Latent Semantic Analysis Algorithm Research Articles

Evaluating LDA and LSA for Topic Modeling in the Indonesian Natural Disaster

Topic Modeling is a method for analyzing topics, documents, and articles in Natural Language Processing. The LDA (Latent Dirichlet Allocation) and LSA (Latent Semantic Analysis) algorithms are widely used in topic modelling. This study focuses on analyzing articles related to natural disasters in the Indonesian language. The dataset for this study was obtained through data scraping from Google News, which served as a container for several articles and online news sources. The research method is divided into several stages: dataset scraping, data preprocessing, topic modelling with LDA and LSA, model visualization, and model evaluation. The results of the research show that both algorithms can generate topics on datasets relevant to natural disasters in Indonesia such as floods, earthquakes, landslides, tsunamis and so on. From the evaluation results using Coherence Scores, it is found that LDA has better performance than LSA seen from Coherence Scores in modeling topics related to Natural Disasters in Indonesia. The results of this research are expected to provide convenience for information system developers in terms of monitoring natural disasters in Indonesia and make it easier for researchers who want to effectively read a document based on the disaster related in the document.

The method of grouping and classifying music curriculum teaching resources in the context of double reduction

Abstract A subsumption classification method is proposed to improve the classification accuracy of teaching resources in the music curriculum through a double reduction policy. Subject words are selected from high-similarity and high-frequency word sets, and a subject word tree is constructed using an automatic tree construction method based on a probabilistic latent semantic analysis algorithm. To complete the subsumption classification of teaching resources, an improved multi-graph kernel convolutional network is employed to group tree leaf nodes. According to the classification evaluation results, the recall, accuracy, and F1 values are 90%, 96.48%, and 88.81%, respectively, and the macro F1 value is as high as 81.59%. It can be seen that the method can effectively classify the teaching resources of music courses with the best effect of subsumption classification, which helps to improve the appropriateness and adequacy of teaching resources utilization.

Effective News Text Summarization Techniques

n proposed work, we successfully implemented a news text summarization system using Natural Language Processing (NLP) techniques and the Latent Semantic Analysis (LSA) algorithm. The purpose of our project was to extract important information from a large volume of news articles and present it in a concise and easily understandable manner. To achieve this, we utilized the LSA algorithm, which is known for its ability to capture the underlying semantic structure of text. LSA employs a mathematical model to analyse relationships between words in a document, creating a semantic representation where words with similar contexts are grouped together in a vector space. The LSA-based summarization process involved several steps. First, we pre-processed the news articles by removing stop words, punctuation, and other non-relevant elements. Then, we constructed a term-document matrix, where rows represented words and columns represented documents, with matrix values representing word frequencies. Next, we applied Singular Value Decomposition (SVD) to the term-document matrix. SVD helped reduce the matrix's dimensionality by identifying the most important latent semantic concepts. This resulted in a lower-dimensional representation that captured the essential information. Finally, we identified the most important sentences in the news articles by measuring the cosine similarity between each sentence and the summary. Sentences with the highest cosine similarity scores were selected as summary sentences. The proposed system demonstrated the effectiveness of the LSA algorithm for news text summarization. By capturing the semantic structure of the text, it generated summaries that allowed users to understand the key points of a news article quickly and easily. Our implementation had practical applications for content recommendation systems, news aggregation platforms, and personalized news feeds. However, it is important to acknowledge the limitations of the LSA algorithm. It may struggle with handling idiomatic expressions and can be sensitive to the quality of the input data. These considerations highlight the need for ongoing research and development to enhance the performance and robustness of news text summarization systems.

New perspective of learning objects in e-learning system

PurposeTogether, learning objects (LOs) and e-pedagogical practices have the potential to improve the performance of e-learning systems in several ways. They can make e-learning more personalised and adaptable, providing students with a more engaging and effective learning experience.Design/methodology/approachThe development of LOs and e-pedagogical practices have significantly influenced and changed the performance of e-learning systems. LOs are self-contained, reusable units of instructional content that create instructional materials, such as online courses, tutorials and assessments. They provide a flexible and modular approach to designing and delivering e-learning content, allowing educators to easily customise and adapt their materials to the needs of their students. e-pedagogical practices refer to the use of technology to enhance and support the teaching and learning process. They include strategies such as online collaboration, gamification and adaptive learning to improve student engagement, motivation and achievement.FindingsTo achieve this objective, this study consists of two main phases. First, the authors extract metadata from LOs using latent semantic analysis algorithms, which are considered a strong tool in web-mining exploration techniques. Second, they identify LOs according to a particular form of similarity using fuzzy c-means (FCM) algorithms. To improve classification accuracy, the FCM is used as a clustering algorithm.Originality/valueFinally, in order to assess the effectiveness of LOs with FCM, a series of experimental studies using a real-world dataset are conducted. The results of this study indicate that the proposed approach exceeds the traditional approach and produces good results.

Discussions about the Russian Revolution of 1917 on the Web: methodological approaches to the study of web forums as a historical source

The subject of the study is a methodology for analyzing the electronic content of social networks (forums) as a historical source. The discussion of the revolution of 1917 during the centenary of this historical event was used as a material for analysis. The aim of the study was to test approaches to the methodology of working with large arrays of online texts, and the possible combination of two approaches to working with online texts - quantitative analysis tools (distant reading) and traditional methods of working with historical text (slow reading). As part of the "distant reading", thematic modeling is used using the LDA (latent Dirichlet placement) and LSA (latent semantic analysis) algorithm in the R programming environment in the R studio program (version 4.2.1). During the "slow reading" we analyze the entire volume of the text directly.The novelty of the research lies in the application of thematic modeling to sources in the R programming environment in conjunction with classical methods of analyzing historical texts. Within the framework of the study, a methodology for analyzing the content of social networks (forums) has been tested, focused on substantial arrays of text that are physically impossible to read in full or at least in a significant part, using exclusively traditional means of interaction of the researcher with the corpus of sources. A step-by-step research algorithm is proposed, in which the researcher needs to analyze the text by "distant reading" methods, identifying the topics of texts consisting of terms (words). Then, using these keywords, you should find the relevant text fragments in which the identified topic was discussed most actively, and analyze the fragments in more detail using traditional methods of working with a text source. A possible way to improve the quality of identifying topics necessary for the researcher in social networks and forums by the LDA algorithm is proposed, namely, preliminary splitting of a large text and subsequent analysis of fragments by the LDA method as separate documents.

TEXT SUMMARIZATION USING MACHINE LEARNING AND DEEP NEURAL NETWORK

The creation of summaries and their evaluation have grown in popularity over time. The two primary techniques for creating summaries from a given text corpus are extractive text summarization and abstractive text summarization. To create abstractive and extractive summaries, a variety of techniques and methodologies are available. Using machine learning and deep neural networks, an automatic extractive and abstractive text summarization technique is developed. To produce a summary from a given text, the Extractive model utilizes (Spacy, TF-IDF Vectorizer, and Latent semantic Analysis algorithm (LSA)) and the Abstract model employs (Encoder-decoder RNN with LSTM units and attention mechanism). The goal of the model is to demonstrate the benefits of applying the Extractive and Abstractive approach to text summarization in practical settings.

Cross-Domain Aspect Detection and Categorization using Machine Learning for Aspect-based Opinion Mining

• There is an increase in usage of e-commerce reviews. • Users find difficulty in getting appropriate information from huge reviews. • ABOM helps users in this regard. • One of the key elements of ABOM is Aspect extraction. • Unsupervised Machine Learning approach has been used to extract aspects. There is an increase in the development of social media and electronic commerce sites day by day. In order to express their opinions about the products purchased user's write comments, messages and reviews. The reviews present in the e-commerce sites are also increasing. Users find difficulty in getting appropriate information about the right topic from this large data. Aspect-based Opinion Mining (ABOM) helps users in this regard. In many real-world applications ABOM is used to get the details about the aspects of entities, where the opinion is expressed for those aspects and entities. One of the key elements of ABOM is Aspect extraction. Unsupervised Machine Learning approach has been used to extract aspects from the reviews as it does not require pre-labelled data. In this regard Latent Dirichlet Allocation (LDA) and Latent Semantic Analysis (LSA) are two most commonly used unsupervised Topic Modeling approaches. The topics are extracted from three different datasets such as Amazon Mobile Reviews, Hotel Reviews and IMDb Movie Reviews using LDA and LSA algorithms. These extracted topics are aspects of our interest. The results of topic modeling algorithms are quite difficult to be interpreted by the common user. The different visualization methods are used to display the results of topic modeling algorithms in an interactive way. Two different multi-class classifiers such as Multinomial Naive Bayes (MNB) and Support Vector Machine (SVM) have been constructed for aspect categorization. These classifiers are evaluated by considering the evaluation measures such as Precision, Recall and F1 score. As a result, SVM classifier has good performance than MNB classifier for aspect categorization task of aspect-based opinion mining.

Content Classification based-on Latent Semantic Analysis and Support Vector Machine (LSA-SVM)

The diversity of the content of a web page can have a negative impact if used by the wrong user. Almost a half of internet users are children. Therefore, it is important to classify web pages to find out which pages are worthy of being seen by children and that are not feasible. One method that can be used is the Support Vector Machine (SVM) algorithm. SVM is a binary classification whose working principle is to find the best hyperplane to separate the two classes. To obtain better classification accuracy, the SVM is combined with the Latent Semantic Analysis (LSA) algorithm. The data used in this study were taken from the DMOZ web data which has been classified into two categories. The data is then entered into the pre-processing stage for further feature extraction using LSA. The LSA algorithm is used to find out the semantic similarities of words and text contained in web pages. The results of feature extraction are then classified using SVM with RBF kernel. Based on the testing result, we obtain a classification accuracy of 64%.

Semantic analysis of Arab leaders on social media

ABSTRACT Social media has led to the creation of complex networks of political and social leaders, and has allowed for instantaneous interaction with broad audiences. Semantic analysis of these interactions is essential for policy-makers and various experts, such as in the anti-terrorism community. However, comprehensive quantitative analysis of the attitudes of Arab political and religious leaders is currently limited. To handle this problem, we first create a novel dataset consisting of 1,145,525 tweets posted in 2009–2018 from 449 political, social, and religious leaders from 12 Arab countries. Next, to semantically process various Arabic dialects in their informal expressions on social media, we develop a Latent Semantic Analysis algorithm that is then used to identify the individuals with the highest semantic matches to our pre-specified categories of interest. Finally, we construct an original Temporal Semantic Similarity (TSS) measure that allows the tracking of those expressions over time. Experimental results demonstrate that the proposed semantic approach can effectively and efficiently process documents in each category. We have made all the data, code and results publicly available on the project homepage: https://github.com/ArabLeader.

Email thread sentiment sequence identification using PLSA clustering algorithm

Email messaging is the most common way of providing effective communication between internauts. Consequently, the total sent and received emails count will be increased. But, the internaut can't remember all such emails. Even though email thread identification approaches give satisfactory benefits to the internauts, but they may fail to alert them for a cause to identify the sentiments behind an email thread. To address, this issue Probabilistic Latent Semantic Analysis clustering algorithm has been used in this paper to identify the email sentiment thread sequence. The sentiment and the thread sequence within the emails have been discovered as clustering sentiment polarity and temporal categories with the help of PLSA clusters. At the initial stage, we used three feature extraction methods, latent semantic analysis (LDA), bag of words (BoW), TF-IDF and SentiWordNet (SWN) lexicon for generating sentiment features of email. Next, Probabilistic Latent Semantic Analysis algorithm is used to form email clusters based on sentiment features. Thus, it helps to identify thread sentiment and sequence of sentiment threads. Email threads give a mechanism by which any user will be able to find out the sequence in the thread on the basis of sentiment analysis of email related to a specific set of communication during a specific time period. Various parameters evaluation measures have been considered in this work to evaluate the proposed model such as accuracy, precision, recall and F-measure, and the proposed algorithm is compared with other standard algorithms. Furthermore, a statistical test has also been performed.

ИСПОЛЬЗОВАНИЕ ЛАТЕНТНО-СЕМАНТИЧЕСКОГО АНАЛИЗА ПРИ ПОДГОТОВКЕ ДАННЫХ ДЛЯ ИДЕНТИФИКАЦИИ АНОНИМНЫХ ПОЛЬЗОВАТЕЛЕЙ ПО ЦИФРОВЫМ ОТПЕЧАТКАМ

Digital fingerprints changings over time as a result of system, plugins, browsers, installation of various programs updates and fonts is a serious problem in the method of tracking and identifying users using a browser (Fingerprinting (FP) of a web browser). The set of parsed attributes can contain both metrical and categorical (mostly non-numeric) values, for example, parameters such as user-agent, webgl, canvas, etc. Considering this, it is required to pre-encode them for the convenience of further processing. For these purposes, artificial intelligence technologies, including the processing of text in natural languages NLP (Natural Language Processing), are widely used. The aim of the research is to analyze the peculiarities of the implementation of latent-semantic analysis (LSA) in the preparation and analysis of FP data for the identification of anonymous users. Methods. A comparative analysis of the common ways of converting categorical values of fingerprint attributes (FP) into numeric One-Hot-Encoding, Label-Encoder, LSA for identifying anonymous users with a predetermined number of possible values of categorical features is carried out. Results. The advantage of the LSA algorithm over One-Hot-Encoding, Label-Encoder is shown. The possibility of clustering implementation within the framework of the user identification problem by visualizing FP (FP) relative to hidden semantic topics using the LSA model of latent semantic analysis is shown. It is shown that with a small number of hid& den topics using the obtained vectors of objects and vectors of terms for assessing the similarity of two FPs, the proposed model allows us to confidently classify the input FP to a common topic. With the help of the obtained vectors of objects and vectors of terms for assessing the similarity of two FP objects, it becomes possible to apply various measures of cluster proximity: Euclidean distance, cosine measure, etc.

Otomatisasi Peringkasan Teks Pada Dokumen Hukum Menggunakan Metode Latent Semantic Analysis

Pada masa sekarang, jumlah kriminalitas di Indonesia cukup banyak. Banyaknya jumlah kriminalitas di Indonesia akan berdampak pada jumlah dokumen hukum yang akan ditangani oleh aparat penegak hukum. Dalam pemahaman dokumen hukum, aparat penegak hukum seperti advokat, hakim, maupun jaksa harus membaca keseluruhan dokumen dimana akan memerlukan waktu yang cukup lama. Oleh karena itu dibutuhkan peringkasan agar aparat penegak hukum lebih mudah memahaminya. Sehingga salah satu solusi yang dibutuhkan yaitu dengan membuat sebuah ringkasan pada dokumen hukum dimana dokumen tersebut berbentuk PDF. Dalam hal meringkas teks, metode yang dapat digunakan yaitu algoritma Latent Semantic Analysis. Algoritma tersebut digunakan untuk menguraikan atau menganalisa makna yang masih tersembunyi dari suatu bahasa, kode atau jenis representasi lainnya, guna memperoleh informasi yang penting.
 Dari pengujian terhadap 10 dokumen yang diringkas oleh pakar, diperoleh hasil precision, recall, f-measure dan accuracy secara berurutan pada peringkasan teks otomatis dengan metode Latent Semantic Analysis untuk compression rate 75% yaitu 53%, 27%, 35% dan 71% lalu untuk compression rate 50% yaitu 54%, 56%, 55% dan 75%, serta untuk compression rate 25% yaitu 51%, 79%, 61% dan 75%. Berdasarkan hasil penelitian dan pengujian yang telah dilakukan dapat ditarik kesimpulan bahwa Metode Latent Semantic Analysis dapat digunakan untuk melakukan peringkasan dokumen hukum.

Latent Semantic Analysis in Automatic Text Summarization: A state of the art analysis

Increasing availability of information in the web and its ease of access necessitate the need for efficient and effective automatic text summarisation. Automatic text summarisation condenses the source text (a single document or multiple documents) into a compact version preserving its overall meaning and information content. Till now, researchers have employed different approaches for creating well-formed summaries. One of the most popular methods is the latent semantic analysis (LSA). In this paper, various prominent works to produce extractive and abstractive text summaries based on different variants of LSA algorithm are analysed.

Development of Cross Language Clone Detector for C, C & Java Repositories using Natural Language Processing

Reusing the code with or without modification is common process in building all the large codebases of system software like Linux, gcc , and jdk. This process is referred to as software cloning or forking. Developers always find difficulty of bug fixes in porting large code base from one language to other native language during software porting. There exist many approaches in identifying software clones of same language that may not contribute for the developers involved in porting hence there is a need for cross language clone detector. This paper uses primary Natural Language Processing (NLP) approach using latent semantic analysis to find the cross language clones of other neighboring languages in terms of all 4 types of clones using latent semantic analysis algorithm that uses Singular value decomposition. It takes input as code(C, C++ or Java) and matches all the neighboring code clones in the static repository in terms of frequency of lines matched.

Cross the data desert: generating textual-visual summary on the evolutionary microblog stream

Effectively and efficiently summarizing social media is crucial and non-trivial to analyse social media. On social streams, events which are the main concept of semantic similar social messages, often bring us a firsthand story of daily news. However, to identify the valuable news, it is almost impossible to plough through millions of multi-modal messages one by one with traditional methods. Thus, it is urgent to summarize events with a few representative data samples on the streams. In this paper, we provide a vivid textual-visual media summarization approach for microblog streams, which exploits the incremental latent semantic analysis (LSA) of detected events. Firstly, with a novel weighting scheme for keyword relationship, we can detect and track daily sub-events on a keyword relation graph (WordGraph) of microblog streams effectively. Then, to summarize the stream with representative texts and images, we use cross-modal fusion to analyze the semantics of microblog texts and images incrementally and separately, with a novel incremental cross-modal LSA algorithm. The experimental results on a real microblog dataset show that our method is at least 1.31% better and 23.67% faster than existing state-of-the-art methods, and cross-modal fusion can improve the summarization performance by 4.16% on average.

Textual Analysis of Corporate Annual Disclosures: A Comparison between Bankrupt and Non-Bankrupt Companies

ABSTRACTIn this study we focus on extracting qualitative information from the management discussion and analysis (MD&A) section of an annual report and compare whether there are textually evident differences in textual expressions used between bankrupt and non-bankrupt companies. We extract high-frequency words, related concept links, and topics from MD&As and find that some high-frequency words appear to suggest differences between bankrupt and non-bankrupt companies regarding their financial position and ongoing status. However, the usefulness of concept links is mixed. Some concept links for high-frequency words do not seem to center around a theme or a key word, yet others provide some contextual information supporting our conjectures about the ongoing business status of non-bankrupt companies. Finally, we perform topic extraction based on a latent semantic analysis algorithm in order to investigate whether issues and themes discussed differ between non-bankrupt and bankrupt companies. We find that most of the top topics extracted merely recapture the characteristics of industries in which companies operate and do not provide information in differentiating between bankrupt and non-bankrupt companies. The reasons are discussed in the paper.

A comparison of automated training-by-example selection algorithms for Evidence Based Software Engineering

ContextStudy search and selection is central to conducting Evidence Based Software Engineering (EBSE) research, including Systematic Literature Reviews and Systematic Mapping Studies. Thus, selecting relevant studies and excluding irrelevant studies, is critical. Prior research argues that study selection is subject to researcher bias, and the time required to review and select relevant articles is a target for optimization. ObjectiveThis research proposes two training-by-example classifiers that are computationally simple, do not require extensive training or tuning, ensure inclusion/exclusion consistency, and reduce researcher study selection time: one based on Vector Space Models (VSM), and a second based on Latent Semantic Analysis (LSA). MethodAlgorithm evaluation is accomplished through Monte-Carlo Cross-Validation simulations, in which study subsets are randomly chosen from the corpus for training, with the remainder classified by the algorithm. The classification results are then assessed for recall (a measure of completeness), precision (a measure of exactness) and researcher efficiency savings (reduced proportion of corpus studies requiring manual review as a result of algorithm use). A second smaller simulation is conducted for external validation. Results and conclusionsVSM algorithms perform better in recall; LSA algorithms perform better in precision. Recall improves with larger training sets with a higher proportion of truly relevant studies. Precision improves with training sets with a higher portion of irrelevant studies, without a significant impact from the training set size. The algorithms reduce the influence of researcher bias and are found to significantly improve researcher efficiency.To improve recall, the findings recommend VSM and a large training set including as many truly relevant studies as possible. If precision and efficiency are most critical, the findings suggest LSA and a training set including a large proportion of truly irrelevant studies.

An index-based algorithm for fast on-line query processing of latent semantic analysis.

Latent Semantic Analysis (LSA) is widely used for finding the documents whose semantic is similar to the query of keywords. Although LSA yield promising similar results, the existing LSA algorithms involve lots of unnecessary operations in similarity computation and candidate check during on-line query processing, which is expensive in terms of time cost and cannot efficiently response the query request especially when the dataset becomes large. In this paper, we study the efficiency problem of on-line query processing for LSA towards efficiently searching the similar documents to a given query. We rewrite the similarity equation of LSA combined with an intermediate value called partial similarity that is stored in a designed index called partial index. For reducing the searching space, we give an approximate form of similarity equation, and then develop an efficient algorithm for building partial index, which skips the partial similarities lower than a given threshold θ. Based on partial index, we develop an efficient algorithm called ILSA for supporting fast on-line query processing. The given query is transformed into a pseudo document vector, and the similarities between query and candidate documents are computed by accumulating the partial similarities obtained from the index nodes corresponds to non-zero entries in the pseudo document vector. Compared to the LSA algorithm, ILSA reduces the time cost of on-line query processing by pruning the candidate documents that are not promising and skipping the operations that make little contribution to similarity scores. Extensive experiments through comparison with LSA have been done, which demonstrate the efficiency and effectiveness of our proposed algorithm.

Agentic, communal, and spiritual traits are related to the semantic representation of written narratives of positive and negative life events

We used a computational method to quantitatively investigate the relationship between personality and written narratives of life events. Agentic (i.e., self-directedness), communal (i.e., cooperativeness), and spiritual (self-transcendence) traits were of special interest because they represent individual differences in intentional values and goals, in contrast to temperament traits, which describe individual differences in automatic responses to emotional stimuli. We also investigated which pronouns were most common in relation to personality constructs that were significantly related to the narratives. Personality was assessed among 79 adolescents at one point in time using the NEO Personality Inventory—Revised (NEO-PI-R) and the temperament and character inventory (TCI). Six months later, adolescents were asked to write down the most positive or the most negative event that had happened to them in the last 3 months. Adolescents were explicitly instructed to answer the following questions within their narratives: What happened? Who were involved? Why do you think it happened? How did you feel when it happened? How do you think the involved persons felt? The descriptions were quantified using a computational method in which the latent semantic analysis algorithm generates a semantic representation of the narratives. Only self-directedness, cooperativeness, and self-transcendence were related to the semantic representation of the narratives. Moreover, cooperativeness and self-transcendence were associated with less frequent usage of singular pronouns (e.g., me respectively mine). Agentic, communal, and spiritual traits are involved when adolescents describe positive and negative life experiences. Moreover, high levels of communal and spiritual traits are related to less self-focused narratives.

Application of Refined LSA and MD5 Algorithms in Spam Filtering

The paper proposes a spam filtering method that uses integrated and refined Latent Semantic Analysis (LSA) and Message-Digest Algorithm 5 (MD5) algorithms to address a series of universal problems in spam filtering, including remarkably lowered filtering precision and notably unbalanced filtering efficiency as a result of lack of latent semantic analysis of mail contents. In introducing LSA, its weighting function is improved by integrating fuzzy membership to improve effectiveness of LSA in processing mail contents. On top of this, MD5 algorithm is used to generate “E-mail fingerprint”, thus enabling quick matching and realizing highly efficient and accurate processing of mass- mailing spam. The result of the simulation experiment testifies effectiveness of the method.