Document representation methods for clustering bilingual documents

Abstract

Globalization places people in a multilingual environment. There is a growing number of users to access and share information in several languages for public or private purpose. In order to deliver relevant information in different languages, efficient multilingual documents management is worthy of study. Generally, classification and clustering are two typical methods for documents management. However, lack of training data and high efforts for corpus annotation will increase the cost for classifying multilingual documents which needs to bridge language gaps as well. Clustering is more suitable to implement in such practical applications. There are two main factors involved in documents clustering, document representation method and clustering algorithm. In this paper, we focus on document representation method and demonstrate that the choice of representation methods has impacts on quality of clustering results. In our experiment, we use parallel corpora (English-Chinese documents on topic of technology information) and comparable corpora (English and Chinese documents on topics of mobile technology and wind energy) as dataset. We compare four different types of document representation methods: Vector Space Model, Latent Semantic Indexing, Latent Dirichlet Allocation and Doc2Vec. Experimental results show that, accuracy of Vector Space Model were not competitive with other methods in all clustering tasks. Latent Semantic Indexing is overly sensitive to corpora itself, for it behaved differently when clustering two different topics of

A Neural Network Language Document Representation Technique for Web-Page Classification

“…Experimental results show that the method performs better than tf-idf with/without stops words and word2vec with/without stop words. A major drawback in their work is that, stops words increase the dimensionality of the feature vectors which impacts badly on the classification accuracy and computational burden [8].Also the classification algorithm used was a linear SVM, other kernels such as string and RBF kernels could produce better results [20].…”

“…To achieve high classification result of the Web Page Classification (WPC) system, an excellent representation of textual data (Preprocessing/DR) should contain as much information as possible from the original document [8]. Also, the accuracy of most classification algorithms depends on the quality and size of training data which is inherently dependent on the document representation technique [9].…”

“…Several researchers have contributed to the document representation stage of the web page classification system because irrelevant and redundant features often degrade the performance of the classification algorithms both in speed and classification accuracy and also its tendency to reduce overfitting [10]. Drawing from literature, state-of-the-art DR technique's used in WPC systems are: bag of words model, Term-Frequency Inverse Document Frequency (TF-IDF), Latent Semantic Indexing (LSI), Probabilistic Latent Semantic Indexing (PLSI), Latent Dirichlet Allocation (LDA), LSI and TF-IDF, N-Gram and TF-IDF, Word2Vec and TF-IDF, TF-IDF and firefly Algorithm, Word2Vec and LDA [8], [11], [12], [13], [14], [15]. Each of these technique are fraught with one challenge or the other such as semantic mismatch and multiple meanings of word and so on.…”

Comparing Attitudes to Climate Change in the Media using sentiment analysis based on Latent Dirichlet Allocation

“…In this paper, we therefore treat the topics as clusters, and apply the Silhouette Coefficient instead. This method has been previously used for finding the optimal number of topics (Panichella et al, 2013;Ma et al, 2016), and is suitable for our LDA approach, since LDA is fully unsupervised. Nevertheless, in future work, it may be worth evaluating some probability measures such as loglikelihood and perplexity, and comparing the performance using these methods.…”

“…In the silhouette analysis (Ma et al, 2016), silhouette coefficients close to +1 indicate that the samples in the cluster are far away from the neighbouring clusters. In contrast, a negative silhouette coefficient means that the samples might have been assigned to the wrong cluster.…”