Topic Detection by Clustering Keywords

Wartena, Christian; Brussee, Rogier

doi:10.1109/dexa.2008.120

Cited by 101 publications

(51 citation statements)

References 9 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…Most labels used as queries in this task are given in plural form and are reduced to singular form for matching with the base forms. More details on the model can be found in [38,39]. Assuming that q is a term like other terms, we also call this distribution the co-occurrence distribution of q.…”

Section: Algorithm and Resultsmentioning

confidence: 99%

Automatic tagging and geotagging in video collections and communities

Larson

Soleymani

Serdyukov

et al. 2011

Proceedings of the 1st ACM International Conference on Multimedia Retrieval

View full text Add to dashboard Cite

Automatically generated tags and geotags hold great promise to improve access to video collections and online communities. We overview three tasks offered in the MediaEval 2010 benchmarking initiative, for each, describing its use scenario, definition and the data set released. For each task, a reference algorithm is presented that was used within MediaEval 2010 and comments are included on lessons learned. The Tagging Task, Professional involves automatically matching episodes in a collection of Dutch television with subject labels drawn from the keyword thesaurus used by the archive staff. The Tagging Task, Wild Wild Web involves automatically predicting the tags that are assigned by users to their online videos. Finally, the Placing Task requires automatically assigning geo-coordinates to videos. The specification of each task admits the use of the full range of available information including user-generated metadata, speech recognition transcripts, audio, and visual features.

show abstract

Section: Algorithm and Resultsmentioning

confidence: 99%

Automatic tagging and geotagging in video collections and communities

Larson

Soleymani

Serdyukov

et al. 2011

Proceedings of the 1st ACM International Conference on Multimedia Retrieval

View full text Add to dashboard Cite

show abstract

“…Pada penelitian [6], jumlah lonjakan kata kunci dideteksi dengan cara menghitung frekuensi kemunculan suatu kata kunci dengan pembobotan Term Frequency-Inverse Document Frequency (TF-IDF) dan juga algoritma n-gram. Lalu pada penelitian [7], topik dideteksi dengan menghitung probabilitas kemunculan bersama suatu kata kunci dengan metode Jensen-Shannon Divergence dan mengelompokan kata kunci tersebut dengan algoritma K-Means.…”

Section: Penelitian Terkaitunclassified

Deteksi Topik Fashion Pada Twitter Dengan Latent Dirichlet Allocation

Al-khairi¹,

Wibisono²,

Putro³

2017

Preprint

View full text Add to dashboard Cite

Abstrak -Bagi orang-orang yang bergerak di bidang fashion mengetahui tren fashion adalah hal yang penting. Salah satu cara untuk mengetahui tren adalah dengan mendeteksi topik mengenai fashion yang dibicarakan di media sosial. Penelitian ini mengimplementasikan algoritma Latent Dirichlet Allocation untuk mendeteksi topik fashion di Twitter. Tweet yang didapat, diklasifikasi dengan metode Naive Bayes lalu dibersihkan dengan cara menghapus URL, simbol, angka dan merubah setiap kata menjadi huruf kecil. Tweet lalu dibentuk menjadi kumpulan kata dan dikelompokan dengan algoritma Latent Dirichlet Allocation. Berdasarkan hasil eksperimen, konfigurasi paramater 20 topik dengan 1000 iterasi memperoleh skor UMass terbaik dengan nilai -56.342, dan konfigurasi parameter 50 topik dengan 1000 iterasi memperoleh skor PMI terbaik dengan nilai 6.272.

show abstract

“…Christian Wartena and Rogier Brussee proposed the discovery of Topic by Clustering Keywords [7]. This approach consists of two steps.…”

Section: Related Workmentioning

confidence: 99%

“…This work aims at proposing and implementing a Topic updation using Testor theory [4] model for discovering and updating meaningful and concise labels for the dynamically updated clusters which are grouped using Semantic Similarity based Histogram based Incremental Document Clustering (SHC) [5] and Enhanced Similarity Histogram Clustering using Intra Centroid Vector Similarity (ESHC-IntraCVS) [6] based on Semantic-based similarity for the scientific literature documents and newsgroup dataset. This proposed technique is compared with the existing Topic Discovery by clustering keywords [7] proposed by Wartena and Brussee, (2008) and TF-IDF classifier [8] by Seymore and Rosenfeld (1997) method using F-measure and Purity as evaluation metrics.…”

Section: Introductionmentioning

confidence: 99%

Correlated Concept based Topic Updation Model for Dynamic Corpora

Jayabharathy¹,

Kanmani²,

Sivaranjani³

2014

IJCA

View full text Add to dashboard Cite

A rapid growth of documents available on the Internet, digital libraries, medical documents, news wires and other scientific document corpuses has motivated the researchers to propose many text mining techniques that help users to quickly retrieve trace and summarize the information in an effective way. Topic detection is one such technique which discovers precise, meaningful and concise labels for the formulated static document clusters. This technique helps the user to navigate and retrieve the needed information quickly and efficiently. Topic updation is the process of identifying and renewing the discovered labels whenever the document clusters are updated dynamically. This paper focuses on topic updation model based on Testor theory. The proposed work is experimented using 20newsgroup and scientific literature data set. The experimental results demonstrate that the proposed algorithm exhibit better performance, compared to the existing algorithms for topic detection.

show abstract

Topic Detection by Clustering Keywords

Cited by 101 publications

References 9 publications

Automatic tagging and geotagging in video collections and communities

Automatic tagging and geotagging in video collections and communities

Deteksi Topik Fashion Pada Twitter Dengan Latent Dirichlet Allocation

Correlated Concept based Topic Updation Model for Dynamic Corpora

Contact Info

Product

Resources

About