Information Retrieval Tools Research Articles

Research on Intelligent Construction Algorithm of Subject Knowledge Thesaurus Based on Literature Resources

The implementation of National Science and Technology Innovation Strategy demands exponential growing in knowledge services on literature information institutions. It is the most important knowledge organization tool for Information Retrieval, which can be widely used for semantic citation, organization and retrieval of literature resources. This study aims to develop an innovative algorithm for constructing subject thesaurus based on massive literature resource data and mining academic neologisms, also the semantic relationship between academic neologisms and subject system. We firstly collect a dataset of literature corpus, corresponding work for data pre-processing carried out. Then using the FastText model to complete academic neologisms mining, we construct an automatic categorization model of academic neologisms based on the Bert and TextCNN algorithm. The algorithm proposed in this study is validated by 8.1 million multi-source and heterogeneous literature data in the field of marine disciplines. The result shows that the algorithm can effectively replace 90% of the manual annotation volume, mine a large number of high-quality marine neologisms and successfully build the marine science knowledge base with a pass rate of 82.6% reviewed by expert, which present high accuracy and certain engineering application prospects.

Robust PDF Document Conversion using Recurrent Neural Networks

The number of published PDF documents in both the academic and commercial world has increased exponentially in recent decades. There is a growing need to make their rich content discoverable to information retrieval tools. Achieving high-quality semantic searches demands that a document's structural components such as title, section headers, paragraphs, (nested) lists, tables and figures (including their captions) are properly identified. Unfortunately, the PDF format is known to not conserve such structural information because it simply represents a document as a stream of low-level printing commands, in which one or more characters are placed in a bounding box with a particular styling. In this paper, we present a novel approach to document structure recovery in PDF using recurrent neural networks to process the low-level PDF data representation directly, instead of relying on a visual re-interpretation of the rendered PDF page, as has been proposed in previous literature. We demonstrate how a sequence of PDF printing commands can be used as input into a neural network and how the network can learn to classify each printing command according to its structural function in the page. This approach has three advantages: First, it can distinguish among more fine-grained labels (typically 10-20 labels as opposed to 1-5 with visual methods), which results in a more accurate and detailed document structure resolution. Second, it can take into account the text flow across pages more naturally compared to visual methods because it can concatenate the printing commands of sequential pages. Last, our proposed method needs less memory and it is computationally less expensive than visual methods. This allows us to deploy such models in production environments at a much lower cost. Through extensive architectural search in combination with advanced feature engineering, we were able to implement a model that yields a weighted average F1 score of 97% across 17 distinct structural labels. The best model we achieved is currently served in production environments on our Corpus Conversion Service (CCS), which was presented at KDD18. This model enhances the capabilities of CCS significantly, as it eliminates the need for human annotated label ground-truth for every unseen document layout. This proved particularly useful when applied to a huge corpus of PDF articles related to COVID-19.

Information Retrieval Tools as Predictors for Information Resources Utilization in Academic Libraries in Nigeria

Information Retrieval Tools as Predictors for Information Resources Utilization in Academic Libraries in Nigeria

MEDLINE search retrieval issues: A longitudinal query analysis of five vendor platforms.

This study compared the results of data collected from a longitudinal query analysis of the MEDLINE database hosted on multiple platforms that include PubMed, EBSCOHost, Ovid, ProQuest, and Web of Science. The goal was to identify variations among the search results on the platforms after controlling for search query syntax. We devised twenty-nine cases of search queries comprised of five semantically equivalent queries per case to search against the five MEDLINE database platforms. We ran our queries monthly for a year and collected search result count data to observe changes. We found that search results varied considerably depending on MEDLINE platform. Reasons for variations were due to trends in scholarly publication such as publishing individual papers online first versus complete issues. Some other reasons were metadata differences in bibliographic records; differences in the levels of specificity of search fields provided by the platforms and large fluctuations in monthly search results based on the same query. Database integrity and currency issues were observed as each platform updated its MEDLINE data throughout the year. Specific biomedical bibliographic databases are used to inform clinical decision-making, create systematic reviews, and construct knowledge bases for clinical decision support systems. They serve as essential information retrieval and discovery tools to help identify and collect research data and are used in a broad range of fields and as the basis of multiple research designs. This study should help clinicians, researchers, librarians, informationists, and others understand how these platforms differ and inform future work in their standardization.

COVID-19 trial graph: a linked graph for COVID-19 clinical trials.

ObjectiveClinical trials are an essential part of the effort to find safe and effective prevention and treatment for COVID-19. Given the rapid growth of COVID-19 clinical trials, there is an urgent need for a better clinical trial information retrieval tool that supports searching by specifying criteria, including both eligibility criteria and structured trial information.Materials and MethodsWe built a linked graph for registered COVID-19 clinical trials: the COVID-19 Trial Graph, to facilitate retrieval of clinical trials. Natural language processing tools were leveraged to extract and normalize the clinical trial information from both their eligibility criteria free texts and structured information from ClinicalTrials.gov. We linked the extracted data using the COVID-19 Trial Graph and imported it to a graph database, which supports both querying and visualization. We evaluated trial graph using case queries and graph embedding.ResultsThe graph currently (as of October 5, 2020) contains 3392 registered COVID-19 clinical trials, with 17 480 nodes and 65 236 relationships. Manual evaluation of case queries found high precision and recall scores on retrieving relevant clinical trials searching from both eligibility criteria and trial-structured information. We observed clustering in clinical trials via graph embedding, which also showed superiority over the baseline (0.870 vs 0.820) in evaluating whether a trial can complete its recruitment successfully.ConclusionsThe COVID-19 Trial Graph is a novel representation of clinical trials that allows diverse search queries and provides a graph-based visualization of COVID-19 clinical trials. High-dimensional vectors mapped by graph embedding for clinical trials would be potentially beneficial for many downstream applications, such as trial end recruitment status prediction and trial similarity comparison. Our methodology also is generalizable to other clinical trials.

An ensemble of inverse moment estimators for sufficient dimension reduction

Sufficient dimension reduction (SDR) is known to be a useful tool in data visualization and information retrieval for high dimensional data. Many well-known SDR approaches investigate the inverse conditional moments of the predictors given the response. Motivated by the idea of the aggregate dimension reduction, we propose an ensemble of inverse moment estimators to explore the central subspace. The new approach can substantially improve the estimation accuracy for the directions beyond the regression mean functions. A ladle estimator is proposed to determine the structural dimension of the central subspace. We further present two variable selection procedures to improve the interpretability of the reduced variables. Both simulation studies and a real data application show the efficacy of the newly proposed method.

Evidence-based Mechanistic Integrative Modeling of Stroke Risk Factors: A Translational Research Study

Introduction: According to global statistics, stroke is known as the main health problem in the world. Many clinical and molecular research, which are stored in the different repository with the various format have been conducted in the area of stroke domain. The heterogeneity of these research data does not make a comprehensive view of the disease. Recently, translational research has been developed to fill the gap between these studies. In this study, we used the integrative disease modeling method to model the underlying mechanism of stroke risk factors.Material and Methods: This study was conducted in three steps: data gathering, model construction, and mechanism discovery. First, using semantic and information retrieval tools, we extracted the cause and effect statement from the literature to create the mechanistic model, and the validated molecular data to evaluate the constructed model. Then, the integrative model was created and evaluated. Finally, we used Gene Set Enrichment Analysis to identify the main biological process and signaling pathways in the mechanism of the disease. Results: In the evidence-based information retrieval from the literature, 1837 causal statement was extracted. The initial network was created with 648 nodes (molecular, clinical, and environmental factors) and 1837 edges (interactions). Also, 51 genes/proteins and nine single nucleotide polymorphisms were matched with data in the model. The inflammatory response, response to lipid, regulation of body fluid levels, and regulation of response to stress, complement and coagulation cascades, and PPAR signaling pathway were the main biological processes and signaling pathways enriched in GSEA analysis.Conclusion: This study showed that we can identify the underlying mechanism of stroke risk factors and use a proper strategy to prevent it, using Integrative Disease Modeling.

Cluster-based information retrieval by using (K-means)-hierarchical parallel genetic algorithms approach

Cluster-based information retrieval is one of the information retrieval (IR) tools that organize, extract features and categorize the web documents according to their similarity. Unlike traditional approaches, cluster-based IR is fast in processing large datasets of document. To improve the quality of retrieved documents, increase the efficiency of IR and reduce irrelevant documents from user search. In this paper, we proposed a (K-means)-hierarchical parallel genetic algorithms approach (HPGA) that combines the K-means clustering algorithm with hybrid PG of multi-deme and master/slave PG algorithms. K-means uses to cluster the population to k subpopulations then take most clusters relevant to the query to manipulate in a parallel way by the two levels of genetic parallelism, thus, irrelevant documents will not be included in subpopulations, as a way to improve the quality of results. Three common datasets (NLP, CISI, and CACM) are used to compute the recall, precision, and F-measure averages. Finally, we compared the precision values of three datasets with Genetic-IR and classic-IR. The proposed approach precision improvements with IR-GA were 45% in the CACM, 27% in the CISI, and 25% in the NLP. While, by comparing with Classic-IR, (K-means)-HPGA got 47% in CACM, 28% in CISI, and 34% in NLP.

Abstract databases on agriculture as a means of scientific communication

The international branch databases, such as "Agricola", "Agris", "Agora" and a number of statistical ones, are presented, their brief description, filling, volume, and also the problems which arise at their use are given.The system of abstracting national publications on agriculture is described, which is an important part of information support of research, technical development, educational process, as the abstract database of scientific publications is not only the most effective tool for information retrieval, but also a generalized picture of the current state of science in the country. The article highlights the role of the National Scientific Agricultural Library of NAAS — the leading scientific and information center of the agricultural sector of Ukraine — in the formation of sectoral information and abstract resource, namely the abstract journal "Agroindustrial Complex of Ukraine", which is published four times a year since the second half of 1999 400 abstracts which reflect more than 31 thousand scientific publications in various fields of agricultural science, which reflect the scientific-theoretical, applied and combined range of diverse research to further increase the efficiency of agricultural production on the basis of resource conservation, environmental friendliness, competitiveness. The main measures of the preparatory stage of creating a branch abstract and information resource are analyzed. Theoretical, organizational and methodological principles of abstracting, ways of further development of secondary information materials are covered. The information on the issue of thematic issues of the abstract magazine "Agroindustrial Complex of Ukraine" in general and in particular for the last year is given. The analysis of the editions presented in the abstract magazine for the last some years is presented, the editions which are most fully reflected in the given information resource are resulted. The thematic content is analyzed. The methods of dissemination of agricultural information are given — meetings, conferences, website of NSАL NAAS. The transformation of the journal during the whole period of its existence caused by the requirements of time, the problems that arise when creating a secondary information resource are considered.

Assessing the influence of point-of-interest features on the classification of place categories

Points of interest (POIs) digitally represent real-world amenities as point locations. POI categories (e.g. restaurant, hotel, museum etc.) play a prominent role in several location-based applications such as social media, navigation, recommender systems, geographic information retrieval tools, and travel-related services. The majority of user queries in these applications center around POI categories. For instance, people often search for the closest pub or the best value-for-money hotel in an area. To provide valid answers to such queries, accurate and consistent information on POI categories is an essential requirement. Nevertheless, category-based annotations of POIs are often missing. The task of annotating unlabeled POIs in terms of their categories — known as POI classification — is commonly achieved by means of machine learning (ML) models, often referred to as classifiers. Central to this task is the extraction of known features from pre-labeled POIs in order to train the classifiers and, then, use the trained models to categorize unlabeled POIs. However, the set of features used in this process can heavily influence the classification results. Research on defining the influence of different features on the categorization of POIs is currently lacking. This paper contributes a study of feature importance for the classification of unlabeled POIs into categories. We define five feature sets that address operation based, review-based, topic-based, neighborhood-based, and visual attributes of POIs. Contrary to existing studies that predominantly use multi-class classification approaches, and in order to assess and rank the influence of POI features on the categorization task, we propose both a multi-class and a binary classification approach. These, respectively, predict the place category among a specified set of POI categories, or indicate whether a POI belongs to a certain category. Using POI data from Amsterdam and Athens to implement and evaluate our study approach, we show that operation based features, such as opening or visiting hours throughout the day, are the most important place category predictors. Moreover, we demonstrate that the use of feature combinations, as opposed to the use of individual features, improves the classification performance by an average of 15%, in terms of F1-score.

Extracting human cortical responses to sound onsets and acoustic feature changes in real music, and their relation to event rate

Evoked cortical responses (ERs) have mainly been studied in controlled experiments using simplified stimuli. Though, an outstanding question is how the human cortex responds to the complex stimuli encountered in realistic situations. Few electroencephalography (EEG) studies have used Music Information Retrieval (MIR) tools to extract cortical P1/N1/P2 to acoustical changes in real music. However, less than ten events per music piece could be detected leading to ERs due to limitations in automatic detection of sound onsets. Also, the factors influencing a successful extraction of the ERs have not been identified. Finally, previous studies did not localize the sources of the cortical generators. This study is based on an EEG/MEG dataset from 48 healthy normal hearing participants listening to three real music pieces. Acoustic features were computed from the audio signal of the music with the MIR Toolbox. To overcome limits in automatic methods, sound onsets were also manually detected. The chance of obtaining detectable ERs based on ten randomly picked onset points was less than 1:10,000. For the first time, we show that naturalistic P1/N1/P2 ERs can be reliably measured across 100 manually identified sound onsets, substantially improving the signal-to-noise level compared to <10 trials. More ERs were measurable in musical sections with slow event rates (0.2 Hz-2.5 Hz) than with fast event rates (>2.5 Hz). Furthermore, during monophonic sections of the music only P1/P2 were measurable, and during polyphonic sections only N1. Finally, MEG source analysis revealed that naturalistic P2 is located in core areas of the auditory cortex.

Sistemas de recuperación de información implementados a partir de CORD-19: herramientas clave en la gestión de la información sobre COVID-19

La investigación sobre el coronavirus ha generado una producción de documentos científicos extraordinaria. Su tratamiento y asimilación por parte de la comunidad científica ha necesitado de la ayuda de sistemas de recuperación de información diseñados específicamente. Algunas de las principales instituciones mundiales dedicadas a la lucha contra la pandemia han desarrollado el conjunto de datos CORD-19 que destaca sobre otros proyectos de similar naturaleza. Los documentos recopilados en esta fuente han sido procesados por distintas herramientas de recuperación de información, a veces prototipos o sistemas que ya estaban implementados. Se ha analizado la tipología y características principales de estos sistemas concluyendo que hay tres grandes categorías no excluyentes entre ellas: búsqueda terminológica, visualización de información y procesamiento de lenguaje natural. Destaca enormemente que la gran mayoría de ellos emplean preferentemente tecnologías de búsqueda semántica con el objeto de facilitar la adquisición de conocimiento s los investigadores y ayudarlas en su ingente tarea. La crisis provocada por la pandemia ha sido aprovechada por los buscadores semánticos para encontrar su sitio.

Uso de Scopus y Web of Science para investigar y evaluar en comunicación social: análisis comparativo y caracterización

El objetivo es caracterizar y comparar las opciones que ofrecen las principales bases de datos a los investigadores en comunicación social para buscar y evaluar información académica. Para ello, mediante una evaluación experta se examinaron las características funcionales de Scopus y Web of Science. Como resultado se presenta una revisión detallada de dimensiones como la cobertura, la recuperación de información y las herramientas de análisis de fuentes y autores presentes en cada una, además de métricas de impacto específicas. Entre las conclusiones se observan similitudes en las funciones disponibles, pero diferencias significativas en el número de revistas en ámbito de las ciencias sociales y las humanidades, lo que deja a Scopus en mejor posición en el caso de tener que optar.

A survey of tag clouds as tools for information retrieval and content representation

Tag clouds are tools that have been widely used on the Internet since their conception. The main applications of these textual visualizations are information retrieval, content representation and browsing of the original text from which the tags are generated. Despite the extensive use of tag clouds, their enormous popularity and the amount of research related to different aspects of them, few studies have summarized their most important features when they work as tools for information retrieval and content representation. In this paper we present a summary of the main characteristics of tag clouds found in the literature, such as their different functions, designs and negative aspects. We also present a summary of the most popular metrics used to capture the structural properties of a tag cloud generated from the query results, as well as other measures for evaluating the goodness of the tag cloud when it works as a tool for content representation. The different methods for tagging and the semantic association processes in tag clouds are also considered. Finally we give a list of alternative for visual interfaces, which makes this study a useful first help for researchers who want to study the content representation and information retrieval interfaces in greater depth.

An Integrated Search Framework for Leveraging the Knowledge-Based Web Ecosystem

The explosion of information constrains the judgement of search terms associated with Knowledge-Based Web Ecosystem (KBWE), making the retrieval of relevant information and its knowledge management challenging. The existing information retrieval (IR) tools and their fusion in a framework need attention, in which search results can effectively be managed. In this article, we demonstrate the effective use of information retrieval services by a variety of users and agents in various KBWE scenarios. An innovative Integrated Search Framework (ISF) is proposed, which utilises crawling strategies, web search technologies and traditional database search methods. Besides, ISF offers comprehensive, dynamic, personalized, and organization-oriented information retrieval services, ranging from the Internet, extranet, intranet, to personal desktop. In this empirical research, experiments are carried out demonstrating the improvements in the search process, as discerned in the conceptual ISF. The experimental results show improved precision compared with other popular search engines.

A Comparative Study of Literature Based on Mathematical Statistics in Big Data Environment

Large data environment, clustering technology has become a powerful tool for information retrieval and text mining. The essence of it is to set up a target and divide a group of objects into different groups according to the similarity. This process can set the target as similar elements in the same group, or different groups of different elements, thus achieving different purposes is a kind of undirected learning research method. This paper compares the two literature by using the frequency and clustering of the computer mathematical statistics, and clearly shows that the key to achieving harmony is to build a good society.

CREATION FOR A CADASTRAL INFORMATION SYSTEM FOR GONIN GORA LAYOUT IN CHIKUN LOCAL GOVERNMENT AREA OF KADUNA STATE, NIGERIA

This study was carried out to determine the geographic information system (GIS) as a tool to examine spatial patterns of Gonin Gora layout in Chikun Local Government Area of Kaduna state. The study involves acquiring spatial and attribute data of the study area, creating relevant spatial layers containing vital information about the parcel within the area of study, designing an appropriate database using relational database model (RDBM) cadastral information system and demonstrating the usability of the cadastral information system (CIS) that was created by carrying out simple queries. The research procedure also involves data acquisition using Promark 3 GPS instrument, database design and spatial queries using ARCGIS (10.1). Three(3) single criteria queries were generated for parcel earmarked for religion worship, parcel set for educational purposes, parcels for commercial purposes were identified and multi- criteria seven (7) queries were generated. The database was created and it was successfully test run for cadastral information system as effective tool for quick retrieval of information

An efficient generic approach for automatic taxonomy generation using HMMs

Taxonomies are essential tools for fast information retrieval and classification of knowledge. Many existing techniques for automatic taxonomy generation strongly depend on the specific properties of a particular domain and are consequently hard to apply to other domains. Some attempts have been made to design taxonomies for multiple domains. Unfortunately, they induce high hierarchical classification error rates for some datasets. The automatic design of a taxonomy requires the capability of measuring the similarity between classes. More precisely, the fact that two classes are near intuitively implies that some elements of one class are scattered in the neighborhood of some elements of the other class. This observation is used in this paper to propose a new generic technique for automatic taxonomy generation. A topological analysis of the neighborhood of each instance is first performed. The results of this analysis are used to initialize and train a hidden Markov model for each class. The model of a given class c captures the frequencies of the classes found in the neighborhood of the instances of c, from the most dominant class to the least dominant. The similarities between these models are finally used to derive a taxonomy. Hierarchical classification experiments realized on 20 datasets from various domains showed an average accuracy of $$97.22\%$$ and a standard deviation of $$4.11\%$$ . Comparison results revealed that the proposed approach outperforms existing work with accuracy gains reaching $$38.62\%$$ for one dataset.

Modern Users of Libraries and the Linked Open Data Environment

The goal of any library is to ensure high quality and general availability of information retrieval tools. The paper describes the project implemented by the Russian State Library (RSL) to present Library Bibliographic Classification as a Networked Knowledge Organization System. The project goal is to support content and provide tools for ensuring system’s interoperability with other resources of the same nature (i.e. with Linked Data Vocabularies) in the global network environment. The project was partially supported by the Russian Foundation for Basic Research (RFBR).The RSL General Classified Catalogue (GCC) was selected as the main data source for the Classification system of knowledge organization. The meaning of each classification number is expressed by complete string of wordings (captions), rather than the last level caption alone. Data converted to the Resource Description Framework (RDF) files based on the standard set of properties defined in the Simple Knowledge Organization System (SKOS) model was loaded into the semantic storage for subsequent data processing using the SPARQL query language. In order to enrich user queries for search of resources, the RSL has published its Classification System in the form of Linked Open Data (https://lod.rsl.ru) for searching in the RSL electronic catalogue. Currently, the work is underway to enable its smooth integration with other LOD vocabularies. The SKOS mapping tags are used to differentiate the types of connections between SKOS elements (concepts) existing in different concept schemes, for example, UDC, MeSH, authority data.The conceptual schemes of the leading classifications are fundamentally different from each other. Establishing correspondence between concepts is possible only on the basis of lexical and structural analysis to compute the concept similarity as a combination of attributes.The authors are looking forward to working with libraries in Russia and other countries to create a common space of Linked Open Data vocabularies.

A multistep priority-based ranking for top-N recommendation using social and tag information

Recommender system is a collection of information retrieval tools and techniques used for recommending items to users based on their choices. For improving recommendation accuracy, the use of extra information (e.g., social, trust, item tags, etc.) other than user-item rating data remains an active area of research since last one and half decade. In this paper, we propose a novel methodology for top-N item recommendation, which uses three different kinds of information: user-item rating data, social network among the users, and tags associated with the items. The proposed method has mainly five steps: (i) creation of neighbor users’ item set, (ii) construction of the user-feature matrix, (iii) computation of user priority, (iv) computation of item priority, and finally, (v) recommendation based on the item priority. We implement the proposed methodology with three recommendation dataset. We compare our results with that of the obtained from some state-of-the-art ranking methods and observe that recommendation accuracy is improved in the case of the proposed algorithm for both all users and cold-start users scenarios. The algorithm is also able to generate more cold-start items in the recommended item list.