Search

The demand for methods that automatically create text summaries from images containing many things has recently grown. Our research introduces a fresh and creative way to achieve this. We bring together the WordNet dictionary and the YOLO model to make this happen. YOLO helps us find where the things are in the images, while WordNet provides their meanings. Our process then crafts a summary for each object found. This new technique can have a big impact on computer vision and natural language processing. It can make understanding complicated images, filled with lots of things, much simpler. To test our approach, we used 1381 pictures from the Google Image search engine. Our results showed high accuracy, with 72% for object detection. The precision was 85%, the recall was 72%, and the F1-score was 74%. [ABSTRACT FROM AUTHOR]

Social media is a platform for sharing various hashtags, news, and posts within the community. Increasing text content and information to read and understand is difficult because of the vast definitions. At present, every business must analyze data in social media. Social media platforms are where people worldwide discuss and share social commentary. Automated text summarization is crucial for condensing lengthy contents into concise ones using learning concepts. The unstructured data in social media are significant phrases with support from sources for analyzing sentiments and extracting the importance of the content. Previously, all data were analyzed as related content similarity matches. However, the main drawback is that data needs to be examined to rephrase or summarize the essential terms of extraction. Due to improper content extraction, the accuracy gets poor in precision level to increase the false content rate. To tackle these problems, in this paper presents a Machine Learning (ML) intelligence algorithm with Diverse Beam Search-Based Maximum Mutual Information (DBSMMI) and Ant Colony Optimization (ACO)-optimized Deep Belief Network Based Text Summarization (DBNTS). Initially, the COVID-19 Twitter dataset is preprocessed to remove noise and create a Pheromone value set based on the k-means semantic similarity algorithm. Our work analyzes and clusters the data according to their theme (area). Data analysis is the central concept and is performed using WordNet keyword matching and semantic matching of the words. Then, the similar word is clustered using a semantic similarity-based k-means clustering algorithm. Using DBSMMI to make identical content phrases maximum supports term sentence extraction. The maximum support clustered group is optimized for the respective theme using ant colony optimization with the DBNTS algorithm. The algorithm's efficiency can be tested with an existing classifier algorithms. The ACO semantic recommender system is implemented in this article to recommend relevant news to the Twitter user. The proposed simulation attains the 92.35% of accuracy, and 90.29% of precision performance. The proposed method efficiently improves classification accuracy, and precision performance compared to other methods. [ABSTRACT FROM AUTHOR]

This paper proposes a hybrid model to improve Information Content (IC) related metrics of semantic similarity between words, named IC+SP, based on the essential hypothesis that IC and the shortest path are two relatively independent semantic evidences and have approximately equal influences to the semantic similarity metric. The paradigm of IC+SP is to linearly combine the IC-related metric and the shortest path. Meanwhile, a transformation from the semantic similarity of the concepts to that of the words is presented by maximizing every component of IC+SP. 13 improved IC-related metrics based on IC+SP are formed and implemented on the experimental platform HESML Lastra-Díaz (Inf Syst 66:97–118, 2017). Pearson's and Spearman's correlation coefficients on well-accepted benchmarks for the improved metrics compare to those for the original ones to evaluate IC+SP. I introduce the Wilcoxon Signed-Rank Test needing no standard distribution hypothesis, while, this hypothesis is required by T-Test on the sample of small size. T-Test, as well as the Wilcoxon Signed-Rank Test, conduct on the differences of the correlative coefficients for improved and original metrics. It is expected that the improved IC-related metrics could significantly outperform their corresponding original ones, and the experimental results, including the comparisons of mean and maximum of correlation coefficients as well as the p-value and confidence interval of both tests, accomplish the anticipation in the vast majority of cases. [ABSTRACT FROM AUTHOR]

Many applications of Natural Language Processing (NLP) like machine translation, document clustering, and information retrieval make use of Word Sense Disambiguation (WSD). WSD automatically predicts the sense of an ambiguous word that exactly fits it as per the given situation. While it may seem very easy for humans to interpret the meaning of natural language, machines require the processing of huge amounts of data for similar tasks. In this paper, we propose an automatic WSD system for the Kashmiri language based on the Naive Bayes classifier. This work is the first attempt towards developing a WSD system for the Kashmiri language to the best of our knowledge. Bag-of-Words (BoW) and Part-of-Speech (PoS) based features are used in this study for developing the WSD system. Experiments are carried out on a manually crafted sense-tagged dataset for 60 ambiguous Kashmiri words. These 60 words are selected based on the frequency in the raw corpus collected. Senses for annotation purposes of these ambiguous words are extracted from Kashmiri WordNet. The performance of the proposed system is measured using accuracy, precision, recall and F-1 measure metrics. The proposed WSD model reported the best performance (accuracy = 89.92, precision = 0.84, recall = 0.89, F-1 measure = 0.86) when both PoS and BoW features were used at the same time. [ABSTRACT FROM AUTHOR]

In the realm of natural language processing (NLP), Word Sense Disambiguation (WSD) is a crucial task, and WSD systems are used in many NLP applications. Systems built on WordNet (e.g., Lesk) have been witnessing encouraging progress in the domain of word sense disambiguation. Yet, the performance of WordNet based WSD systems may have limits when disambiguating polysemous words. The purpose of this research was to investigate the discrepancies between a systematic fusing approach of WordNet WSD and a single best performing system. In the experimental test, the fusing approach was used to disambiguate, and in the control test, a single best disambiguating system was used to disambiguate. The accuracies, recalls, and disambiguation times of two groups were compared after the two groups were tested by the same test dataset. The result of the experiment is that the performance accuracy and recall of the experimental group is better than that of the control group. The decision result of the multiple systems was fused to strengthen the performance accuracy and comprehensive of the system. At the disambiguation time, the experimental group showed a worthy disambiguation rate of disambiguation. [ABSTRACT FROM AUTHOR]

IntroductionSemantic relations are crucial in various cognitive processes, highlighting the need to understand concept interactions and how such relations are represented in the brain. Psycholinguistics research requires computationally grounded datasets that include word similarity measures controlled for the variables that play a significant role in lexical processing. This work presents a dataset for noun pairs in Basque and European Spanish based on two well-known Natural Language Processing resources: text corpora and knowledge bases.MethodsThe dataset creation consisted of three steps, (1) computing four key psycholinguistic features for each noun; concreteness, frequency, semantic, and phonological neighborhood density; (2) pairing nouns across these four variables; (3) for each noun pair, assigning three types of word similarity measurements, computed out of text, Wordnet and hybrid embeddings.ResultsA dataset of noun pairs in Basque and Spanish involving three types of word similarity measurements, along with four lexical features for each of the nouns in the pair, namely, word frequency, concreteness, and semantic and phonological neighbors. The selection of the nouns for each pair was controlled by the mentioned variables, which play a significant role in lexical processing. The dataset includes three similarity measurements, based on their embedding computation: semantic relatedness from text-based embeddings, pure similarity from Wordnet-based embeddings and both categorical and associative relations from hybrid embeddings.DiscussionThe present work covers an existent gap in Basque and Spanish in terms of the lack of datasets that include both word similarity and detailed lexical properties, which provides a more useful resource for psycholinguistics research in those languages.

Abstract This paper proposes a hybrid model to improve Information Content (IC) related metrics of semantic similarity between words, named IC+SP, based on the essential hypothesis that IC and the shortest path are two relatively independent semantic evidences and have approximately equal influences to the semantic similarity metric. The paradigm of IC+SP is to linearly combine the IC-related metric and the shortest path. Meanwhile, a transformation from the semantic similarity of the concepts to that of the words is presented by maximizing every component of IC+SP. 13 improved IC-related metrics based on IC+SP are formed and implemented on the experimental platform HESML Lastra-Díaz (Inf Syst 66:97–118, 2017). Pearson’s and Spearman’s correlation coefficients on well-accepted benchmarks for the improved metrics compare to those for the original ones to evaluate IC+SP. I introduce the Wilcoxon Signed-Rank Test needing no standard distribution hypothesis, while, this hypothesis is required by T-Test on the sample of small size. T-Test, as well as the Wilcoxon Signed-Rank Test, conduct on the differences of the correlative coefficients for improved and original metrics. It is expected that the improved IC-related metrics could significantly outperform their corresponding original ones, and the experimental results, including the comparisons of mean and maximum of correlation coefficients as well as the p-value and confidence interval of both tests, accomplish the anticipation in the vast majority of cases.

A recommender system is a model that automatically recommends some meaningful cases (such as clips/films/goods/items) to the clients/people/consumers/users according to their (previous) interests. These systems are expected to recommend the items according to the users' interests. There are two traditional general recommender system models, i.e., Collaborative Filtering Recommender System (ColFRS) and Content-based Filtering Recommender System (ConFRS). Also, there is another model that is a hybrid of those two traditional recommender systems; it is called Hybrid Recommender System (HRS). An HRS usually outperforms simple traditional recommender systems. The problems such as scalability, cold start, and sparsity belong to the main problems that any recommender system should solve. The memory-based (modeless) recommender systems benefit from good accuracies. But they suffer from a lack of admissible scalability. The model-based recommender systems suffer from a lack of admissible accuracies. But they benefit from good scalability. In this paper, it is tried to propose a hybrid model based on an automatically improved ontology to solve the scalability, cold start, and sparsity problems. Our proposed HRS also uses an innovative approach of clustering as an augmented section. When there are enough ratings, it uses a collaborative filtering approach to predict the missing ratings. When there are not enough ratings, it uses a content-based filtering approach to predict the missing ratings. In the content-based filtering section of our proposed HRS, ontology concepts are used to improve the accuracy of ratings' prediction. If our target client is severely sparse, we cannot trust even the ratings predicted by the content-based filtering section of our proposed HRS. Therefore, our proposed HRS uses additive clustering to improve the prediction of the missing ratings if the target client is severely sparse. It is experimentally shown that our model outperforms many of the newly developed recommender systems. [ABSTRACT FROM AUTHOR]

After musing on the history and varying media of the concept of 'gone viral', Associate Professor of Architecture at the University of California, Berkeley, Kyle Steinfeld further investigates computational design through the lens of cultural practices. Even the seemingly most contemporary and innovative technological ideas and gizmos can be traced back to a series of legacy notions that remain silently present in new advances. The article discusses such 'hinge' moments and searches for them in AI. [ABSTRACT FROM AUTHOR]

With tremendous evolution in the internet world, the internet has become a household thing. Internet users use search engines or personal assistants to request information from the internet. Search results are greatly dependent on the entered keywords. Casual users may enter a vague query due to lack of knowledge of the domain-specific words. We propose a query reformulation system that determines the context of the query, decides on keywords to be replaced and outputs a better-modified query. We propose strategies for keyword replacements and metrics for query betterment checks. We have found that if we project keywords into the vector space of word projection using word embedding techniques and if the keyword replacement is correct, clusters of a new set of keywords become more cohesive. This assumption forms the basis of our proposed work. To prove the effectiveness of the proposed system, we applied it to the ad-hoc retrieval tasks over two benchmark corpora viz TREC-CDS 2014 and OHSUMED corpus. We indexed Whoosh search engine on these corpora and evaluated based on the given queries provided along with the corpus. Experimental results show that the proposed techniques achieved 9 to 11% improvement in precision and recall scores. Using Google's popularity index, we also prove that the reformulated queries are not only more accurate but also more popular. The proposed system also applies to Conversational AI chatbots like ChatGPT, where users must rephrase their queries to obtain better results. [ABSTRACT FROM AUTHOR]

In this paper we present some modifications to and extensions of a Wordnet for Bulgarian designed to make it more appropriate for applications in the area of language learning. However, in order to support education, we need to ensure the appropriate selection of sets of synonyms (synsets) from BTB- Wordnet, depending on the education level of the learners, and various types of exercises based on integration of the learning topic and semantic information within Wordnet. For this purpose, our focus is mainly on the combination of the lexemes (lemmas), with their meanings and examples, and the specially designed pictures as illustrations of those meanings within the synsets. We report on our preliminary results.

In the era of pervasive internet use and the dominance of social networks, researchers face significant challenges in Persian text mining, including the scarcity of adequate datasets in Persian and the inefficiency of existing language models. This paper specifically tackles these challenges, aiming to amplify the efficiency of language models tailored to the Persian language. Focusing on enhancing the effectiveness of sentiment analysis, our approach employs an aspect-based methodology utilizing the ParsBERT model, augmented with a relevant lexicon. The study centers on sentiment analysis of user opinions extracted from the Persian website 'Digikala.' The experimental results not only highlight the proposed method's superior semantic capabilities but also showcase its efficiency gains with an accuracy of 88.2% and an F1 score of 61.7. The importance of enhancing language models in this context lies in their pivotal role in extracting nuanced sentiments from user-generated content, ultimately advancing the field of sentiment analysis in Persian text mining by increasing efficiency and accuracy.

Saraiki, being the fourth most widely spoken language in Pakistan and being used in some parts of India and Afghanistan, is of significant geographical, historical, and cultural importance. However, it remains neglected in terms of proper documentation and identification of its unique linguistic features. The current study is centered on identifying the lexico-semantic categories of Saraiki nouns and then developing their hierarchical relationships (Miller et al., 1993). This quantitative research is designed to contribute to the process of developing Saraiki WordNet and is related to Natural Language Processing (NLP). A corpus of 3 million words was developed on the basis of data collected from different genres of the Saraiki language, including newspapers, academic essays, literary texts, and religious books. Both expansion and merge approaches were used to analyze the data. A wordlist of 1500 most occurring nouns was extracted from the corpus using Antconc 3.4.4.0, followed by manual tagging in Microsoft Excel 2010. Resultantly, 39 most occurring nouns from the wordlist were used to develop 173 related synsets, and lexico-semantic relationships among these nouns were identified with the help of 30 hierarchies (Miller et al., 1993). This study is limited to areas like Bahawalpur, Multan, and Muzaffarabad. It would be a milestone for Saraiki language learners, SWN development, Saraiki lexical resources, online SL dictionaries, and a guide for researchers.

Specialized languages can activate different sets of semantic features when compared to general language or express concepts through different words according to the domain. The specialized lexicon, i.e., lexical units that denote more specific concepts and knowledge emerging from specific domains, however, co-exists with the common lexicon, i.e., the set of lexical units that denote concepts and knowledge shared by the average speakers, regardless of their specific training or expertise. Communication between specialists and non-specialists can show a big gap between language(s), and therefore lexical units, used by the two groups. However, quite often, semantic and conceptual overlapping between specialized and common lexical units occurs and, in many cases, the specialized and common units refer to close concepts or even point to the same reality. Considering the modeling of meaning in functional lexical resources, this paper puts forth a solution that links common and specialized lexica within the WordNet model framework. We propose a new relation expressing semantic proximity between common and specialized units and define the conditions for its establishment. Besides contributing to the observation and understanding of the process of knowledge specialization and its reflex on the lexicon, the proposed relation allows for the integration of specialized and non-specialized lexicons into a single database, contributing directly to improving communication in specialist/non-specialist contexts, such as teaching–learning situations or health professional-patient interactions, among many others, where code-switching is frequent and necessary. [ABSTRACT FROM AUTHOR]

The process of plagiarism detection is one of the challenges in revealing the originality of a document, especially in the fields of science and research. Natural language processing methods can recognize and determine the level of similarity between different documents. In this paper, we tackle the task of extrinsic plagiarism detection based on semantic and syntactic approaches. The objective is to identify segments of a document that show strong similarity with a group of reference documents dealing with the same topic. In this paper, we present our hybrid approach that implements semantic and syntactic features based on Latent Dirichlet Allocation (LDA) and Wu & Plamer algorithm. The proposed approach has been evaluated on a PAN13 public dataset with a total accuracy of 85%. [ABSTRACT FROM AUTHOR]

Recommender systems are widely used in many applications. They can be viewed as the predictor systems that are to suggest accurate and highly preferred items to consumers or clients. These systems can be considered to be information filtering systems. They counter some important challenges such as cold start (it means the absence of enough data for a new item to make accurate recommendations), scalability, and sparsity. The memory-based recommender systems have high accuracy but lack scalability. Also, the model-based systems are scalable but not accurate. Current recommender systems use hybrid methods to deal with the most important shortages of traditional filtering approaches. Current recommender systems are usually a hybrid of content-based filtering and collaborative filtering, and so on. In this paper, a hybrid recommender system is presented to meet the stated challenges, increase system performance and provide more accurate recommendations. This system uses both content-based filtering and collaborative filtering. In addition, using an automatically collected wordnet, we create an ontology that has been used in the content-based filtering section of our proposed approach. Furthermore, this framework applies KNN (k nearest neighbors) algorithm and clustering to improve its functionality. The proposed system is evaluated on a real benchmark. The experimentations show the proposed method has a better performance compared with the current superior related methods. The experimentations also show that our recommender system has desirable scalability compared with the state-of-the-art recommender systems. [ABSTRACT FROM AUTHOR]

Most mediation systems use a caching policy in order to overcome their performance challenges. One of the most widely adopted strategies is known as semantic caching. Semantic caches are called so because they store the descriptions of all submitted queries. Although they may seem to be based on semantics because of their name, this is not really the case. In fact, they actually compare the syntax of the cached queries to the syntax of the new query to retrieve responses from the cache. This can lead to significant delays, especially if multiple requests are stored in the cache. In this work, we propose a new semantic approach based on ontologies to compute the semantic similarity between two given queries, and we provide also a new algorithm to filter all regions of the cache that do not semantically cover a user query. In this way, the use of the cache would be optimal and fast at the same time, despite the large number of regions in the cache. In fact, only the most beneficial regions will be processed to retrieve data from the cache. [ABSTRACT FROM AUTHOR]

E-learning is inevitable during these pandemic days and most of the learners find it comfortable to learn online. However, the main challenge is to locate the appropriate data in line with the learner's requirement. Considering the necessity of this issue, this article presents an e-content recommendation system that considers the user's learning capability. This work categorizes the documents into three categories such as basic, intermediate and advanced levels. Based on the users' learning capability, corresponding documents are recommended and this idea enhances the overall learning experience. This work is based on three phases such as data pre-processing, feature extraction and classification. The collected documents are pre-processed for preparing the documents suitable for further processes. Features such as Parts–OF–Speech (POS) tagging, Term Frequency - Inverse Document Frequency (TF-IDF) and semantic similarity based on WordNet are extracted and the multiclass Support Vector Machine (SVM) is employed for distinguishing between the classes. The performance of the work is tested and the results prove the efficacy of the work with 98 % accuracy rates, in contrast to the comparative techniques.

When integrating data from different sources, there are problems of synonymy, different languages, and concepts of different granularity. This paper proposes a simple yet effective approach to evaluate the semantic similarity of short texts, especially keywords. The method is capable of matching keywords from different sources and languages by exploiting transformers and WordNet-based methods. Key features of the approach include its unsupervised pipeline, mitigation of the lack of context in keywords, scalability for large archives, support for multiple languages and real-world scenarios adaptation capabilities. The work aims to provide a versatile tool for different cultural heritage archives without requiring complex customization. The paper aims to explore different approaches to identifying similarities in 1- or n-gram tags, evaluate and compare different pre-trained language models, and define integrated methods to overcome limitations. Tests to validate the approach have been conducted using the QueryLab portal, a search engine for cultural heritage archives, to evaluate the proposed pipeline. [ABSTRACT FROM AUTHOR]

In recent years, researchers have proposed several feature-based methods to measure semantic similarity using knowledge resources like Wikipedia and WordNet. While Wikipedia covers millions of concepts with multiple features, it has some limitations such as articles with limited content and concept ambiguity. Disambiguating these concepts remains a challenge. Conversely, WordNet offers unambiguous terms by covering all possible senses, making it a useful resource for disambiguating Wikipedia concepts. Additionally, WordNet can enrich the limited content of Wikipedia articles. Thus, we present a new approach that combines both resources to enhance previous feature-based methods of semantic similarity. We begin by analyzing the limitations of previous research, followed by introducing a novel method to disambiguate Wikipedia concepts using WordNet's synonym structure, resulting in more effective disambiguation. Furthermore, we use WordNet to supplement the features in Wikipedia articles and redefine the feature similarity functions. Finally, we train non-linear fitting-based models to measure semantic similarity. Our approach outperforms other previous methods on various benchmarks. To further showcase our approach, we apply our models to develop a movie recommender system using the MovieLens dataset, which consistently outperforms other systems. [ABSTRACT FROM AUTHOR]

Word sense disambiguation (WSD) remains a persistent challenge in the natural language processing (NLP) community. While various NLP packages exist, the Lesk algorithm in the NLTK library demonstrates suboptimal accuracy. In this research article, we propose an innovative methodology and an open-source framework that effectively addresses the challenges of WSD by optimizing memory usage without compromising accuracy. Our system seamlessly integrates WSD into NLP tasks, offering functionality similar to that provided by the NLTK library. However, we go beyond the existing approaches by introducing a novel idea related to WSD. Specifically, we leverage deep neural networks and consider the language patterns learned by these models as the new gold standard. This approach suggests modifying existing semantic dictionaries, such as WordNet, to align with these patterns. Empirical validation through a series of experiments confirmed the effectiveness of our proposed method, achieving state-of-the-art performance across multiple WSD datasets. Notably, our system does not require the installation of additional software beyond the well-known Python libraries. The classification model is saved in a readily usable text format, and the entire framework (model and data) is publicly available on GitHub for the NLP research community.

An emerging application domain concerning content-based recommender systems provides a better consideration of the semantics behind textual descriptions. Traditional approaches often miss relevant information due to their sole focus on syntax. However, the Semantic Web community has enriched resources with cultural and linguistic background knowledge, offering new standards for word categorization. This paper proposes a framework that combines the information extractor ReVerb with the WordNet taxonomy to monitor global semantic relatedness scores. Additionally, an experimental validation confronts human-based semantic relatedness scores with theoretical ones, employing Mann–Kendall trend tests to reveal fairness and biases. Overall, our framework introduces a novel approach to semantic relatedness monitoring by providing valuable insights into fairness and biases.

The proposed work models a novel plagiarism detection system based on the semantic features to uncover the cases of plagiarism. The system constructs the dynamic relation matrix for each suspicious and source sentence pair to measure the degree of similarity using semantic features. Two Weighted Inverse Distance and GlossDice procedures show several text properties (synonyms, shortest path, etc.) to overcome the limitations of the existing features and new similarity metric for plagiarism detection are presented in this paper. Moreover, this research investigates the independent performance of various features to detect plagiarized cases and combine the best features by assigning different weight contributions to further enhance the system performance. Weighted Inverse Distance integrated with SynJaccard boosts the system performance and shows promising results. Initially, all the experiments were performed on PAN-PC-11dataset, and then PAN-14 text alignment dataset was used to validate the results of the proposed system. The effectiveness of the proposed system has been measured using standard performance measures i.e. Precision, Recall, F-measure, Granularity, and Plagdet score. The proposed system has outperformed the other baseline systems with precision (0.9459), recall (0.8861), f-measure (0.8917), and plagdet (0.8857) on the PAN-PC-11 dataset. For PAN-14 text alignment, the system exhibits precision (0.9257), recall (0.9055), f-measure (0.8931), and plagdet (0.8806). [ABSTRACT FROM AUTHOR]

In this survey, 60 research papers are reviewed based on various web data classification techniques, which are used for effective classification of web data and measuring the semantic relatedness between the two words. The web data classification techniques are classified into three types, such as semantic-based approach, search engine-based approach, and WordNet-based approach, and the research issues and challenges confronted by the existing techniques are reported in this survey. Moreover, the analysis is carried out based on the research works using the categorized web data classification techniques, dataset, and evaluation metrics are carried out. From the analysis, it is clear that semantic-based approach is the widely used techniques in the classification of web data. Similarly, Miller-Charles dataset is the most commonly used dataset in most of the research papers, and the evaluation metrics, like precision, recall, and F-measure are widely utilized in web data classification. The insights from this manuscript can be utilized to understand various research gaps and problems in this area. Those can be considered in the future by developing novel optimization algorithms, which might enhance the performance of web data classifications. [ABSTRACT FROM AUTHOR]

Many e-commerce websites currently provide online reviews to share e-shoppers' experience with the products. To help e-shoppers obtaining information efficiently, these websites usually summarize product information based on their certain predefined aspects. However, e-shopper's aspects should be annotated to make sure that more highly related information of online reviews can be fetched for fulfilling e-shopper's requirements. Hence, this study integrates an annotation approach with similarity techniques (Keyword pair similarity and Aspect-sentence similarity) to propose a new framework to fetch more highly correlated sentences for e-shoppers. Experimental results show that most of the combinations in the proposed approach have high prediction performance in the Top 10 sentences with Precision (0.90 or higher). [ABSTRACT FROM AUTHOR]

In this paper, we propose a method exploiting syntactic structure, semantic relations and word embeddings for recognizing textual entailment. The sentence pairs are analyzed using their syntactic structure and categorization of sentences in active voice, sentences in passive voice and sentences holding copular relations. The main syntactic relations such as subject, verb and object are extracted and lemmatized using a lemmatization algorithm based on parts-of-speech. The subject-to-subject, verb-to-verb and object-to-object similarity is identified using enhanced Wordnet semantic relations. Further similarity is analyzed using modifier relation, number relation, nominal modifier relation, compound relation, conjunction relation and negative relation. The experimental evaluation of the method on Stanford Natural Language Inference dataset shows that the accuracy of the method is 1.4% more when compared to the state-of-the-art zero shot domain adaptation methods. [ABSTRACT FROM AUTHOR]

Abstract This research paper presents a novel approach for recommending products to customers based on their cared aspects by performing sentiment analysis on customer feedback. The proposed approach utilizes the WordNet database to identify and extract aspects from customer reviews and feedback, and then applies sentiment analysis techniques to determine the sentiment associated with each aspect. The resulting sentiment scores are then used to generate personalized product recommendations that align with the customer’s preferences and priorities. Here we extract the comments from an e-commerce website that is Amazon, and we then choose the most cared aspects from those comments. The dataset is publicly available online which contains reviews of each product. The chosen most cared aspects are price, colour, battery, and screen. These cared aspects are keywords that shopping online and recommending, will help to categorize the comments based on price, colour, battery, and screen. After categorizing the comments, it will be defined as the set of explicit comments. After an explicit comment set is defined, sentiment analysis is performed to systematically identify the interest of the customer through comments. Here the comments are classified into the polarity of given texts in an explicit comment set into positive, negative, and neutral. Finally, scores were calculated for all brands which will help to recommend the product.

Abstract Opinion mining or sentiment analysis (SA) is a key component of real-world applications for e-commerce organizations, manufacturers, and customers. SA deals with the computational evaluation of people’s views, thoughts, and feelings in the text, whether they are visible or concealed. The Aspect based SA level is becoming one of the most active phases in this area. In this paper, we propose an approach to enrich K-Nearest Neighbors (KNN) to deal with Implicit Aspect Identification task (IAI). Through the use of WordNet semantic relations, we propose an enhancement for KNN distance computation to support the IAI task. For a conclusive empirical evaluation, experiments are conducted on two datasets of electronic products and restaurant reviews and the effects of our approach are examined and analyzed according to three criteria: KNN distance used to compute the similarity, the number of nearest neighbors (K) and the KNN behavior towards Overfitting and Underfitting. The experimental results show that our approach helps KNN improve its performance and better deal with Overfitting and Underfitting for Implicit Aspect Identification.

This research paper presents a novel approach for recommending products to customers based on their cared aspects by performing sentiment analysis on customer feedback. The proposed approach utilizes the WordNet database to identify and extract aspects from customer reviews and feedback, and then applies sentiment analysis techniques to determine the sentiment associated with each aspect. The resulting sentiment scores are then used to generate personalized product recommendations that align with the customer's preferences and priorities. Here we extract the comments from an e-commerce website that is Amazon, and we then choose the most cared aspects from those comments. The dataset is publicly available online which contains reviews of each product. The chosen most cared aspects are price, colour, battery, and screen. These cared aspects are keywords that shopping online and recommending, will help to categorize the comments based on price, colour, battery, and screen. After categorizing the comments, it will be defined as the set of explicit comments. After an explicit comment set is defined, sentiment analysis is performed to systematically identify the interest of the customer through comments. Here the comments are classified into the polarity of given texts in an explicit comment set into positive, negative, and neutral. Finally, scores were calculated for all brands which will help to recommend the product. [ABSTRACT FROM AUTHOR]

Extracting hierarchical structure in graph data is becoming an important problem in fields such as natural language processing and developmental biology. Hierarchical structures can be extracted by embedding methods in non-Euclidean spaces, such as Poincaré embedding and Lorentz embedding, and it is now possible to learn efficient embedding by taking advantage of the structure of these spaces. In this study, we propose embedding into another type of metric space called a metric cone by learning an only one-dimensional coordinate variable added to the original vector space or a pre-trained embedding space. This allows for the extraction of hierarchical information while maintaining the properties of the pre-trained embedding. The metric cone is a one-dimensional extension of the original metric space and has the advantage that the curvature of the space can be easily adjusted by a parameter even when the coordinates of the original space are fixed. Through an extensive empirical evaluation we have corroborated the effectiveness of the proposed cone embedding model. In the case of randomly generated trees, cone embedding demonstrated superior performance in extracting hierarchical structures compared to existing techniques, particularly in high-dimensional settings. For WordNet embeddings, cone embedding exhibited a noteworthy correlation between the extracted hierarchical structures and human evaluation outcomes. [ABSTRACT FROM AUTHOR]

With the recent growth of the Internet, the volume of data has also increased. In particular, the increase in the amount of unstructured data makes it difficult to manage data. Classification is also needed in order to be able to use the data for various purposes. Since it is difficult to manually classify the ever-increasing volume data for the purpose of various types of analysis and evaluation, automatic classification methods are needed. In addition, the performance of imbalanced and multi-class classification is a challenging task. As the number of classes increases, so does the number of decision boundaries a learning algorithm has to solve. Therefore, in this paper, an improvement model is proposed using WordNet lexical ontology and BERT to perform deeper learning on the features of text, thereby improving the classification effect of the model. It was observed that classification success increased when using WordNet 11 general lexicographer files based on synthesis sets, syntactic categories, and logical groupings. WordNet was used for feature dimension reduction. In experimental studies, word embedding methods were used without dimension reduction. Afterwards, Random Forest (RF), Support Vector Machine (SVM) and Multi-Layer Perceptron (MLP) algorithms were employed to perform classification. These studies were then repeated with dimension reduction performed by WordNet. In addition to the machine learning model, experiments were also conducted with the pretrained BERT model with and without WordNet. The experimental results showed that, on an unstructured, seven-class, imbalanced dataset, the highest accuracy value of 93.77% was obtained when using our proposed model. [ABSTRACT FROM AUTHOR]

This study aimed to measure the semantic distance between Vietnamese nouns to determine their similarities and differences in terms of semantics. The study utilized a dataset of 1000 randomly selected noun pairs from the Vietnamese WordNet and applied different semantic distance measurement formulas, such as Path-length, Leacock & Chodorow, Wu & Palmer, Resnik, Lin, and Jiang-Conrath measures. The results showed that the Resnik measure is suitable for evaluating noun similarity among primitive concepts of the same class and for measuring proximity between two classes of primitive concepts. On the other hand, the Leacock & Chodorow measure is appropriate for measuring the semantic distance between nouns belonging to two distant classes of primitive concepts. The methods used in this study have various potential applications in different fields of linguistics, such as determining synonyms, identifying semantically similar words, defining lexical-semantic fields, and detecting plagiarism. [ABSTRACT FROM AUTHOR]

Specialized languages can activate different sets of semantic features when compared to general language or express concepts through different words according to the domain. The specialized lexicon, i.e., lexical units that denote more specific concepts and knowledge emerging from specific domains, however, co-exists with the common lexicon, i.e., the set of lexical units that denote concepts and knowledge shared by the average speakers, regardless of their specific training or expertise. Communication between specialists and non-specialists can show a big gap between language(s), and therefore lexical units, used by the two groups. However, quite often, semantic and conceptual overlapping between specialized and common lexical units occurs and, in many cases, the specialized and common units refer to close concepts or even point to the same reality. Considering the modeling of meaning in functional lexical resources, this paper puts forth a solution that links common and specialized lexica within the WordNet model framework. We propose a new relation expressing semantic proximity between common and specialized units and define the conditions for its establishment. Besides contributing to the observation and understanding of the process of knowledge specialization and its reflex on the lexicon, the proposed relation allows for the integration of specialized and non-specialized lexicons into a single database, contributing directly to improving communication in specialist/non-specialist contexts, such as teaching–learning situations or health professional-patient interactions, among many others, where code-switching is frequent and necessary.

The results of manual mapping of Polish plWordNet onto English Princeton WordNet revealed a number of gaps and mismatches between those interlinked lexical resources. Preliminary studies have shown that they embrace wordnet-specific and language-specific differences, and in this exploratory study we focus on the latter, also called lacunae. Capitalising on the system of equivalence types and features for linking wordnet senses (Rudnicka et al. 2019), we present a semi-automatic, rule-based diagnostic system developed specifically for systematic detection and classification of gaps and mismatches between wordnets. First, focusing on noun synsets, we aim to identify those network fragments that are the most prone to reveal lexical and referential gaps (Svensén 2009). Second, we attempt to identify areas in an interlinked Polish-English wordnet that require resource expansion or modification of the existing network of inter-lingual relations. [ABSTRACT FROM AUTHOR]

In this study, a new approach called Contextualized Deep SemSpace is proposed for intent detection. First, the synset vectors are determined by training the generalized SemSpace method with the WordNet 3.1 data. Then, each word in an intent dataset is transformed into a synset vector by a contextualized approach, and finally, the synset vectors are trained with a deep learning model using BLSTM. Since the proposed approach adapts the contextualized semantic vectors to the dataset with a deep learning model, it treats like one of contextualized deep embeddings like BERT, ELMo, and GPT-3 methods. In order to measure the success of the proposed approach, some experiments have been carried out on six well-known intent detection benchmark datasets (ATIS, Snips, Facebook, Ask Ubuntu, WebApp, and Chatbot). Although the dependence of its vocabulary on WordNet causes a serious number of out of vocabulary problems, results showed that the proposed approach is the most successful intent classifier in the literature. According to these results, it can be said that deep learning-based contextualized synset vectors can be used successfully in many problems. [ABSTRACT FROM AUTHOR]

Summary: Sentiment analysis (SA) is information retrieval and computational linguistics that expresses sentiment about a specific document. Recently, research on SA has received greater interests due to the vast expansion of internet data. The comments and reviews posted online creates difficulties to the researchers, because the comments posted are written in unstructured formats with informal expressions, languages and possibly mixed languages. Most of the works on SA have been focusing only on English language but with respect to Indian languages, few works are done, especially on Hindi, Telugu, Bengali, and so forth. SentiWordNet is a useful lexical resource for determining the sentiment of a document or piece of text. There is currently no lexical resource for the Marathi language used for sentiment analysis. Therefore, this proposed research presents SA in one of the languages of Indian origin, that is, Marathi. In this paper, a new method is proposed for the development of a Marathi SentiWordNet (M‐SWN), synset‐based expansion approach using Marathi WordNet, which consists of Positive, Negative, and Neutral Polarity score, mapped from Hindi‐SentiWordNet (H‐SWN). Finally, sentiment analysis of movie reviews is carried out using M‐SWN using Bi‐LSTM based SVM classifier model. [ABSTRACT FROM AUTHOR]

Automated paraphrasing is considered an important and emerging educational phenomenon that has a key role in academia. Paraphrasing tools facilitate students as well as professionals' work. However, there is a need to assess and evaluate the quality of automated paraphrasing by such paraphrasing tools. This study aims at evaluating the quality of automated paraphrasing carried out by one of such tools, that is, QuillBot, by exploring the lexical relatedness and differences in the original and the paraphrased text. Hence, the limitations of automated paraphrasing tools are highlighted. In this regard, QuillBot is used for paraphrasing as it is considered a standard paraphrasing tool by many. Automated paraphrasing of literary and nonliterary text carried out by QuillBot is used as data. AntConc, a corpus tool, is used for exploring the lexical features of the data. In addition, word relatedness is also computed from HSO measure through the WS4J demo, an online WordNet ontological tool. Differences were found at the lexical level between the original and paraphrased text. Automated paraphrasing carried out by QuillBot of the non-literary text is more aligned with the original text than that of the literary text. In the lexical category, mostly content words were substituted with synonymous words. The morphological structure of words was also modified. At times, these alterations would obscure the original meaning, while on other occasions, they would enhance and refine it.Therefore, automated paraphrasing should not be taken for granted instead, it should be subject to manual review and revisions. In addition to automated paraphrasing of literary text carried out by such tools, it should be rechecked and updated manually while standard tools, like QuillBot, could be relied upon for paraphrasing of non-literary text. [ABSTRACT FROM AUTHOR]

There has been a meteoric rise in the total amount of digital texts as a direct result of the proliferation of internet access. As a direct result of this, document clustering has evolved into a crucial method that must be used in order to successfully extract relevant information from big document collections. When employing the document clustering approach, documents are automatically sorted into groups whose members have a high degree of similarity to one another. These groups are created by applying the document clustering technique. Because they do not take into account the semantic linkages that exist between the texts, traditional clustering approaches are unable to provide an acceptable description of a collection of texts. This is because traditional clustering techniques. Document clusters, in which texts are ordered according to their meaning rather than their use of keywords, have been extensively utilized as a means of overcoming these challenges as a result of the incorporation of semantic information. This has been possible as a result of the fact that document clusters can group together related texts. In this investigation, we looked at a total of 27 distinct papers that were published over the previous five years and categorized the documents based on the semantic similarities that existed between the various pieces. A detailed literature evaluation is included to each and every one of the publications that were selected for further consideration. Comparative research is carried out on a wide variety of evaluation strategies, including as algorithms, similarity metrics, instruments, and processes. Following that, there is a drawn-out discussion that analyzes the similarities and differences between the activities. [ABSTRACT FROM AUTHOR]

In this article, we present research directly inspired by the Princeton WordNet lexical ontology project (Miller, Fellbaum), which was a response to the real need for ontologies corresponding to the natural conceptualization common to all language users, within a given natural language, or within a specific sublanguage. Lexical ontologies for a given language or language subsystem determined by the scope of communication needs turn out to be useful and even necessary for constructing formal models of linguistic competence and, consequently, for designing and implementing AI systems with linguistic communicative competence, both passive and active. An important milestone of the research program presented in this work is the acquisition of tools in the form of extensive lexical ontologies of a new type, referred to in this work as LexiconGrammar Verbnets. In the article, we refer to the works of authors such as: Alain Colmerauer, Charles Fillmore, Christiane Fellbaum, Gaston Gross, Maurice Gross, Thomas R. Gruber, Richard Kittredge, George A. Miller, Martha Palmer, Kazimierz Polański, and Piek Vossen. [ABSTRACT FROM AUTHOR]

In the area of natural language processing, measuring sentence similarity is an essential problem. Searching for semantic meaning in natural language is a related issue. The task of measuring sentence similarity is to find semantic symmetry in two sentences, not matter how they are arranged. It is important to measure the similarity of sentences accurately. To compute the similarity between sentences, existing methods have been constructed from approaches for large texts. Since these methods work in very high-dimensional spaces, they are inefficient, require human input, and are not flexible enough for some applications. In this study, we propose a hybrid method (HydMethod) which considers not only semantic information including lexical databases, word embeddings, and corpus statistics, but also implied word order information. With lexical databases, our method models human common sense knowledge, and that knowledge can then be adapted to be used in different domains with the incorporation of corpus statistics. Therefore, the methodology is applicable across several domains. As part of our experiments, we used two standard datasets - Pilot Short Text Semantic Similarity Benchmark and MS paraphrase - in order to demonstrate the efficacy of our proposed method. As a result, the proposed method outperforms the existing approaches when tested on these two datasets, giving the highest correlation value for both word and sentence similarity. Moreover, it achieves a maximum of 32% higher increase than only using word vector or WorldNet based methodology. With Rubenstein and Goodenough word & sentence pairs, our algorithm's similarity measure shows a high Pearson correlation coefficient of 0.8953.

Due to the advances made in recent years, embedding methods caused a significant increase in the accuracy of text or graph processing methods. Embedding methods exhibit a compact vector representation of the basic elements (words, synsets, nodes,..) of the underlying system to encode the semantic information between the elements. Of course, due to the polysemous nature of words, in some NLP tasks, the use of sense/synset embedding is better than word embedding. However, in the literature, the introduction of embedding for synsets has received less attention. Existing synset embedding methods have complex calculations to calculate synset embedding based on word embeddings or base upon a defined pairwise synset similarity. In this paper, considering the graphical structure of the WordNet and the high-level knowledge encoded in it, we will create a synset embedding directly from the WordNet graph and its synset relations. Node2Vec graph embedding is used to map nodes of this graph to a vector space. We evaluate the performance of different graph structures (e.g. weighted/weightless, directed/undirected graphs). Moreover, we propose a weighting strategy to weight different synset relation types in the resulting WordNet graph. Experimental results of evaluation of the proposed synset embedding on the task of measuring lexical semantic similarities shows that mean squared error of similarities for the proposed synset embedding method on MEM and WordSim353 datasets are 0.065 and 0.035, resp., which is better than the mean squared error of Word2Vec on these datasets, (0.073 and 0.045, resp.). Furthermore, we use the Pearson correlation and Spearman correlation to compare the performance of the proposed synset embedding method with the state-of-the-art ones. The obtained results show the efficiency of the proposed method on various datasets. .The spearman correlation of the SimLex999 is improved by 0.02, while it improves WordSim353 Pearson correlation by 0.14.

Emotion lexicons are useful in research across various disciplines, but the availability of such resources remains limited for most languages. While existing emotion lexicons typically comprise words, it is a particular meaning of a word (rather than the word itself) that conveys emotion. To mitigate this issue, we present the Emotion Meanings dataset, a novel dataset of 6000 Polish word meanings. The word meanings are derived from the Polish wordnet (plWordNet), a large semantic network interlinking words by means of lexical and conceptual relations. The word meanings were manually rated for valence and arousal, along with a variety of basic emotion categories (anger, disgust, fear, sadness, anticipation, happiness, surprise, and trust). The annotations were found to be highly reliable, as demonstrated by the similarity between data collected in two independent samples: unsupervised (n = 21,317) and supervised (n = 561). Although we found the annotations to be relatively stable for female, male, younger, and older participants, we share both summary data and individual data to enable emotion research on different demographically specific subgroups. The word meanings are further accompanied by the relevant metadata, derived from open-source linguistic resources. Direct mapping to Princeton WordNet makes the dataset suitable for research on multiple languages. Altogether, this dataset provides a versatile resource that can be employed for emotion research in psychology, cognitive science, psycholinguistics, computational linguistics, and natural language processing. [ABSTRACT FROM AUTHOR]

When integrating data from different sources, there are problems of synonymy, different languages, and concepts of different granularity. This paper proposes a simple yet effective approach to evaluate the semantic similarity of short texts, especially keywords. The method is capable of matching keywords from different sources and languages by exploiting transformers and WordNet-based methods. Key features of the approach include its unsupervised pipeline, mitigation of the lack of context in keywords, scalability for large archives, support for multiple languages and real-world scenarios adaptation capabilities. The work aims to provide a versatile tool for different cultural heritage archives without requiring complex customization. The paper aims to explore different approaches to identifying similarities in 1- or n-gram tags, evaluate and compare different pre-trained language models, and define integrated methods to overcome limitations. Tests to validate the approach have been conducted using the QueryLab portal, a search engine for cultural heritage archives, to evaluate the proposed pipeline.

Unstructured text processing is the first step for several applications such as question answering systems, information retrieval, and recipe classification. In the field of recipe classification, number of frameworks have been proposed. However, it is still very tedious and time consuming to extract the food items from the unstructured text and then process for classification. In this research, an automatic food item detection from unstructured text is proposed based on semantic sense modeling. The candidate nouns are detected which can be food items and then the similarity of those nouns is computed with possible food categories. The candidate noun is treated as food item if the similarity is high. For similarity between possible food item and food category is computed by WordNet ontology. The proposed framework is evaluated on benchmark datasets and competitive performance have been achieved. The F-score on large dataset that contains around 20 K recipes is 0.89 which is improved from 0.56. [ABSTRACT FROM AUTHOR]

In communication, textual data are a vital attribute. In all languages, ambiguous or polysemous words' meaning changes depending on the context in which they are used. The ability to determine the ambiguous word's correct meaning is a Know‐distill challenging task in natural language processing (NLP). Word sense disambiguation (WSD) is an NLP process to analyze and determine the correct meaning of polysemous words in a text. WSD is a computational linguistics task that automatically identifies the polysemous word's set of senses. Based on the context some word comes into view, WSD recognizes and tags the word to its correct priori known meaning. Semitic languages like Arabic have even more significant challenges than other languages since Arabic lacks diacritics, standardization, and a massive shortage of available resources. Recently, many approaches and techniques have been suggested to solve word ambiguity dilemmas in many different ways and several languages. In this review paper, an extensive survey of research works is presented, seeking to solve Arabic word sense disambiguation with the existing AWSD datasets. This article is categorized under:Algorithmic Development > Text MiningTechnologies > Machine Learning [ABSTRACT FROM AUTHOR]

Machine-readable databases such as FrameNet (based on frame semantics) and WordNet (based on lexical semantic relations) appeared in the 1990s and became part of the lexicographic scene. The current study argues that FrameNet and WordNet can contribute to addressing the lexicographic challenge of sense delineation and elicit better performance from learners of English as a second language. The study examined the decoding and encoding performance of university students (n = 48) after exposure to modified lexicographic entries from FrameNet, WordNet, and the online Oxford Learner's Dictionary. The classroom experiment assessed the accuracy of sense selection, user perplexity, and the accuracy of synonym production, and measured the response time for each question. An online survey followed the test, in order to collect further information about students' dictionary use and evaluation of guide words and definitions. Results revealed significant intergroup differences in the response time, perplexity level, and encoding performance. Learners who consulted the modified FrameNetbased entries were the fastest and most successful among the three groups. Future studies can benefit from simplifying the name of frames in FrameNet and modifying the microstructure of the database for pedagogical purposes. [ABSTRACT FROM AUTHOR]