Sequential latent semantic indexing

Krivenko, Mikhail; Vasilyev, Vitaly

doi:10.1145/1581114.1581117

Cited by 3 publications

(4 citation statements)

References 5 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Statistical approaches use only the given textual context without any additional meta-information. Various models have been already investigated in this direction: topic modeling (Prabhakaran et al, 2016;Uban et al, 2021;Krivenko and Vasilyev, 2009), clustering approaches (Mei and Zhai, 2005;Behpour et al, 2021), and so forth. Among the aforementioned models a sequential variant of LSI (Krivenko and Vasilyev, 2009) is the approach most similar to ours in terms of problem formulation.…”

Section: Related Workmentioning

confidence: 99%

“…Various models have been already investigated in this direction: topic modeling (Prabhakaran et al, 2016;Uban et al, 2021;Krivenko and Vasilyev, 2009), clustering approaches (Mei and Zhai, 2005;Behpour et al, 2021), and so forth. Among the aforementioned models a sequential variant of LSI (Krivenko and Vasilyev, 2009) is the approach most similar to ours in terms of problem formulation. Apart from that, other models utilize information from knowledge bases like the web (Roy et al, 2002) or citation graphs (Erten et al, 2004;Chang and Blei, 2010).…”

Section: Related Workmentioning

confidence: 99%

See 1 more Smart Citation

Incremental Topic Modeling for Scientific Trend Topics Extraction

Gerasimenko,

Chernyavskiy,

Nikiforova

et al. 2023

Computational Linguistics and Intellectual Technologies”

View full text Add to dashboard Cite

Rapid growth of scientific publications and intensive emergence of new directions and approaches poses a challenge to the scientific community to identify trends in a timely and automatic manner. We denote trend as a semantically homogeneous theme that is characterized by a lexical kernel steadily evolving in time and a sharp, often exponential, increase in the number of publications. In this paper, we investigate recent topic modeling approaches to accurately extract trending topics at an early stage. In particular, we customize the standard ARTM-based approach and propose a novel incremental training technique which helps the model to operate on data in real-time. We further create the Artificial Intelligence Trends Dataset (AITD) that contains a collection of early-stage articles and a set of key collocations for each trend. The conducted experiments demonstrate that the suggested ARTM-based approach outperforms the classic PLSA, LDA models and a neural approach based on BERT representations. Our models and dataset are open for research purposes.

show abstract

Section: Related Workmentioning

confidence: 99%

Section: Related Workmentioning

confidence: 99%

Incremental Topic Modeling for Scientific Trend Topics Extraction

Gerasimenko,

Chernyavskiy,

Nikiforova

et al. 2023

Computational Linguistics and Intellectual Technologies”

View full text Add to dashboard Cite

show abstract

“…At first all terms within the text documents can be built. Once a term by document matrix is made, SVD is employed to decompose the term by document matrix to construct a linguistics vector space which can be used totrepresenttconceptualttermdocumenttassociation [25].  Principle component Analysis: PCA it is a successful and important technique as it has the ability to reduce the size of data via converted the main attribute space into smaller spaces.…”

Section:  Latent Semantic Indexing Technique: Latentmentioning

confidence: 99%

Support Vector Machine and Back Propagation Neural Network Approach for Text Classification

Saad¹,

Shaker²

2017

JUHD

View full text Add to dashboard Cite

Abstract-Text classification is the process of inserting text into one or additional categories. Text categorization has many of significant application, Mostly in the field of organization, and for browsing within great groups of document. It is sometimes completed by means of "machine learning.". Since the system is built based on a wide range of document features."Feature selection." is an important approach within this process, since there are typically several thousand possible features terms. Within text categorization, The target goal of features selection is to improve the efficiency of procedures and reliability of classification by deleting features that have no relevance and non-essential terms. While keeping terms which hold enough data that facilitate with the classification task. The target goal of this work is to increase the efficient text categorization models. Within the "text mining" algorithms, a document is appearing as "vector" whose dimension is that the range of special keywords in it, which can be very large. Classic document categorization may be computationally costly. Therefore, feature extraction through the singular valued decomposition is employed for decrease the dimensionality of the documents, we are applying classification algorithms based on "Back propagation" and "Support Vector Machine." methodology. before the classification we applied "Principle Component Analysis." technique in order to improve the result accuracy . We then compared the performance of these two algorithms via computing standard precision and recall for the documentscollection.

show abstract

“…Latent semantic Analysis uses singular value decomposition (SVD) technique in which large term-document matrix is changed into a set of k orthogonal factors so that the unique textual data is converted to a reduced semantic space. Also, new document vectors and new coordinates of the queries are calculated and compared in smaller k-dimensional space [4].…”

Section: Introductionmentioning

confidence: 99%

Text Categorization Comparison between Simple BPNN and Combinatorial Method of LSI and BPNN

Tekwani¹,

Motwani²

2014

IJCA

View full text Add to dashboard Cite

This paper proposed a text categorization comparison between simple BPNN and Combinatorial method of LSI and BPNN. In the traditional error back propagation network, the weight adjustment process gets block in a local minima and also the training speed of such network is very slow which leads to reduced performance and reduced efficiency of the network. Also the Learning time of overall network is very high.Hence, to improve the categorization accuracy, a new combinatorial method of LSI (latent semantic Indexing) and BPNN (back propagation neural network) is proposed. The latent semantics demonstration is an accurate data structure in low-dimensional space in which documents, terms and queries are rooted and also compared. Singular value decomposition (SVD) technique is used in Latent semantic Analysis in which large term-document matrix is decomposed into a set of k orthogonal factors by which the original textual data is changed to a smaller semantic space. New document vectors are found in reduced k-dimensional space. Also new coordinates of the queries are found. Here we implement combinatorial method of LSI and BPNN based technique for the classification of 20Newsgroup dataset which include categories of Sports, CS, and Medicine. The proposed technique implemented is compared with the existing BPNN technique. . Hence, this new method greatly reduces the dimension and better classification results can be achieved.

show abstract

Sequential latent semantic indexing

Cited by 3 publications

References 5 publications

Incremental Topic Modeling for Scientific Trend Topics Extraction

Incremental Topic Modeling for Scientific Trend Topics Extraction

Support Vector Machine and Back Propagation Neural Network Approach for Text Classification

Text Categorization Comparison between Simple BPNN and Combinatorial Method of LSI and BPNN

Contact Info

Product

Resources

About