得词的向量表示, 融合全局词嵌入向量与局部词嵌入向量计算词向量间的语义相关度, 并通过主题相关词权重进行词的语义增强计算. 实验表明, 本文提出的模型在主题一致性表示上更准确, 且提升了模型在短文本上的分类正确率.     

基于信息内容和拓扑关系的社会媒体用户兴趣分类

随着社会的发展,信息已经成为社会发展越来越重要的部分,人类的信息传播活动越来越明显地展示出分众特征,对用户的分类成为人类信息活动的一个重要研究课题.从这一目标出发,分别基于信息内容、拓扑关系和两者综合的方法,按兴趣主题对社会媒体用户进行分类.对于基于信息内容的用户分类,采用LDA主题模型从用户所发布的内容中提取其主题分布,基于这一分布,采用支持向量机、决策树、贝叶斯等多种模型按兴趣主题对用户进行分类.对于基于拓扑关系的分类,依据相同兴趣主题的用户倾向于拥有共同的粉丝这一发现,构建分类模型来按兴趣主题对用户进行分类.然后提出综合信息内容和拓扑关系的分类方法来对用户进行分类.最后基于大规模Twitter数据的实验发现,采用综合方法对用户进行的兴趣分类性能明显高于采用单一信息内容或粉丝拓扑方法的性能.     

基于上下文的话题演化和话题关系抽取研究

自动挖掘大规模语料中的语义信息以及演化关系近年来已受到广大专家学者的关注。话题被认为是文档集合中的潜在语义信息,话题演化用于研究话题内容随时间的变化。该文提出了一种基于上下文的话题演化和话题关系抽取方法。分析发现,一个话题常和某些其他话题共现在多篇文档中,话题间的这种共现信息被称为话题的上下文。上下文信息可以用于计算同时间段话题间的语义关系以及识别不同时间段中具有相同语义的话题。该文对2008年～2012年两会报告以及2007年～2011年NIPS科技文献进行实验,通过人工分析,利用话题的上下文信息,不但可以提高话题演化的正确率,而且还能挖掘话题之间的语义关系,在话题演化的基础上,显示话题关系的演化。     

Event detection over twitter social media streams

In recent years, microblogs have become an important source for reporting real-world events. A real-world occurrence reported in microblogs is also called a social event. Social events may hold critical materials that describe the situations during a crisis. In real applications, such as crisis management and decision making, monitoring the critical events over social streams will enable watch officers to analyze a whole situation that is a composite event, and make the right decision based on the detailed contexts such as what is happening, where an event is happening, and who are involved. Although there has been significant research effort on detecting a target event in social networks based on a single source, in crisis, we often want to analyze the composite events contributed by different social users. So far, the problem of integrating ambiguous views from different users is not well investigated. To address this issue, we propose a novel framework to detect composite social events over streams, which fully exploits the information of social data over multiple dimensions. Specifically, we first propose a graphical model called location-time constrained topic (LTT) to capture the content, time, and location of social messages. Using LTT, a social message is represented as a probability distribution over a set of topics by inference, and the similarity between two messages is measured by the distance between their distributions. Then, the events are identified by conducting efficient similarity joins over social media streams. To accelerate the similarity join, we also propose a variable dimensional extendible hash over social streams. We have conducted extensive experiments to prove the high effectiveness and efficiency of the proposed approach.     

Latent topic model for audio retrieval

Latent topic model such as Latent Dirichlet Allocation (LDA) has been designed for text processing and has also demonstrated success in the task of audio related processing. The main idea behind LDA assumes that the words of each document arise from a mixture of topics, each of which is a multinomial distribution over the vocabulary. When applying the original LDA to process continuous data, the word-like unit need be first generated by vector quantization (VQ). This data discretization usually results in information loss. To overcome this shortage, this paper introduces a new topic model named Gaussian-LDA for audio retrieval. In the proposed model, we consider continuous emission probability, Gaussian instead of multinomial distribution. This new topic model skips the vector quantization and directly models each topic as a Gaussian distribution over audio features. It avoids discretization by this way and integrates the procedure of clustering. The experiments of audio retrieval demonstrate that Gaussian-LDA achieves better performance than other compared methods.