Semantic Ranking Research Articles

Cross-Modal Hashing With Feature Semi-Interaction and Semantic Ranking for Remote Sensing Ship Image Retrieval

Cross-Modal Hashing With Feature Semi-Interaction and Semantic Ranking for Remote Sensing Ship Image Retrieval

LCNA-LSTM CNN based attention model for recommendation system to improve marketing strategies on e-commerce

<p>E-commerce industries have grown at an unpredicted and unprecedented rate in the 21st century, booming in the midst of the COVID-19 catastrophe and revolutionizing our way of life. The COVID-19 pandemic has demonstrated the widespread strong uptake of e-commerce. In today scenario, success of any business depends upon e-commerce platform. As e-commerce uses Machine learning algorithms for processing but they also encounter serious issues, i.e., cold start, sparsity, scalability and many more. In this research work, researchers address the cold start issue, as efficiency drops due to new users or lower engagement of users. Same is resolved in proposed LSTM CNN Based Attention Model (LCNA), a Longest short-term memory (LSTM) recurring neural networks, Convolution Neural Network (CNN) and deep attention layer based model for collaborative filtering to solve the problem of cold start for a new user. The proposed model in this study uses deep attention layer for semantic ranking with cosine similarity to improve the recommendation. Proposed framework primarily functions in stages, starting with the creation of interactive map matrices, then improving ranking using CNN, LSTM, and deep attention layer, and concluding with the framework’s prediction of three key metrics: mean absolute error (MAE), root MSE (RMSE), and accuracy. The framework is put to the test in several metrices using various recommender metrics on the electronics dataset from the Amazon dataset.</p>

Who can verify this? Finding authorities for rumor verification in Twitter

A large body of research work has proposed verification techniques for rumors spreading in social media that mainly relied on subjective evidence, e.g., propagation networks or user interactions. Alternatively, in this work, we introduce the task of authority finding in social media, in which we aim to find authorities, for given rumors spreading specifically in Twitter, who can help verify them by providing exclusive/convincing evidence that supports or denies those rumors. We release the first test collection for Authority FINding in Arabic Twitter (AuFIN). The collection comprises 150 rumors (expressed in tweets) associated with a total of 1,044 authority accounts and a user collection of 395,231 Twitter accounts (members of 1,192,284 unique Twitter lists). Moreover, we propose a hybrid model that employs pre-trained language models and combines lexical, semantic, and network signals to find authorities. Our experiments show that the textual representation of users is insufficient, and incorporating the Twitter network features improved the recall of authorities by 34%. Moreover, semantic ranking is inferior to the lexical and network-based ranking in terms of precision, but superior in terms of recall. Therefore, combining both the semantic and network-based ranking achieved the best overall performance achieving a precision of 0.413 and 0.213 at depth 1 and 5 respectively. We show that rumor expansion by exploiting Knowledge Bases improves the recall of authorities by up to 15%. Furthermore, we find that SOTA models for topic expert finding perform poorly on finding authorities. Finally, drawing upon our experiments, we discuss failure factors and make recommendations for future research directions in addressing this task.

CNS: Application of Distributed Indexing Based on Large-Scale Graph Data in Legal Protection of Network Personal Information

In order to improve the legal protection effect of online personal information, this paper combines intelligent methods to analyze the application of distributed indexing for large-scale graph data in the legal protection of online personal information. Moreover, this paper designs a TreeRank semantic ranking model based on the characteristics of RDF data. In addition, this paper uses the matching algorithm to obtain the relevance ranking of the class to which the primary keyword belongs, improves the intelligent algorithm, and builds an intelligent distributed index structure. Finally, this paper simulates the distributed indexing system based on large-scale graph data in combination with the legal protection requirements of personal information, and verifies the validity of the model through experimental research.

Semantic ranking structure preserving for cross-modal retrieval

Cross-modal retrieval not only needs to eliminate the heterogeneity of modalities, but also needs to constrain the return order of retrieval results. Accordingly, we propose a novel common representation space learning method, called Semantic Ranking Structure Preserving (SRSP) for Cross-modal Retrieval in this paper. First, the dependency relationship between labels is used to minimize the discriminative loss of multi-modal data and mine potential relationships between samples to get richer semantic information in the common space. Second, we constrain the correlation ranking of representations in common space, so as to break the modal gap and promote the multi-modal correlation learning. The comprehensive experimental comparison results show that our algorithm substantially enhances the performance and consistently outperforms very recent algorithms in terms of widely used cross-modal benchmark datasets.

Ranking-Based Deep Cross-Modal Hashing

Cross-modal hashing has been receiving increasing interests for its low storage cost and fast query speed in multi-modal data retrievals. However, most existing hashing methods are based on hand-crafted or raw level features of objects, which may not be optimally compatible with the coding process. Besides, these hashing methods are mainly designed to handle simple pairwise similarity. The complex multilevel ranking semantic structure of instances associated with multiple labels has not been well explored yet. In this paper, we propose a ranking-based deep cross-modal hashing approach (RDCMH). RDCMH firstly uses the feature and label information of data to derive a semi-supervised semantic ranking list. Next, to expand the semantic representation power of hand-crafted features, RDCMH integrates the semantic ranking information into deep cross-modal hashing and jointly optimizes the compatible parameters of deep feature representations and of hashing functions. Experiments on real multi-modal datasets show that RDCMH outperforms other competitive baselines and achieves the state-of-the-art performance in cross-modal retrieval applications.

MR-NET: Exploiting Mutual Relation for Visual Relationship Detection

Inferring the interactions between objects, a.k.a visual relationship detection, is a crucial point for vision understanding, which captures more definite concepts than object detection. Most previous work that treats the interaction between a pair of objects as a one way fail to exploit the mutual relation between objects, which is essential to modern visual application. In this work, we propose a mutual relation net, dubbed MR-Net, to explore the mutual relation between paired objects for visual relationship detection. Specifically, we construct a mutual relation space to model the mutual interaction of paired objects, and employ linear constraint to optimize the mutual interaction, which is called mutual relation learning. Our mutual relation learning does not introduce any parameters, and can adapt to improve the performance of other methods. In addition, we devise a semantic ranking loss to discriminatively penalize predicates with semantic similarity, which is ignored by traditional loss function (e.g., cross entropy with softmax). Then, our MR-Net optimizes the mutual relation learning together with semantic ranking loss with a siamese network. The experimental results on two commonly used datasets (VG and VRD) demonstrate the superior performance of the proposed approach.

Zero-Shot Cross-Media Embedding Learning With Dual Adversarial Distribution Network

Existing cross-media retrieval methods are mainly based on the condition where the training set covers all the categories in the testing set, which lack extensibility to retrieve data of new categories. Thus, zero-shot cross-media retrieval has been a promising direction in practical application, aiming to retrieve data of new categories (unseen categories), only with data of limited known categories (seen categories) for training. It is challenging for not only the heterogeneous distributions across different media types, but also the inconsistent semantics across seen and unseen categories need to be handled. To address the above issues, we propose dual adversarial distribution network (DADN) , to learn common embeddings and explore the knowledge from word-embeddings of different categories. The main contributions are as follows. First, zero-shot cross-media dual generative adversarial networks architecture is proposed, in which two kinds of generative adversarial networks (GANs) for common embedding generation and representation reconstruction form dual processes. The dual GANs mutually promote to model semantic and underlying structure information, which generalizes across different categories on heterogeneous distributions and boosts correlation learning. Second, distribution matching with maximum mean discrepancy criterion is proposed to combine with dual GANs, which enhances distribution matching between common embeddings and category word-embeddings. Finally, adversarial inter-media metric constraint is proposed with an inter-media loss and a quadruplet loss, which further model the inter-media correlation information and improve semantic ranking ability. The experiments on four widely used cross-media datasets demonstrate the effectiveness of our DADN approach.

Exploring the Importance of Entities in Semantic Ranking

In recent years, entity-based ranking models have led to exciting breakthroughs in the research of information retrieval. Compared with traditional retrieval models, entity-based representation enables a better understanding of queries and documents. However, the existing entity-based models neglect the importance of entities in a document. This paper attempts to explore the effects of the importance of entities in a document. Specifically, the dataset analysis is conducted which verifies the correlation between the importance of entities in a document and document ranking. Then, this paper enhances two entity-based models—toy model and Explicit Semantic Ranking model (ESR)—by considering the importance of entities. In contrast to the existing models, the enhanced models assign the weights of entities according to their importance. Experimental results show that the enhanced toy model and ESR can outperform the two baselines by as much as 4.57% and 2.74% on NDCG@20 respectively, and further experiments reveal that the strength of the enhanced models is more evident on long queries and the queries where ESR fails, confirming the effectiveness of taking the importance of entities into account.

Segmentation in Weakly Labeled Videos via a Semantic Ranking and Optical Warping Network.

Weakly supervised video object segmentation (WSVOS) focuses on generating pixel-level object masks for videos only tagged with class labels, which is an essential yet challenging task. For WSVOS, the algorithm is just aware of rough category information rather than the concrete object size and location cues, besides it lacks reliable annotated exemplars to learn temporal evolution in the investigated videos. Basically, there are three challenging factors which may influence the performance of WSVOS: foreground object discovery in each frame, coarse object semantic consistency within each video, and fine-grained segmentation smoothness within neighbor frames. In this paper, we establish a semantic ranking and optical warping network (SROWN) to simultaneously solve these three challenges in a unified framework. For the first challenge, we apply the still image saliency detection method and discover the foreground object for each frame via a segmentation network. Due to the huge discrepancies between the image saliency and the video object segmentation, we step further and propose two subnetworks to solve the other two challenges. For the second one, we propose an attentive semantic ranking subnetwork to mine video-level tags, which can learn discriminative features for semantic ranking and lead to semantic consistent segmentation masks. For the third one, we propose an optical flow warping subnetwork to constrain fine-grained segmentation smoothness within neighbor frames, which can suppress the large deformation and thus obtain smooth object boundaries for adjacent frames. Experiments on two benchmark datasets, i.e., DAVIS dataset and YouTube-Objects dataset, demonstrate the effectiveness of the proposed approach for segmenting out video objects under weak supervision.

RankTagViz: A Semantic Ranking and Tags Visualization of User Travelogues

The travelogues contain prosperous geo-referenced information such as tours, weather, and expenses etc. Reading travelogues and finding useful trip information is a tedious task for many people. So as to correlate the georeferenced text meaningfully the semantic ranking is taken in account on the user contributing travelogues. In this way there is a requirement for creating programmed travelogue mining methods to pass on valuable data in a travelogue to its peruses in the more successful way. Visualizing a geo-referenced data with location tags and descriptions makes it helpful for readers to understand the fundamental substance of the travelogue. For user expedient, this paper proposes a Travelogue RankTagViz approach that semantically ranks and visualizes the tags of user travelogues. The RankTagViz contains two phases. In the first phase, the user travelogues are ranked based on semantic. In the next phase, the travelogues are visualized based on the tag and images. During semantic ranking, a semantic dimension reduction method is proposed to pre-process and mine useful information from the travelogues. After that a semantic rank mechanism is proposed to rank travelogues based on tags and POI (Point Of Interest). For tag visualization, location based tags and images are extracted from the travelogues and a novel UI is intended to give a superior client encounter, by arranging both the literary and visual data produced by committed voyagers in an alluring way. Test comes about on an arrangement of gathered travelogues show the proposed techniques' capacity to rank and imagine travelogues.

Semantic Ranking based on Keywords Extractions in the web

The collection of terminologies from text is animportant for several issues, i.e., the future extraction and also summarization of keyword in semantic network. Semantic Ranking is typically performed using statistical data from text. Our proposed strategy develops a semantic chart utilizing understood connections, which depend on semantic relatedness between content hubs and therefore positions hubsutilizing Ranking calculations. This research work focuses only Semantic Rank for keyword mining from text. This method for keywords extraction shows that Semantic Rank makesconstructivelyin the data set.

Training Visual-Semantic Embedding Network for Boosting Automatic Image Annotation

Image auto-annotation which annotates images according to their semantic contents has become a research focus in computer vision, as it helps people to edit, retrieve and understand large image collections. In the last decades, researchers have proposed many approaches to solve this task and achieved remarkable performance on several standard image datasets. In this paper, we train neural networks using visual and semantic ranking loss to learn visual-semantic embedding. This embedding can be easily applied to nearest-neighbor based models to boost their performance on image auto-annotation. We test our method on four challenging image datasets, reporting comparison results with existing works. Experimental results show that our method can be applied to several state-of-the-art nearest-neighbor based models including TagProp and 2PKNN, and significantly improves their performance.

Hidden information retrieval and evaluation method and tools utilising ontology reasoning applied for financial fraud analysis.

The paper summarizes a semantic association evaluation and reasoning method, utilising domain and problem solving ontologies. The method combines algorithms for data aggregation and logic reasoning utilising concrete financial data. As an outcome method supports suspicious behaviour recognition of money loundering schemes. These scenarios and schemes are implemented by the analysts using ontology-based constructs. Provided tools cover all stages of data processing starting from structural data extraction and migration, aggregation to reasoning using logic (DL and FOL) constructs. Advances in automatic reasoning and the availability of semantic processing tools encourage analysts to extend existing link analysis methods towards contextual data processing. To demonstrate presented method, a proof of concept environment IAFEC Ontology Toolkit has been described. It delivers initial financial fraud identification schemes (rules) based on set of problem solving ontologies. The novelty in such approach comes from incorporating heterogeneous types of data, which usually are processed by graph methods. The semantic tool, extend capabilities of graph-based (homogeneous) approach by delivering context-aware indirect association identification, and inference path explanation and inspection capabilities. Presented material describes the method and analytical algorithms, which demonstrate description logic reasoning and graph-based semantic association identification and ranking. Developed method has been implemented, as a Protégé OWL 5.0 environment extension, supplemented with web-services delivering distributed data processing, aggregation (not available in ontology languages). The environment provides declarative processing capabilities enabling analysts to design configurable processing flow, and new financial fraud identification schemes.

E-Commerce Grading Approaches of Ultimate Ranking System- A Review

e-commerce ranking algorithms are the algorithms utilized for the purpose of product suggestions and product listings against the input query or the browsed product on the given shopping portal. The product ranking portals requires the number of computations on the basis of various factors for the calculation of the ranks of the given products. Various local and online factors can be grouped for the overall evaluation of the product ranks according to the input keywords specifically on the first stage product lookup on the shopping sites. In this paper, the product ranking solution has been proposed with the versatile approach using the popularity and accessibility factors. Also the reliability factors are evaluated which analyzes the trust factor for the page by using the online security evaluation programs. Various experiments would be conducted over the large number of input product data obtained from the application programming interfaces (API) from the prominent shopping portals active online. The proposed model is expected to resolve the issue by evaluating the proposed model performance in comparison with the existing model on the basis of various factor associated with the time complexity and reliability. Keywordssystem, semantic ranking, machine learning, e- commerce ranking, semantic web, neural network.

An RDF Metadata-Based Weighted Semantic Pagerank Algorithm

PageRank evaluates the importance of Web pages with link relations. However, there is no direct method of evaluating the meaning of links in a hyperlink-based Web structure. This feature may cause problems in that pages containing many in-links are highly ranked without considering the meaning of the link relations among the pages. We therefore propose a novel ranking approach to directly analyze the meaning of links by transforming a hyperlink-based Web structure into a semantic-link-based Web structure. We extract semantic metadata from Web pages and construct a semantic-link-based Web structure using RDF model. We define a metric to evaluate the weight of the links for stratifying rank values based on their importance in the semantic-link-based Web structure. We implement the weighted semantic ranking algorithm in the MapReduce framework to consider large-scale semantic metadata. The results of our experiment show that our approach outperforms existing PageRank algorithms.

Enhancing Fuzzy Topsis for web service selection

With proliferation of web services, quality-of-service QoS capability is becoming a significant factor for distinguishing services. QoS includes various non-functional issues such as performance, reliability and availability, etc. The functional requirements are assured through Semantic Matchmaking algorithm whereas the non functional requirements are guaranteed though QoS factors. Selection of a web service has become very crucial due to the availability of multiple web service with similar functionality. To solve this issue there is a need to distinguish the services based on their non-functional criteria QoS. This paper proposes an enhanced QoS-based semantic web service selection and ranking solution with the application of Fuzzy Topsis method. Further, this paper proposes a new dynamic replication algorithm to increase the availability of the service and decrease network delays. This is demonstrated using a diabetes monitoring system.

Automatic Discovery of Relevant Web Services with Semantic Ranking

The main objective of this study is to discover and rank relevant web services from the given set of WSDL documents automatically for the specified user requirements through semantic search. Current web services are described using WSDL that only provides syntactic description of a web service. This syntactic description is related with the implementation details of a service and thus customized towards the requirements of the programmer. Semantic descriptions of the web services are necessary to automatically discover many relevant web services for a specific user request. In the proposed architecture, the semantic descriptions of a given set of WSDL documents are extracted using WordNet Ontology Framework and Suggested Upper Merged Ontology (SUMO) concepts and these semantic descriptions are stored in the database along with the WSDL documents. The given set of WSDL documents are semantically categorized in to six predefined categories based on the semantic descriptions. These semantically categorized WSDL documents are searched for discovering relevant web services for a refined service request. As these relevant web services are semantically ranked and displayed, the users are facilitated with the required information. The effectiveness of this proposed architecture is evaluated with set of WSDL documents that are collected from OWLS-TC test collection. The analysis of results shows that the proposed architecture provides better improvement on recall, precision and average query response time.

Improving semantic Web services discovery and ranking: A lightweight, integrated approach

Los Servicios Web Semanticos (SWSs) se han convertido en un area de investigacion muy activa, en la que diversos frameworks, como WSMO u OWL-S, definen ont ologias de la Web Semantica para describir servicios Web, de forma que puedan descubrirse, ordenarse, componerse y ejecutarse automaticamente, de acuerdo a los requisitos y preferencias del usuario. En concreto, diversas tecnicas de descubrimiento y ranking han sido propuestas, incluyendo herramientas relacionadas. Sin embargo, las propuestas existentes ofrecen una expresividad limitada para definir preferencias, siendo muy dependientes de los formalismos subyacentes. Ademas, las tecnicas semanticas actuales sufren de problemas de rendimiento, interoperabilidad e integracion, impidiendo su explotacion efectiva. Para solucionar estos problemas, la investigacion se centra actualmente en el desarrollo de descripciones de SWSs ligeras, las cuales faciliten la interoperabilidad de las propuestas actuales, asi como en soluciones para el descubrimiento y el ranking que proporcionen mejor rendimiento con una disminucion acotada en su precision. En esta memoria de tesis, resolvemos esos desafios mediante la propuesta de SOUP, un completo modelo ontologico de preferencias sobre el que hemos desarrollado unas herramientas ligeras, llamadas EMMA y PURI, que mejoran el rendimiento del descubrimiento e integran las propuestas actuales de ranking, respectivamente. Nuestras contribuciones han sido evaluadas aplicandolas a escenarios sinteticos y reales. En primer lugar, la expresividad e independencia del modelo de preferencias SOUP se ha validado mediante la descripcion completa de escenarios complejos del SWS Challenge. Ademas, hemos llevado a cabo un estudio experimental de EMMA que muestra una mejora de rendimiento significativa, obteniendo una penalizacion en la precision insignificante. Por ultimo, hemos aplicado PURI dentro del proyecto SOA4All, integrando sus tres mecanismos de ranking (objetivo, basado en NFP, y en logicas difusas) en una solucion de descubrimiento y ranking interoperable.

Ontology-based blog collection and profile-based personalised ranking

Blogging plays a major role in today's social networking. Large numbers of blogs are available in the web. This becomes a tedious task for the user and the blogger, to search for a technical blog from a list of randomly appearing blogs. Thus a method is required to rank the blogs for the user, according to the relevance of the query. The existing works focus either on the behaviour of the blogger or the linking of one blog with another blog. This paper introduces a novel method of ranking blogs based on the relevance of the content to the query. The semantics of the query is analysed to rank the blogs. The proposed method also uses concept-based profiling, to enable the user to retrieve personalised content for the query. The experimental analysis shows that the proposed personalised semantic blog ranking PSBR is efficient in ranking the technical blogs for the user.