Title:
Leveraging Sentiment Analysis for Topic Detection

Keywords:
sentiment analysis, topic detection

Abstract:
... of such sentiment analysis techniques. We then detail a novel topic detection method using point-wise mutual information and term frequency distribution. We ...

Full Text:
Leveraging Sentiment Analysis for Topic DetectionLeveraging Sentiment Analysis for Topic Detection Keke Cai*, Scott Spangler!, Ying Chen!, Li Zhang* *IBM China ... of such sentiment analysis techniques. We then detail a novel topic detection me-thod using point-wise mutual information and term frequency distribution. We ... approach which com-bines a unique sentiment classification approach with a topic detection approach that discovers terms that are highly correlated to different ...
... sentiment analysis framework which combines sentiment classification approaches with sentiment topic detection approaches in one system. 2. We present our semantic-based sentiment ... 3. We define the sentiment topic concept and present sentiment topic detection approach (STD). 4. We verify the effectiveness of our approach ... components of the sentiment analy-sis, including sentiment classification and sentiment topic detection. . We show experimental results in Sec-tion 4. Finally, we ...
... describe the detailed algorithm in the following section.? 3. Sentiment topic detection algorithm 3.1. Sentiment based taxonomies Our overall sentiment analysis starts ...
... includes a sentiment classifica-tion scheme as well as a sentiment topic detection scheme. The sentiment classification component meas-ures the relative sentiment (on ... and then parti-tions the snippets into positive/negative/neutral catego-ries. The sentiment topic detection component detects the most significant topics hidden behind each senti-ment ...

Title:
A Graph Analytical Approach for Topic Detection

Keywords:
KeyGraph-based Topic Detection
Topic detection

Abstract:
<p>Topic detection with large and noisy data collections such as social media ...

References:
Al Sumait, L., Barbará, D., and Domeniconi, C. 2008. On-line lDA: Adaptive topic models for mining text streams with applications to topic detection and tracking. In Proceedings of the International Conference on Data Mining (ICDM). 3--12.
Cataldi, M., Di Caro, L., and Schifanella, C. 2010. Emerging topic detection on twitter based on temporal and social terms evaluation. In Proceedings of the 10th International Workshop on Multimedia Data Mining (MDMKDD). 4:1--4:10.
Mori, M., Miura, T., and Shioya, I. 2006. Topic detection and tracking for news web pages. In Proceedings of the IEEE/WIC/ACM International Conference on Web Intelligence (WI). 338--342.
Wartena, C. and Brussee, R. 2008. Topic detection by clustering keywords. In Proceedings of the IEEE Computer Society DEXA Workshops. 54--58.

Full Text:
... Management]:Database Applications?Data miningGeneral Terms: Algorithms, PerformanceAdditional Key Words and Phrases: Topic detection, , network analysis, community detection, KeyGraph-based Topic DetectionACM Reference Format:Sayyadi, ... H. and Raschid, L. 2013. A graph analytical approach for topic detection. . ACM Trans. InternetTechnol. 13, 2, Article 4 (December 2013), ... date: December 2013.????????4:2 H. Sayyadi and L. RaschidInitial methods for topic detection typically relied on clustering documents. Con-sider a core group of ...
... discuss these algorithms in detail in the next section.The best-of-breed topic detection algorithms, including kNN [Allan et al. 1998], GAC[Yang et al. ... collection.In this article we present KeyGraph, an efficient approach for topic detection thatwas inspired by the keyword cooccurrence graph and efficient graph ... contributions.?We demonstrate that the accuracy of the KeyGraph method for topic detection issimilar to that of the benchmark algorithms, including kNN [Allan ... Article 4, Publication date: December 2013.????????A Graph Analytical Approach for Topic Detection 4:3events, as well as the events that were identified in ... next section, we review related work,Section 3 presents the KeyGraph topic detection algorithm, in Section 4, we comparethe accuracy of KeyGraph to ...
The problem is also referencedas Topic Detection and Tracking (TDT). TDT usually falls into two categories, NewEvent ... word signals. They then chose correlated word signals as featuresfor topic detection. . Wang andMcCallum [2006], Wang et al. [2008], and Al ...
... not address topic evolution.There are also several recent studies on topic detection in social media. Asur et al.[2011] investigated the factors that ... Article 4, Publication date: December 2013.????????A Graph Analytical Approach for Topic Detection 4:5Algorithm 1 Graph Analytical approach for Topic Detection1: %Building the ...
... topic, andmerge these topics.Preliminary results of the KeyGraph approach to topic detection were presented inSayyadi et al. [2009]; research in this article ... Article 4, Publication date: December 2013.????????A Graph Analytical Approach for Topic Detection 4:7Table I. KeyGraph ParametersThe analysis to select parameter values is ...
... H. Sayyadi and L. Raschidresults show that the quality of topic detection with our aggregated KeyGraph is asgood as the best topic detection models in the literature.Assigning Topics to Documents. Recall that each ... Article 4, Publication date: December 2013.????????A Graph Analytical Approach for Topic Detection 4:9Table II. Details of Dataset 1 (TDT4) and Dataset 2 ... well-known dataset that has been widely used tocompare methods for topic detection [Li et al. 2005; Tantrum et al. 2002; Yang et ...
... as named entities so thatthey will be emphasized by our topic detection algorithm. We obtained the best resultsby doubling the term frequency ... doubling their term frequencies.We compare them with the following benchmark topic detection algorithms.(1) kNN. The k-Nearest Neighbor clustering algorithm is a popular ...
... Article 4, Publication date: December 2013.????????A Graph Analytical Approach for Topic Detection 4:11Fig. 3. The performance of the eight approaches (LDA-VEM, kNN, ...
... Article 4, Publication date: December 2013.????????A Graph Analytical Approach for Topic Detection 4:13keywords to describe each topic. We then asked subjects to ... Turk forthe human subject evaluation of the performance of the topics detected by the two al-gorithms. We evaluated the best 14 topics ...
... Article 4, Publication date: December 2013.????????A Graph Analytical Approach for Topic Detection 4:15Table IV. KeyGraph+?s Pseudo-Precision using Mechanical Turk for Spinn3r (32,000 ...
... Article 4, Publication date: December 2013.????????A Graph Analytical Approach for Topic Detection 4:17Table VI. KeyGraph+?s Pseudo-Inverse Recall using Mechanical Turk for Spinn3r(32,000 ...
... Article 4, Publication date: December 2013.????????A Graph Analytical Approach for Topic Detection 4:19Fig. 6. Sensitivity analysis for parameters node min df, edge ...
... we used Amazon?s Mechanical Turk for human subject evaluation ofthe topics detected by the two algorithms. On average, only 40% of the ... for detecting emerging topics, it will not be suitable forretrospective topic detection, , and more importantly, for topic tracking. We plan toincorporate ...
... Article 4, Publication date: December 2013.????????A Graph Analytical Approach for Topic Detection 4:21Table VIII. ContinuedTopic ID Topic Keywords Similar TopicKEYGRAPH10 Bush administration ... lDA: Adaptive topic models for mining textstreams with applications to topic detection and tracking. In Proceedings of the International Conferenceon Data Mining ... 73?82.Cataldi, M., Di Caro, L., and Schifanella, C. 2010. Emerging topic detection on twitter based on temporaland social terms evaluation. In Proceedings ...
... and Applications (AISTA).Mori, M., Miura, T., and Shioya, I. 2006. Topic detection and tracking for news web pages. In Proceedings ofthe IEEE/WIC/ACM ...
... Article 4, Publication date: December 2013.????????A Graph Analytical Approach for Topic Detection 4:23Wartena, C. and Brussee, R. 2008. Topic detection by clustering keywords. In Proceedings of the IEEE Com-puter Society ...

Title:
Cost-Effective Online Trending Topic Detection and Popularity Prediction in Microblogging

Keywords:
Topic detection

References:
James Allan (Ed.). 2002. Topic Detection and Tracking: Event-based Information Organization. Kluwer Academic Publishers, Norwell, MA.
Mario Cataldi, Luigi Di Caro, and Claudio Schifanella. 2010. Emerging topic detection on twitter based on temporal and social terms evaluation. In Proceedings of the 10th International Workshop on Multimedia Data Mining (MDMKDD’10). ACM, New York, NY, Article 4, 10 pages.
Juha Makkonen, Helena Ahonen-Myka, and Marko Salmenkivi. 2004. Simple semantics in topic detection and tracking. Inform. Retriev. 7, 3--4 (Sep. 2004), 347--368.
Zhongchen Miao, Kai Chen, Yi Zhou, Hongyuan Zha, Jianhua He, Xiaokang Yang, and Wenjun Zhang. 2015. Online trendy topics detection in microblogs with selective user monitoring under cost constraints. In Proceedings of the 2015 IEEE International Conference on Communications (ICC’15). 1194--1200.
Georgios Petkos, Symeon Papadopoulos, and Yiannis Kompatsiaris. 2014. Two-level message clustering for topic detection in twitter. In Proceedings of the SNOW 2014 Data Challenge. 49--56.

Full Text:
TOIS3503-1818Cost-Effective Online Trending Topic Detection and PopularityPrediction in MicrobloggingZHONGCHEN MIAO and KAI CHEN, Shanghai Jiao ... feature selection; Information extraction; Social networks;Additional Key Words and Phrases: Topic detection, , prediction, microblogging, costACM Reference Format:Zhongchen Miao, Kai Chen, Yi ... Zhou, Wenjun Zhang, and Hongyuan Zha. 2016. Cost-effective online trending topic detection and popularity prediction in microblogging. ACM Trans. Inf. Syst.35, 3, ...
... even individuals are in need of a reliable onlinereal-time trending topics detection and prediction system on microblogging servicesand other social networks, which ... accurate, and even customizedresults from a third-party perspective.Traditionally, online trending topics detection and prediction systems for microblog-ging comprise three major steps: (1) ... No. 3, Article 18, Publication date: December 2016.Cost-Effective Online Trending Topic Detection and Popularity Prediction in Microblogging 18:3dataset at those time periods ... needed in order to remove the bias and get repre-sentative topic detection results for all times. However, for any third-party analyzers,the tremendous ...
... article can be summarized as follows:(1) We treat online trending topics detection in microblogs as a multi-coverage problem:How to select a subset ... topics that users are more interested in.(3) We integrate trending topics detection and their future popularity prediction intoa single system. We propose ...
... detection with hashed significance thresholds[Schubert et al. 2014], real-time emergent topic detection in blogs [Alvanaki et al. 2012],?TwitterMonitor? trend detection system that ... No. 3, Article 18, Publication date: December 2016.Cost-Effective Online Trending Topic Detection and Popularity Prediction in Microblogging 18:5Kupavskii et al. [2013]. There ... andJordan [2010].From all the above works, we see that various topic detection and analysis systemswith different purposes, structures, and algorithms have been ... performance is restricted in real-time operations.Our proposed online microblogging trending topics detection and popularity predic-tion system differs from the above reported systems ... the idea of selectingsubset users for single tasks such as topic detection or topic prediction in microblogswas proposed. Some other greedy-based algorithms ... the cost-effective framework and propose an integratedsystem for both trending topics detection as well as topic future popularity predic-tion in microblogs. Hence, ...
... user selection;?Module III: Real-time online data retrieval;?Module IV: Online trending topic detection; ;?Module V: Online trending topic popularity prediction.In general, Module I ... Miao et al.Fig. 1. Overall framework of our microblog trending topics detection and prediction system. Subset usersare selected by Modules I and ... No. 3, Article 18, Publication date: December 2016.Cost-Effective Online Trending Topic Detection and Popularity Prediction in Microblogging 18:7pre-processing procedures, will be introduced. ...
... No. 3, Article 18, Publication date: December 2016.Cost-Effective Online Trending Topic Detection and Popularity Prediction in Microblogging 18:9is related to the user?s ... et al. 1998] in this online detection module.The online trending topic detection steps are outlined as follows, while the mathe-matical definition will ...
... methods that are compatible withour framework in accomplishing online trending topics detection task.3.5. Online Trending Topic Popularity PredictionAfter a trending topic is ... No. 3, Article 18, Publication date: December 2016.Cost-Effective Online Trending Topic Detection and Popularity Prediction in Microblogging 18:11Denoting mv as node v?s ... Function for DetectionThis subsection formulates the loss function of trending topic detection in microblogsby selected subset users S.A node v (v ? ... 0.As mentioned in Section 1, selecting subset users for trending topic detection isa multi-coverage ?sensor placement? problem in a microblog propagation network.Therefore, ...
... No. 3, Article 18, Publication date: December 2016.Cost-Effective Online Trending Topic Detection and Popularity Prediction in Microblogging 18:13in Equation (8),Lpredict(E,S) =?e?E??p?ope(t? ,V|?, ...
... No. 3, Article 18, Publication date: December 2016.Cost-Effective Online Trending Topic Detection and Popularity Prediction in Microblogging 18:15ALGORITHM 1: Algorithm SWC for ...
... No. 3, Article 18, Publication date: December 2016.Cost-Effective Online Trending Topic Detection and Popularity Prediction in Microblogging 18:17ALGORITHM 2: Algorithm JNT for ...
... No. 3, Article 18, Publication date: December 2016.Cost-Effective Online Trending Topic Detection and Popularity Prediction in Microblogging 18:19similarities between YGSe and every ...
... No. 3, Article 18, Publication date: December 2016.Cost-Effective Online Trending Topic Detection and Popularity Prediction in Microblogging 18:21Table II. Statistics of Followers ...
... only subset users? microposts as data sourcesfor real-time online trending topic detection, , we also run an experiment using a state-of-art detection ... No. 3, Article 18, Publication date: December 2016.Cost-Effective Online Trending Topic Detection and Popularity Prediction in Microblogging 18:23The detection precision rate is ...
... No. 3, Article 18, Publication date: December 2016.Cost-Effective Online Trending Topic Detection and Popularity Prediction in Microblogging 18:25the table, column ?KS? shows ...
... No. 3, Article 18, Publication date: December 2016.Cost-Effective Online Trending Topic Detection
... it isnot practically suitable to accomplish a real-time online trending topics detection task20In all online testing experiments, a commodity computer with 2.0GHz ... No. 3, Article 18, Publication date: December 2016.Cost-Effective Online Trending Topic Detection and Popularity Prediction in Microblogging 18:29Fig. 2. Performance and cost ...
... No. 3, Article 18, Publication date: December 2016.Cost-Effective Online Trending Topic Detection and Popularity Prediction in Microblogging 18:31Table VIII. Performance Comparison for ... and thus we can accomplish the joint tasks of trending topic detection andprediction several hours in advance of the official lists. This ... a third-party system that is very practical in both earlytrending topic detection and early prediction for real microblogging services, using arelatively small ... of ? increases; and the system willfocus more on user?s topic detection ability when ? drops.To exhibit the effect of ?, we ... No. 3, Article 18, Publication date: December 2016.Cost-Effective Online Trending Topic Detection and Popularity Prediction in Microblogging 18:337. CONCLUSIONS AND FUTURE WORKIn ... FUTURE WORKIn this article, we present a cost-effective online trending topic detection and predictionsystem for microblogging services from a third-party perspective. The ...
... Mining (WSDM?13). ACM, New York, NY, 607?616.DOI:http://dx.doi.org/10.1145/2433396.2433473James Allan (Ed.). 2002. Topic Detection and Tracking: Event-based Information Organization. KluwerAcademic Publishers, Norwell, MA.Foteini Alvanaki, ... DOI:http://dx.doi.org/10.1016/j.comnet.2012.10.007Mario Cataldi, Luigi Di Caro, and Claudio Schifanella. 2010. Emerging topic detection on twitter basedon temporal and social terms evaluation. In Proceedings ...
... No. 3, Article 18, Publication date: December 2016.Cost-Effective Online Trending Topic Detection and Popularity Prediction in Microblogging 18:35Jure Leskovec, Lars Backstrom, and ...
... Zha, Jianhua He, Xiaokang Yang, and Wenjun Zhang. 2015.Online trendy topics detection in microblogs with selective user monitoring under cost constraints.In Proceedings ...

Keywords:
topic detection

Abstract:
... detail. This short paper describes research on generic algorithms for topic detection and segmentation that are applicable on texts of heterogeneous types ...

Full Text:
... detail. This short paper describes research on generic algorithms for topic detection and segmentation that are applicable on texts of heterogeneous types ... are applicable on texts of heterogeneous types and domains. Keywords Topic detection; ; text indexing; summarization. 1. INTRODUCTION Topic segmentation of texts ...

Keywords:
topic detection

References:
J. Allan, J. Carbonell, G. Doddington, J. Yamron and Y. Yang. Topic detection and tracking pilot study: final report. In Proceedings of the DARPA Broadcast News Transcription and Understanding Workshop, 1998.

Full Text:
... and Retrieval ? miscellaneous General Terms: Algorithms, Design, Performance Keywords: Topic detection, , information diffusion, topic evolution, FounderWISE. With the rapid increase ... the topics and obtain useful information by discovering topic-related knowledge. Topic detection and tracking (TDT) [1] is one of such techniques to ... stream. However, almost all previous works on TDT focus on topic detection on a small benchmark TDT corpus. Because the corpus is ... algorithm (INCR) is the most widely used algorithm for online topic detection. . It sequentially processes the input documents, one at a ...
... visits China?. The ?same? relationships can be identified in the topic detection algorithm because the same topics should have the same topic ... same topics should have the same topic identifier in the topic detection algorithm. The ?relevant? relationships can be identified by comparing the ... Figure 3 gives an example of evolution graph for news topics detected between 2007-10-12 and 2007-10-18 for the politics category. Twenty hot ... Allan, J. Carbonell, G. Doddington, J. Yamron and Y. Yang. Topic detection and tracking pilot study: final report. In Proceedings of the ...

Keywords:
topic detection

Abstract:
... sampling and evaluate our proposal by link detection task of Topic Detection

Full Text:
... Gibbs sampling and evaluate our proposalby link detection task of Topic Detection and Tracking.Categories and Subject Descriptors: I.2.6 [ArtificialIntelligence]: Learning; H.3.3 [Information ...

Title:
Topic Detection from Large Scale of Microblog Stream with High Utility Pattern Clustering

Keywords:
topic detection

Abstract:
... that the developed method achieves better performance than other existing topic detection methods, leading to a desirable solution of detecting event from ...

References:
J. Allan, V. Lavrenko, D. Malin, and R. Swan. Detections, bounds, and timelines: Umass and tdt-3. In Proceedings of topic detection and tracking workshop, pages 167--174, 2000.
L. AlSumait, D. Barbara, and C. Domeniconi. On-line lda: adaptive topic models for mining text streams with applications to topic detection and tracking. In ICDM, pages 3--12, 2008.
G. Petkos, S. Papadopoulos, L. Aiello, R. Skraba, and Y. Kompatsiaris. A soft frequent pattern mining approach for textual topic detection. In WIMS, pages 25:1--25:10, 2014.

Full Text:
Topic Detection from Large Scale of Microblog Stream withHigh Utility Pattern ClusteringJiajia ... of eventdetection from conventional media sources have been ad-dressed in Topic Detection and Tracking research program[9][12], they do not work well on ... of data and inability tobacktrack over previously arriving transactions.Studying of topic detection on microblog (or streams) hasbecome an active research area. Most ...
... is organized as follows. Section2 reviews the related work of topic detection. . Section 3describes the HUPC framework in detail. Section 4 ... experimental results. We conclude the paperin Section 5.2. RELATEDWORKMethodologically, general-purposed topic detection meth-ods largely fall in three classes: document-pivot method,probability topic model, ...
... ?rst category focuses on selecting a group of burstyterms for topic detection. . For example, TwitterMonitor [16]?rst identi?ed bursty keywords that have ... de-tection method. The second category focuses on selectingco-occurrence features for topic detection, , such as frequentpatterns, text segments (i.e., n-grams), and memes. ... used for detecting topics from Twitter. Petkoset al. treated the topic detection problem as a frequent pat-tern mining process and proposed a ... g,where Ti(i = 1; 2; :::; LN ) denotes the topic detected fromDL. This topic is either a coherent topic appeared in ... DL?1, or a new emerging topic. Therefore,in this paper, our topic detection task is to detect topicsfTLg from the latest text batch ...
... infor-mation, which in fact is a kind of information redundancyfor topic detection. . In this paper, we hope to reduce this re-dundancy ...
... quali?ed to participate in the8Table 1: The comparison among di?erent topic detection methods on two datasets.TwitterSet SinaSetHUPC-Stream MP On-LDA SFPM HUPC-Stream MP ...
... situation: 1) only a few of tweets inTable 3: Coherent topics detected from ?ve sequen-tial segmentsBatch id Topical words34 iphon, ipad, mini, ...
... R. Swan.Detections, bounds, and timelines: Umass and tdt-3.In Proceedings of topic detection and trackingworkshop, pages 167{174, 2000.[3] L. AlSumait, D. Barbara, and ... On-linelda: adaptive topic models for mining text streamswith applications to topic detection and tracking. InICDM, pages 3{12, 2008.[4] Z. Chen and B. ... Skraba, andY. Kompatsiaris. A soft frequent pattern miningapproach for textual topic detection. . In WIMS, pages25:1{25:10, 2014.[19] S. Petrovic, M. Osborne, and ...

Keywords:
topic detection

References:
L. AlSumait, D. Barbará, and C. Domeniconi. On-line lda: Adaptive topic models for mining text streams with applications to topic detection and tracking. In ICDM '08, pages 3--12, 2008.
M. Cataldi, L. Di Caro, and C. Schifanella. Emerging topic detection on twitter based on temporal and social terms evaluation. In MDMKDD '10, pages 4--13, 2010.

Full Text:
... DescriptorsH.3.4 [Information Systems]: Information Storage andRetrieval?Systems and SoftwareKeywordsOnline social networks, topic detection, , network analysis1. INTRODUCTIONOnline social networks allow hundreds of millions ... the various approaches. Besides thedifficulties of developing new techniques for topic detection, ,these tasks necessitate generally a heavy preprocessing stepwhich is performed ... data. SONDY is an opensource platform integrating optimized implementations ofsome topic detection and graph mining algorithms in thesame platform. The application relies ...
... stop-wordsremoval, content stemming, message stream discretiza-tion, and message stream resizing.1http://mediamining.univ-lyon2.fr/sondy/10052. Topic detection and exploration service: for identifyingand temporally locating trending topics and ... : for importing new algorithms tobe used by the the topic detection or network analysisservices.To illustrate the capabilities of SONDY , the ... filter: reduces words to their stem to im-prove efficiency of topic detection algorithms, such astopic model based techniques.2.3 Topic Detection and ExplorationThis service enables (i) applying different topic detectionalgorithms on ... exploration service.summarize results. So far we have implemented the follow-ing topic detection
... can scale to massive networks and calculatesstructural properties.Some tools for topic detection and visualization have beendeveloped in the recent years. They mainly ... tag clouds of topics extracted from tweets using a sim-ple topic detection algorithm that uses a search engine as anexternal knowledge base.5. ... work, we intend to enrich theavailable range of techniques for topic detection, , trends anal-ysis, as well as network analysis to offer ... ofalgorithms for comparison. Algorithms include for exampleOnline LDA [1] for topic detection and community detectiontechniques for network analysis. To increase performances,we plan ... On-linelda: Adaptive topic models for mining text streamswith applications to topic detection and tracking. InICDM ?08, pages 3?12, 2008.[2] V. Batagelj and ...

Title:
Image-regulated graph topic model for cross-media topic detection

Keywords:
topic detection

Full Text:
... recog-nition and scenario understanding [16] other than the initialtasks for topic detection and understanding. Moreover, topicmodel is introduced to solve the ?sparsity ...
... uncovering topic for normal documents, they rarely ap-ply to the topic detection task. Hashtag Graph based TopicModel for Tweet Mining (HGTM) [17] ...

Keywords:
topic detection and tracking

Abstract:
... news issue construction is proposed. The first step is a topic detection process, in which newly appearing stories are clustered into new ... construction experiments. F-measure of the best results is either above (topic detection) ) or close to (topic detection and tracking) 90%. Four news issue construction results are successfully ...

References:
M. Spitters and W. Kraaij. TNO at TDT2001: Language Model-Based Topic Detection. Topic Detection and Tracking Workshop Report, 2001.
D. Trieschnigg and W. Kraaij. TNO Hierarchical topic detection report at TDT 2004. Topic Detection and Tracking Workshop Report, 2004.
M.-Q. Yu, W.-H. Luo, Z.-T. Zhou and S. Bai. ICT's Approaches to HTD and Tracking at TDT2004. Topic Detection and Tracking Workshop Report, 2004.
M. Connell, A. Feng, G. Kumaran, and et al. UMass at TDT 2004. Topic Detection and Tracking Workshop Report, 2004.

Full Text:
... news issue construction is proposed. The first step is a topic detection process, in which newly appearing stories are clustered into new ... construction experiments. F-measure of the best results is either above (topic detection) ) or close to (topic detection and tracking) 90%. Four news issue construction results are successfully ... Software ? Performance evaluation. General Terms: Experimentation Keywords News issue, Topic detection and tracking, Clustering 1. INTRODUCTION People often have information needs ... update an existing news issue or create a novel one. Topic Detection and Tracking (TDT) tasks are intended to structure news stories ...
... into news issues. We select a clustering method as our topic detection algorithm. Experiments are made to compare the results of various ... model are tuned to achieve the best performance. (2) Our topic detection and tracking algorithm firstly clusters newly appearing stories into new ... 2 gives a brief review of related work in Hierarchical Topic Detection. . Section 3 presents our topic detection and tracking algorithm. The news issue construction algorithm is proposed ... discussion of future work in Section 7. 2. RELATED WORK Topic detection and tracking (TDT) techniques have developed for years [2, 3, ... but there still lack accurate, efficient and practical solutions. Hierarchical Topic Detection (HTD) task was proposed and evaluated in TDT2004. The best ... part of our work can be regarded as a two-layer topic detection and tracking algorithm for practical use. We suggest it unnecessary ...
... clustering in [21]. Their system is a well-developed application for topic detection and tracking, while our work differs in the topic hierarchy, ... the topic hierarchy, topic similarity calculation and clustering algorithms. 3. TOPIC DETECTION AND TRACKING BASED ON CLUSTERING A news issue usually contains ... topics, and so we borrow its name. We propose our topic detection and tracking system based on clustering, considering the characteristics of ... models are updated after the operation of combination. The whole topic detection and tracking process is performed online automatically. 3.1 Pre-Processing and ... t, their similarity is calculated as: ???='),'(*),()',(ddwttt wdweightwdweightddsimilarity (4) 3.3 Topic Detection Using New Stories New coming stories are clustered into new ... their pair-wise similarities, which is similar to the process of Topic Detection in TDT. The clustering results, called as new topic candidates, ... single pass and reallocation clustering method in [8], we perform topic detection process within new stories for two reasons. First, the same ... averages), a traditional agglomerative clustering method, is arranged to perform topic detection
... not putting new and old stories together to operate the topic detection algorithm and generate all the topics. Firstly, we can easily ... for N days are frozen, like what [8] did in topic detection. . Similarities between every new topic candidate and previous one ... News Issue Construction Based On Topic Combination Similar to the topic detection and tracking process described in Section 3, the news issue ...
... EXPERIMENTS AND DISCUSSIONS Experiments are firstly done to evaluate our topic detection algorithm. Parameters are tuned and the results are compared with ... tracking algorithm. News issue construction results are demonstrated finally. 6.1 Topic Detection Results We did some experiments to test the topic detection algorithm we had selected firstly, based on the pre-processing process, ... compared in our experiments. 6.1.1 Selection of Clustering Method Our topic detection algorithm is actually a topic clustering algorithm. We surveyed the ... compared on Dataset2, with the help of cluto. Since the topic detection process deals with new stories and new stories are coming ... not appear too many stories in few minutes, so the topic detection algorithm does not have the requirement of dealing with a ...
... environment and the size is large enough to test our topic detection algorithm. We compare different clustering algorithms implemented in cluto firstly. ... achieved on Dataset2. Dataset2 is a relatively easier dataset for topic detection, , for some confusing stories have been deliberately added to ...
... weight 3 is our chosen method of using titles. 6.2 Topic Detection and Tracking Results Dataset2 is used to test our topic detection and tracking algorithm. The news stories of Dataset2 are from ... 0:01 to 01:00 are firstly clustered into topics as the topic detection process does. Then news stories appearing from 01:01 to 02:00 ... to perform the topic tracking process with existing topics. The topic detection and tracking process is repeated 23 times until all the ... (except the first part which is only processed by the topic detection algorithm) of the dataset have been processed. We evaluate all ... once one more part of the dataset is processed. The topic detection threshold d? is set as 0.225 and the topic tracking ... 19:00 21:00 23:00Time of the dayPrecisonRecallF-measure Figure 6. Performance of topic detection and tracking at 24 hours of the day on Dataset2 ... Figure 6 we find out that the performance of the topic detection and tracking process is good as a whole. The final ... Compare this result with the best result achieved in the topic detection process which takes Dataset2 as a whole and perform the ...
... are done as Section 4 describes. Firstly, we perform the topic detection and tracking process with Web pages from news websites and ... issue construction. Newly crawled Web pages are used to do topic detection. . The result is used to update the previous topics ... their feedbacks are summarized as follows: (1) Results of the topic detection and tracking process do not seem as good as what ...
... Our algorithm includes three steps. The first step is a topic detection process, in which newly appearing stories are clustered into new ... M. Spitters and W. Kraaij. TNO at TDT2001: Language Model-Based Topic Detection. . Topic Detection and Tracking Workshop Report, 2001. [9] Y. Zhao and G. ... Results, http://www.nist.gov/speech/tests/tdt/tdt2004/papers/NIST-TDT2004.ppt [16] D. Trieschnigg and W. Kraaij. TNO Hierarchical topic detection report at TDT 2004. Topic Detection and Tracking Workshop Report, 2004. [17] M.-Q. Yu, W.-H. Luo, ... S. Bai. ICT?s Approaches to HTD and Tracking at TDT2004. Topic Detection and Tracking Workshop Report, 2004. [18] M. Connell, A. Feng, ... Feng, G. Kumaran, and et al. UMass at TDT 2004. Topic Detection and Tracking Workshop Report, 2004. [19] G.P.C. Fung, J.X. Yu, ...

Title:
Cross-Platform Emerging Topic Detection and Elaboration from Multimedia Streams

Keywords:
Topic detection

Abstract:
... and more attractive to provide users a solution of emerging topic detection and elaboration. And this posts a real challenge to both ... large outlier noises. This article provides a method on emerging topic detection and elaboration using multimedia streams cross different online platforms. Specifically, ...

References:
Foteini Alvanaki, Michel Sebastian, Krithi Ramamritham, and Gerhard Weikum. 2011. EnBlogue: Emergent topic detection in Web 2.0 streams. In Proceedings of the ACM International Conference on Management of Data. 1271--1274.
Shiva Prasad Kasiviswanathan, Prem Melville, Arindam Banerjee, and Vikas Sindhwani. 2011. Emerging topic detection using dictionary learning. In Proceedings of the ACM International Conference on Information and Knowledge Management. 745--754.

Full Text:
TOMM1104-5454Cross-Platform Emerging Topic Detection and Elaboration fromMultimedia StreamsBING-KUN BAO, CHANGSHENG XU, and WEIQING MIN, ... more and more attractiveto provide users a solution of emerging topic detection and elaboration. And this posts a real challenge toboth industrial ... large outlier noises. This article provides a method on emerging topic detection andelaboration using multimedia streams cross different online platforms. Specifically, Twitter, ... Retrieval]: Information Search andRetrieval?ClusteringGeneral Terms: AlgorithmsAdditional Key Words and Phrases: Topic detection, , cross-platform, cross-media, coclusteringACM Reference Format:Bing-Kun Bao, Changsheng Xu, Weiqing ... Changsheng Xu, Weiqing Min, and Mohammod Shamim Hossain. 2015. Cross-platformemerging topic detection and elaboration from multimedia streams. ACM Trans. Multimedia Comput. Com-mun. ... B.-K. Bao et al.Fig. 1. The overview of emerging social topic detection and cross-platform elaboration.tweets per minute during his 200-meter race in ...
... all these three platforms to delivera robust method of emerging topic detection and elaboration. Specifically, Twitter, NewYork Times and Flickr are selected, ... extracted emerg-ing data and data relationships, a robust and effective topic detection and elaborationmethod is expected.The proposed framework includes three stages: emerging ... proposed framework includes three stages: emerging keyword extraction, emerg-ing social topic detection, , and cross-media elaboration, as shown in Figure 1. In ... in stageof emerging social detection, we propose a novel multimedia topic detection method,called Robust Cross-Platform Multimedia Co-Clustering (RCPMM-CC), to simultane-ously cocluster the ... 11, No. 4, Article 54, Publication date: April 2015.Cross-Platform Emerging Topic Detection and Elaboration from Multimedia Streams 54:3Fig. 2. Two kinds of ... of the proposed framework, thatis, Emerging Keyword Extraction, Emerging Social Topic Detection, , and Cross-MediaElaboration. Experimental results are reported in Section 6 ... future work in Section 8.2. RELATED WORKAs our work involves topic detection, , topic elaboration, and coclustering, the review ofrelated work is ...
... But it heavily relies on the2As this article only discusses topic detection but not the extension of the proposed method, the experimentsare ... its popularity over time. Aiello et al. [2013] found thattrending topic detection method based on n-grams cooccurrence and a time-dependentboost (df ? ... platforms. For example, Alvanaki et al. [2011] devel-oped an emerging topic detection system, called En Blogue, by identifying shifts inthe correlations between ... 11, No. 4, Article 54, Publication date: April 2015.Cross-Platform Emerging Topic Detection and Elaboration from Multimedia Streams 54:5system does not process the ...
... 11, No. 4, Article 54, Publication date: April 2015.Cross-Platform Emerging Topic Detection and Elaboration from Multimedia Streams 54:7Section 4.3, and Section 4.4 ... last, we will present our proposed RCPMM-CC algorithm for emerging topic detection. .4.1. Problem DefinitionThe target of our work is to simultaneously ...
... 11, No. 4, Article 54, Publication date: April 2015.Cross-Platform Emerging Topic Detection and Elaboration from Multimedia Streams 54:9Fig. 3. The coclustering process ...
... 11, No. 4, Article 54, Publication date: April 2015.Cross-Platform Emerging Topic Detection and Elaboration from Multimedia Streams 54:11ALGORITHM 1: Robust Cross-Platform Multimedia ...
... those with high weights. Note thatthe results from emerging social topic detection include not only the clusters of Twit-ter keywords, news and ... 11, No. 4, Article 54, Publication date: April 2015.Cross-Platform Emerging Topic Detection
... experimental results with real world data on emergingkeyword extraction, emerging topic detection and topic elaboration respectively.7.1. Real World DatasetTo evaluate our work, ...
... platform. This baseline is used to evaluate theperformance on emerging topic detection by not introducing the complementary mediaplatforms. BCC-TN and BCC-TF are ... toleration. Thesebaselines are used to evaluate the performance on emerging topic detection with onlyone introduced complementary media platform. ITCC-ALL is high-order coclusteringwith ... 11, No. 4, Article 54, Publication date: April 2015.Cross-Platform Emerging Topic Detection and Elaboration from Multimedia Streams 54:15TableIII.TheTop20EmergingKeywordsinEachTimeInterval03/11-03/1703/18-03/2403/25-03/3103/11-03/1703/18-03/2403/25-03/311Japan(0.2524)Rebecca(0.0397)Japan(0.0487)11Japanese(0.0268)Austin(0.0213)iPad(0.0180)2tsunami(0.1027)iPad(0.0296)Gaga(0.0337)12Patrick(0.0255)moon(0.0176)cricket(0.0173)3earthquake(0.0689)H1N1(0.0268)Lady(0.0298)13facebook(0.0220)April(0.0166)earthquake(0.0173)4Oscars(0.0598)Sheen(0.0264)Jackie(0.0265)14Fukushima(0.0217)March(0.0153)good(0.0158)5Rebecca(0.0567)Taylor(0.0250)F1(0.0241)15Potter(0.0214)Oscar(0.0139)Twitter(0.0152)6nuclear(0.0508)Libya(0.0249)Austin(0.0240)16Harry(0.0207)winning(0.0138)love(0.0140)7Austin(0.0503)Charlie(0.0248)tsunami(0.0237)17relief(0.0201)Japan(0.0116)day(0.0134)8Dogg(0.0407)tsunami(0.0223)YouTube(0.0235)18plant(0.0194)apple(0.0116)live(0.0119)9nate(0.0370)mobile(0.0219)March(0.0235)19st(0.0190)love(0.0115)apple(0.0114)10quake(0.0317)Elizabeth(0.0215)Britney(0.0187)20winning(0.0177)supermoon(0.0113)rumors(0.0110)ACM Trans. Multimedia Comput. Commun. ... it is assigned with 0.We use three criteria on emerging topic detection
... 11, No. 4, Article 54, Publication date: April 2015.Cross-Platform Emerging Topic Detection and Elaboration from Multimedia Streams 54:17Table V. Top 10 Emerging ... B.-K. Bao et al.Fig. 5. (a) Precision comparison on emerging topic detection; ; (b) NDCG comparison on emerging topicdetection; (c) mAP comparison ...
... vs. RCP-CC/GS-HOCC/RCPMM-CC. This provesthat both complementary platforms are beneficial to topic detection. . 3) Coclusteringon Twitter and New York Times achieves better ... especially in Figure 5(c), likely due to the lowprecision of topic detection. . 6) For coclustering in pairwise correlated structure on threeplatforms, ... the process of discarding outliers will bring better performance on topic detection, ,especially on topic ranking and finding true emerging keywords for ... 11, No. 4, Article 54, Publication date: April 2015.Cross-Platform Emerging Topic Detection and Elaboration from Multimedia Streams 54:19Fig. 6. (a) mAP comparison ...
... 11, No. 4, Article 54, Publication date: April 2015.Cross-Platform Emerging Topic Detection and Elaboration from Multimedia Streams 54:21K. Jarvelin and J. Kekalainen. ... Kasiviswanathan, Prem Melville, Arindam Banerjee, and Vikas Sindhwani. 2011. Emerg-ing topic detection using dictionary learning. In Proceedings of the ACM International Conference ...

Title:
Major topic detection and its application to opinion summarization

Keywords:
topic detection

Full Text:
Proceedings Template - WORDMajor Topic Detection and Its Application to Opinion Summarization Lun-Wei Ku, Li-Ying Lee, ... process.General Terms Algorithms, Design, Experimentation. Keywords Opinion Summarization, Sentence Retrieval, Topic Detection. . 1. Introduction Watching specific information sources and summarizing the ... posed beforehand, detecting opinions is similar to the task of topic detection on sentence level. Besides telling which opinions are positive or ... such opinions are also important. This paper proposes a major topic detection mechanism to capture main concepts embedded implicitly in a relevant ... a relevant document set is the spirit of our major topic detection. . A term is considered to be representative if it ... set are relevant. That meets the assumption of our major topic detection. . Take set 2 (?clone Dolly sheep?) as an example. ...
... relevant sentence identification to verify the accuracy of the major topic detection. . The more accurate the relevant sentence retrieval is, the ... the relevant sentence retrieval is, the more precise the major topic detection is. Sentences with more qualified terms are considered as relevant. ... opinion summarization system is proposed in this paper, including major topic detection, , relevant sentence retrieval, opinion-oriented sentence identification, and summarization. Our ...

Keywords:
topic detection and tracking (TDT)

References:
Juha, M., Helena, A.M., Marko, S.: Simple Semantics in Topic Detection and Tracking. Information Retrieval 7(3-4), 347-368 (2004).

Full Text:
... a TDT system based on NE detection. Keywords: Named entity, Topic Detection and Tracking (TDT). 1 Introduction Topic Detection and Tracking (TDT) aims to effectively retrieve and organize broadcast ...
... evaluation is important to improve the performance of techniques for topic detection as it depends on the correct classification of named entities. ... [3] Juha, M., Helena, A.M., Marko, S.: Simple Semantics in Topic Detection and Tracking. Information Retrieval 7(3?4), 347?368 (2004) [4] Mohd, M.: ...

Keywords:
automated topic detection

Abstract:
... measure the influence of OCR errors and historical orthography on topic detection, , we created ground truth versions and in addition ground ...

Full Text:
... to measure the influence of OCR errors and historicalorthography on topic detection, , we created ground truthversions and in addition ground truth ... orthography.Categories and Subject DescriptorsH.3.6 [Library Automation]: Large text archivesGeneral TermsDocumentationKeywordsAutomated Topic Detection, , Topic and Subject Classifica-tion1. INTRODUCTIONAs a consequence of Google ... are often fullof historical spelling variants. How seriously does thisaffect topic detection? ? Which improvements could beexpected from an orthographic normalization?3. What ... normalization?3. What other special effects can be observed when ap-plying topic detection to OCRed historical texts?To find a preliminary answer to these ...
... the givenpage. Section 6 gives a resume.The evaluation shows that topic detection generally workswell, though manually assigned topics are often not matched(exactly) ...
... topic may depend on a singletoken.25Figure 1: Recall values for topic detection in 14 OCRed pages when using fixed significance thresholds > ... truth texts with modernized orthography.Figure 2: Average recall values for topic detection in 14 OCRed pages. The four subfigures denote average recall ...
... 53/57/77 50/23/19 51/42/58 52/52/46 37/36/30 23/18/16Figure 3: Precision values for topic detection in 14 pages for fixed significance threshold >0,6, comparing OCRed ... misinterpreted as denoting an oldFigure 4: Average precision values for topic detection in 14OCRed pages. The diagram compares for fixed threshold> 0.6 ...

Title:
User-defined hot topic detection in microblogging

Keywords:
hot topic detection

Abstract:
... of microblogging, an important information communication and spreading channel, hot topic detection in it increasingly attracts researchers' interests. Currently, their interests mostly ... which includes user-defined keywords expansion, relevant microblogging filter and hot topic detection. . We also propose an effective algorithm for user-defined keywords ...

Full Text:
User-defined hot topic detection in microbloggingUser-defined hot topic detection in MicrobloggingYing ChenInstitute of Automation,Chinese Academy of Sciences95 Zhongguancun East ... which includes user-defined keywords expansion, rel-evant microblogging filter and hot topic detection. . We al-so propose an effective algorithm for user-defined keywordsexpansion ... algorithms.Categories and Subject DescriptorsH.4 [Information Systems Applications]: MiscellaneousKeywordsKeywords Expansion, Hot Topic Detection, , Microblogging1. INTRODUCTIONMicroblogging platforms such as Twitter or Sina offer ... time.To our knowledge, little research result concentrates onsuch user-defined hot topic detection approach. In this pa-per, we propose a unified framework to ... posts, which includes auto-matic keywords expansion, relevant microblogging filter andhot topic detection. . The main contributions are summarizedas follows: 1)This is the ...
... Otherresearch reports on Microblogging includes [2, 7, 1].However, personalized hot topic detection from microblog-ging posts is useful to both the individual and ... events just happened in a specificarea is convenient by personalized topic detection. . govern-ment departments who need to keep abreast of emergency ... section, we focus on presenting our major work,i.e., user-defined hot topic detection framework and also in-troducing algorithms for the automatic keywords expansion(AKE), ...
... the missing of relevant microbloggingmay leads to the loss of topic detection in the next step.Thus we set a lower threshold value ... hoc in the deci-sion stage of our logistic classifier.3.3 Hot topic detection algorithmAfter the filter step, hot topic detection algorithm is ap-plied to find the topic word related events. ... the efficiency of ourproposed algorithms for automatic keywords expansion(AKE),user-defined hot topic detection. .4.1 Results of automatic keywords expansionFor the purpose of evaluating ...
... Results of user-defined hot topic detectionWe conduct experiments of hot topic detection on real on-line micro-blog data to reveal the effectiveness of ... micro-blog data to reveal the effectiveness of our field-dependent hot topic detection algorithm. We collect datafrom Sina-Weibo, which is a Twitter-like microblogging ... less than 500 posts. If without micro-blog fil-ter before hot topic detection algorithm, those global eventswould inevitably cover local events, which would ... framework, based on automatic keywords expansion,user-defined microblogging filter and hot topic detection. . ItFigure 4: Comparison of event posts distributionsolves the problem ... work that pays close attention tothe personalized or field-dependent hot topic detection frommicroblogging posts, 2) fusing both the importance and rele-vance factor, ...

Keywords:
topic detection

References:
F. Walls, H. Jin, S. Sista and R. Schwartz, (1999), "Probabilistic Models for Topic Detection and Tracking", IEEE
JingQiu, LeJian Liao, XiuJie Dong, (2008), " Topic Detection and Tracking for Chinese News Web Pages", International conference on Advanced Language Processing and Wen Information Technology, IEEE
Paula Hatch, Nicola Stokes and Joe Carthy, "Topic detection, a new application for Lexical Chaining"
Sungjick Lee, Han-joon Kim, (2008), "News Keyword Extraction for Topic Tracking", 4^th International Conference on Networked Computing and Advanced Information Management, IEEE Topic Detection and Tracking, Available: www.projects.ldu.upenn.edu
Xiang Ying Dai, Qing Cai Chen, Xiao Long Wang and Jun Xu, (2010), "Online Topic Detection and Tracking of financial News based on Hierarchical Clustering", Proceedings of the 9^th International Conference on Machine Learning and Cybernetics, IEEE

Full Text:
... I.5.3 [Clustering]: Algorithms General Terms Algorithms, Performance Keywords Text Mining, Topic detection, , topic tracking. 1. INTRODUCTION Text mining is a new ...
... story is judged to be on-topic; otherwise, off-topic. 2.1. Background [12][25][26]Topic Detection and tracking is fairly a new area of research in ...
... case, which has the same idea with the TDT Hierarchical Topic Detection (HTD). DCEM can simplify the model to describe the events ...
... average-link method, which is then used to implement the retrospective topic detection and the online topic detection of news stories of the stocks. Additionally, the improved single ...
... Jin, S. Sista and R. Schwartz, (1999), ?Probabilistic Models for Topic Detection and Tracking?,IEEE Hongxiang Diao, Zhansheng Bai and Xilin Yu, (2010), ... Information Technology, IEEE JingQiu, LeJian Liao, XiuJie Dong, (2008), ? Topic Detection and Tracking for Chinese News Web Pages?, International conference on ...
... 841-872 OmidDadgar,?TopicDetectionandtracking?,Available: www.tcnj.edu/~mmmartin/.../TDT/TopicDetectionTracking04.ppt Paula Hatch, Nicola Stokes and Joe Carthy, ?Topic detection, , a new application for Lexical Chaining? Shengdong Li, Xueqiang ... International Conference on Networked Computing and Advanced Information Management, IEEE Topic Detection and Tracking, Available: www.projects.ldu.upenn.edu Vishal Gupta, G.S. Lehal, (2009), ?A ... Cai Chen, Xiao Long Wang and Jun Xu, (2010), ?Online Topic Detection and Tracking of financial News based on Hierarchical Clustering?, Proceedings ...

Keywords:
emerging topic detection

Full Text:
... it heavily relies on the users' contribution.Some e?orts on emerging topic detection in social mediastreams are constrained by the short message length ... NewsTwitter KeywordsFlickrJapanTsunamiearthquakenuclearJapan Tokyo earthquake tsunamijapan march sendai2011earthquakeEmerging Keyword Extraction Module??Emerging Topic Detection ModuleCross Media Retrieval Module????1)?8.8?magnitudeEarthquake?hit?Japan2)?Shocking?photos?of?the?earthquake?in?JapanPublicOpion:Figure 1: The framework of Me-digger.Figure 1 ... of Me-digger with fourparts: Data Collection, Emerging Keywords Extraction, E-merging Topic Detection, ... , and Cross Media Retrieval. Themost challenging part is Emerging Topic Detection, , in whichwe enrich the correlation between detected Twitter key-words ...

Title:
Emerging topic detection on Twitter based on temporal and social terms evaluation

Keywords:
topic detection

Abstract:
... this primary role of Twitter and we propose a novel topic detection technique that permits to retrieve in real-time the most emergent ...

References:
J. Allan, editor. Topic detection and tracking: event-based information organization. Kluwer Academic Publishers, Norwell, MA, USA, 2002.
J. Makkonen, H. Ahonen-Myka, and M. Salmenkivi. Simple semantics in topic detection and tracking. Inf. Retr., 7(3--4):347--368, 2004.

Full Text:
E:/Articoli/mdm-kdd 2010/sig-alternate - MDM-KDD.dviEmerging Topic Detection on Twitterbased on Temporal and Social Terms EvaluationMario CataldiUniversit di ... recognize this primary role of Twitter andwe propose a novel topic detection technique that permits toretrieve in real-time the most emergent topics ...
... time range r set by the user (depending onthe preferred topic detection frequency), we define the t-thconsidered interval It asIt =< it, ...
... TGtwill contain the edge ?k, z? but not vice versa).7.3 Topic Detection and RankingSince in our system each topic is defined as ...
per hour), which included more than 300kdifferent keywords.Considering the topic detection method introduced in thispaper, we analyze different experiments: we initially ...
... pages 1?8, New York, NY,USA, 2010. ACM.[7] J. Allan, editor. Topic detection and tracking:event-based information organization. KluwerAcademic Publishers, Norwell, MA, USA, 2002.[8] ...
... ACM.[23] J. Makkonen, H. Ahonen-Myka, and M. Salmenkivi.Simple semantics in topic detection and tracking. Inf.Retr., 7(3-4):347?368, 2004.[24] P. Melville, R. J. Mooney, ...

Title:
An effective multi-clue fusion approach for web video topic detection

Keywords:
topic detection

Abstract:
... about the happened events. Due to the potential application prospects, topic detection attracts increasing research interests in the last decade. On one ... the hot topics. In this paper, different from the traditional topic detection methods, which mainly rely on data clustering, we propose a ... we propose a novel multi-clue fusion approach for web video topic detection. . In our approach, firstly by utilizing the video related ... keywords from the search engine are used as guidance for topic detection. . Finally, these clues are combined together to detect the ...

References:
J. Allan, J. G. Carbonell, G. Doddington, J. Yamron and Y. Yang. Topic detection and tracking pilot study: Final report. In DARPA Broadcast News Transcription and Understanding Workshop, 1998.

Full Text:
... video topic detectionAn Effective Multi-Clue Fusion Approach for Web Video Topic Detection Tianlong Chen 1 , Chunxi Liu 2 , Qingming Huang ... about the happened events. Due to the potential application prospects, topic detection attracts increasing research interests in the last decade. On one ... the hot topics. In this paper, different from the traditional topic detection methods, which mainly rely on data clustering, we propose a ... we propose a novel multi-clue fusion approach for web video topic detection. . In our approach, firstly by utilizing the video related ... keywords from the search engine are used as guidance for topic detection. . Finally, these clues are combined together to detect the ... Indexing - Abstracting methods. General Terms Algorithms, Design, Experimentation. Keywords Topic detection, , multi-clues fusion, tag group 1. INTRODUCTION With the rapid ... real world events, and therefore could enhance the user experience. Topic detection is such a technology to summarize information from the unstructured ... information from the unstructured web video data. The objective of topic detection is to discover new or previously unidentified event which refers ... thing that happens at a specific time and place [9]. Topic detection rises from detecting topics in news articles or blog posts ... event fragments. Although many methods have been proposed for text topic detection, , they are not suitable for web video topic detection. . In order to exact hot topics from the web ... framework for web video topic discovery and visualization. These video topic detection methods are mostly based on clustering the tags or keyframes. ...
... cluster number is an insurmountable problem for the unsupervised clustering topic detection approach. The users? queries recorded in the search engine could ... In summary, hot search queries provide us additional information for topic detection. . In this paper, we propose a novel multi-clue fusion ... find the hot topics. Compared with the existing web video topic detection methods, the main contributions of the proposed approach are summarized ... the proposed approach are summarized as follows: 1. A novel topic detection method is proposed based on tag groups. Unlike previous clustering ... further refine the tag group extraction result and makes the topic detection result more accurate. 3. The utilization of the users? intention ... of the users? intention is presented as the guidance for topic detection. . The hot search queries in the search engine reflect ...
... shows that hot search queries as an informative clue for topic detection could filter most false topics. Table 1. The comparison result ... we present an effective multi-clue fusion approach for web video topic detection. . First of all, dense-bursty tag groups are extracted by ... demonstrate that hot search queries are an effective guidance for topic detection and the proposed topic detection method is effective. In future work, we will try to ... to develop more effective multi-clue fusion strategy for web video topic detection. . 7. ACKNOWLEDGMENTS This work was supported in part by ...
... J. G. Carbonell, G. Doddington, J. Yamron and Y. Yang. Topic detection and tracking pilot study: Final report. In DARPA Broadcast News ...

Keywords:
topic detection

References:
J. Allen, editor. Topic Detection and Tracking: Event-based Information Organization. Springer, 2002.

Full Text:
... Keywords:TDT-5, Cluster Labeling, TopicDetection1. INTRODUCTIONOne curious aspect of the TDT (Topic Detection and Tracking)program organized by NIST [1], which ran from late ...
... the time. As it turns out, some of the ma-jor topics detected by WikiLabels such as `Afghan War documents6http://www.mahout.apache.org7http://www.journalism.org/about_news_index/methodology8PEJ does not provide ...
... on a par with manual labeling.7. REFERENCES[1] J. Allen, editor. Topic Detection and Tracking: Event-basedInformation Organization. Springer, 2002.[2] D. Carmel, H. Roitman, ...