Computer Science › Artificial Intelligence

Advanced Text Analysis Techniques

Works Count: 37479

Wikipedia

Description

This cluster of papers focuses on the automatic extraction of keywords from textual data using various techniques such as graph-based methods, unsupervised approaches, and neural networks. The research explores the application of linguistic knowledge and statistical information to improve the accuracy of keyword extraction from documents.

Keywords

Automatic; Extraction; Textual Data; Keyword; Linguistic Knowledge; Graph-Based; Unsupervised Approach; Neural Networks; Statistical Information; Document

TextRank: Bringing Order into Text

2004-07-01

Rada Mihalcea , Paul Tarau

In this paper, the authors introduce TextRank, a graph-based ranking model for text processing, and show how this model can be successfully used in natural language applications. In this paper, the authors introduce TextRank, a graph-based ranking model for text processing, and show how this model can be successfully used in natural language applications.

Introduction to information retrieval

2009-01-01

Christopher D. Manning , Prabhakar Raghavan , Hinrich Schütze

Class-tested and coherent, this groundbreaking new textbook teaches web-era information retrieval, including web search and the related areas of text classification and text clustering from basic concepts. Written from a … Class-tested and coherent, this groundbreaking new textbook teaches web-era information retrieval, including web search and the related areas of text classification and text clustering from basic concepts. Written from a computer science perspective by three leading experts in the field, it gives an up-to-date treatment of all aspects of the design and implementation of systems for gathering, indexing, and searching documents; methods for evaluating systems; and an introduction to the use of machine learning methods on text collections. All the important ideas are explained using examples and figures, making it perfect for introductory courses in information retrieval for advanced undergraduates and graduate students in computer science. Based on feedback from extensive classroom experience, the book has been carefully structured in order to make teaching more natural and effective. Although originally designed as the primary text for a graduate or advanced undergraduate course in information retrieval, the book will also create a buzz for researchers and professionals alike.

View PDF Chat

How persuasive is a good fit? A comment on theory testing.

2000-04-01

Seth Roberts , Harold Pashler

Quantitative theories with free parameters often gain credence when they closely fit data. This is a mistake. A good fit reveals nothing about the flexibility of the theory (how much … Quantitative theories with free parameters often gain credence when they closely fit data. This is a mistake. A good fit reveals nothing about the flexibility of the theory (how much it cannot fit), the variability of the data (how firmly the data rule out what the theory cannot fit), or the likelihood of other outcomes (perhaps the theory could have fit any plausible result), and a reader needs all 3 pieces of information to decide how much the fit should increase belief in the theory. The use of good fits as evidence is not supported by philosophers of science nor by the history of psychology; there seem to be no examples of a theory supported mainly by good fits that has led to demonstrable progress. A better way to test a theory with free parameters is to determine how the theory constrains possible outcomes (i.e., what it predicts), assess how firmly actual outcomes agree with those constraints, and determine if plausible alternative outcomes would have been inconsistent with the theory, allowing for the variability of the data.

View PDF Chat

Observing Interaction: An Introduction to Sequential Analysis

1988-01-01

Victor L. Willson , Roger Bakeman , John M. Gottman

odological concerns. This review is decidedly mixed. The book begins with a discussion of social interaction and observation and quickly moves into a classic study of interaction, Parten's (1932) study … odological concerns. This review is decidedly mixed. The book begins with a discussion of social interaction and observation and quickly moves into a classic study of interaction, Parten's (1932) study of social interaction in children. The issue of sequence versus marginal summation is brought in with argument favoring retention of sequence at all times until independence from sequence is established. Some discussion of the observation-theory issue in the philosophy of science is brought in (see Willson, 1987, for some commentary on this). The second chapter is devoted to developing a coding scheme for observation. It is here that the lack of attention to the reading literature is apparent. Frick and Semmel's (1978) paper is widely cited for development of coding schemes in reading. Researchers in this field have had to grapple with extremely complex issues. Flanders' (1960) work is often cited in education as an early effort, but does not appear in Bakeman and Gottman's book at all. Frick and Semmel pointed researchers to important considerations such as inference level in observation and its development in the coding scheme. This issue is not given nearly the space it requires, especially with the research showing the problems of reliability with highinference observation. The chapter ends with some examples of coding schemes but little practical advice on how to set up the coding schemes and the definitional menus that are absolutely required when several observers other than the developer are to use the system. Chapter 3 discusses recording methods but is notable for its lack of detail

Extracting semantic representations from word co-occurrence statistics: A computational study

2007-08-01

John A. Bullinaria , Joseph P. Levy

View PDF Chat

Why a Diagram is (Sometimes) Worth Ten Thousand Words

1987-01-03

Jill H. Larkin , Herbert A. Simon

We distinguish diagrammatic from sentential paper‐and‐pencil representations of information by developing alternative models of information‐processing systems that are informationally equivalent and that can be characterized as sentential or diagrammatic. Sentential … We distinguish diagrammatic from sentential paper‐and‐pencil representations of information by developing alternative models of information‐processing systems that are informationally equivalent and that can be characterized as sentential or diagrammatic. Sentential representations are sequential, like the propositions in a text. Diagrammatic representations are indexed by location in a plane. Diagrammatic representations also typically display information that is only implicit in sentential representations and that therefore has to be computed, sometimes at great cost, to make it explicit for use. We then contrast the computational efficiency of these representations for solving several illustrative problems in mathematics and physics. When two representations are informationally equivalent, their computational efficiency depends on the information‐processing operators that act on them. Two sets of operators may differ in their capabilities for recognizing patterns, in the inferences they can carry out directly, and in their control strategies (in particular, the control of search). Diagrammatic and sentential representations support operators that differ in all of these respects. Operators working on one representation may recognize features readily or make inferences directly that are difficult to realize in the other representation. Most important, however, are differences in the efficiency of search for information and in the explicitness of information. In the representations we call diagrammatic, information is organized by location, and often much of the information needed to make an inference is present and explicit at a single location. In addition, cues to the next logical step in the problem may be present at an adjacent location. Therefore problem solving can proceed through a smooth traversal of the diagram, and may require very little search or computation of elements that had been implicit.

View PDF Chat

Distributed representations of structure: A theory of analogical access and mapping.

1997-07-01

John E. Hummel , Keith J. Holyoak

We describe an integrated theory of analogical access and mapping, instantiated in a computational model called LISA (Learning and Inference with Schemas and Analogies). LISA represents predicates and objects as … We describe an integrated theory of analogical access and mapping, instantiated in a computational model called LISA (Learning and Inference with Schemas and Analogies). LISA represents predicates and objects as distributed patterns of activation over units representing semantic primitives. These representations are dynamically bound into propositional structures, thereby achieving the structure-sensitivity of a symbolic system and the flexibility of a connectionist system. LISA also has a number of inherent limitations, including capacity limits and sensitivity to the manner in which a problem is represented. A key theoretical claim is that similar limitations also arise in human reasoning, suggesting that the architecture of LISA can provide computational explanations of properties of the human cognitive architecture. We report LISA's performance in simulating a wide range of empirical phenomena concerning human analogical access and mapping. The model treats both access and mapping as types of guided pattern classification, differing only in that mapping is augmented by a capacity to learn new correspondences. Extensions of the approach to account for analogical inference and schema induction are also discussed.

Finding scientific topics

2004-04-06

Thomas L. Griffiths , Mark Steyvers

A first step in identifying the content of a document is determining which topics that document addresses. We describe a generative model for documents, introduced by Blei, Ng, and Jordan … A first step in identifying the content of a document is determining which topics that document addresses. We describe a generative model for documents, introduced by Blei, Ng, and Jordan [Blei, D. M., Ng, A. Y. & Jordan, M. I. (2003) J. Machine Learn. Res. 3, 993-1022], in which each document is generated by choosing a distribution over topics and then choosing each word in the document from a topic selected according to this distribution. We then present a Markov chain Monte Carlo algorithm for inference in this model. We use this algorithm to analyze abstracts from PNAS by using Bayesian model selection to establish the number of topics. We show that the extracted topics capture meaningful structure in the data, consistent with the class designations provided by the authors of the articles, and outline further applications of this analysis, including identifying “hot topics” by examining temporal dynamics and tagging abstracts to illustrate semantic content.

View PDF Chat

Synergy, antagonism, and what the chequerboard puts between them

2003-06-12

Frank C. Odds

A Beginner's Guide to Partial Least Squares Analysis

2004-11-01

Michael Haenlein , Andreas Kaplan

Since the introduction of covariance-based structural equation modeling (SEM) by Joreskog in 1973, this technique has been received with considerable interest among empirical researchers. However, the predominance of LISREL, certainly … Since the introduction of covariance-based structural equation modeling (SEM) by Joreskog in 1973, this technique has been received with considerable interest among empirical researchers. However, the predominance of LISREL, certainly the most well-known tool to perform this kind of analysis, has led to the fact that not all researchers are aware of alternative techniques for SEM, such as partial least squares (PLS) analysis. Therefore, the objective of this article is to provide an easily comprehensible introduction to this technique, which is particularly suited to situations in which constructs are measured by a very large number of indicators and where maximum likelihood covariance-based SEM tools reach their limit. Because this article is intended as a general introduction, it avoids mathematical details as far as possible and instead focuses on a presentation of PLS, which can be understood without an in-depth knowledge of SEM.

Information behaviour: An interdisciplinary perspective

1997-07-01

T. D. Wilson

The text mining handbook: advanced approaches in analyzing unstructured data

2007-06-01

Ronen Feldman , James R. Sanger

1. Introduction to text mining 2. Core text mining operations 3. Text mining preprocessing techniques 4. Categorization 5. Clustering 6. Information extraction 7. Probabilistic models for Information extraction 8. Preprocessing … 1. Introduction to text mining 2. Core text mining operations 3. Text mining preprocessing techniques 4. Categorization 5. Clustering 6. Information extraction 7. Probabilistic models for Information extraction 8. Preprocessing applications using probabilistic and hybrid approaches 9. Presentation-layer considerations for browsing and query refinement 10. Visualization approaches 11. Link analysis 12. Text mining applications Appendix Bibliography.

What Are the Odds? A Practical Guide to Computing and Reporting Bayes Factors

2014-11-07

Andrew F. Jarosz , Jennifer Wiley

The purpose of this paper is to provide an easy template for the inclusion of the Bayes factor in reporting experimental results, particularly as a recommendation for articles in the … The purpose of this paper is to provide an easy template for the inclusion of the Bayes factor in reporting experimental results, particularly as a recommendation for articles in the Journal of Problem Solving. The Bayes factor provides information with a similar purpose to the p-value – to allow the researcher to make statistical inferences from data provided by experiments. While the p-value is widely used, the Bayes factor provides several advantages, particularly in that it allows the researcher to make a statement about the alternative hypothesis, rather than just the null hypothesis. In addition, it provides a clearer estimate of the amount of evidence present in the data. Building on previous work by authors such as Wagenmakers (2007), Rouder et al. (2009), and Masson (2011), this article provides a short introduction to Bayes factors, before providing a practical guide to their computation using examples from published work on problem solving.

View PDF Chat

Evaluation of unsupervised semantic mapping of natural language with Leximancer concept mapping

2006-05-01

Andrew E. Smith , Michael S. Humphreys

View PDF Chat

Index Construction with Formative Indicators: An Alternative to Scale Development

2001-05-01

Adamantios Diamantopoulos , Heidi Winklhofer

Although the methodological literature is replete with advice regarding the development and validation of multi-item scales based on reflective measures, the issue of index construction using formative measures has received … Although the methodological literature is replete with advice regarding the development and validation of multi-item scales based on reflective measures, the issue of index construction using formative measures has received little attention. The authors seek to address this gap by (1) examining the nature of formative indicators, (2) discussing ways in which the quality of formative measures can be assessed, and (3) illustrating the proposed procedures with empirical data. The aim is to enhance researchers' understanding of formative measures and assist them in their index construction efforts.

Specification, evaluation, and interpretation of structural equation models

2011-08-20

Richard P. Bagozzi , Youjae Yi

Improved automatic keyword extraction given more linguistic knowledge

2003-01-01

Anette Hulth

In this paper, experiments on automatic extraction of keywords from abstracts using a supervised machine learning algorithm are discussed. The main point of this paper is that by adding linguistic … In this paper, experiments on automatic extraction of keywords from abstracts using a supervised machine learning algorithm are discussed. The main point of this paper is that by adding linguistic knowledge to the representation (such as syntactic features), rather than relying only on statistics (such as term frequency and n-grams), a better result is obtained as measured by keywords previously assigned by professional indexers. In more detail, extracting NP-chunks gives a better precision than n-grams, and by adding the PoS tag(s) assigned to the term as a feature, a dramatic improvement of the results is obtained, independent of the term selection approach applied.

View PDF Chat

Using Linear Algebra for Intelligent Information Retrieval

1995-12-01

Michael W. Berry , Susan Dumais , Gavin O'Brien

Currently, most approaches to retrieving textual materials from scientific databases depend on a lexical match between words in users’ requests and those in or assigned to documents in a database. … Currently, most approaches to retrieving textual materials from scientific databases depend on a lexical match between words in users’ requests and those in or assigned to documents in a database. Because of the tremendous diversity in the words people use to describe the same document, lexical methods are necessarily incomplete and imprecise. Using the singular value decomposition (SVD), one can take advantage of the implicit higher-order structure in the association of terms with documents by determining the SVD of large sparse term by document matrices. Terms and documents represented by 200–300 of the largest singular vectors are then matched against user queries. We call this retrieval method latent semantic indexing (LSI) because the subspace represents important associative relationships between terms and documents that are not evident in individual documents. LSI is a completely automatic yet intelligent indexing method, widely applicable, and a promising way to improve users’ access to many kinds of textual materials, or to documents and services for which textual descriptions are available. A survey of the computational requirements for managing LSI-encoded databases as well as current and future applications of LSI is presented.

On Relevance, Probabilistic Indexing and Information Retrieval

1960-07-01

M. E. Maron , J. L. Kuhns

This paper reports on a novel technique for literature indexing and searching in a mechanized library system. The notion of relevance is taken as the key concept in the theory … This paper reports on a novel technique for literature indexing and searching in a mechanized library system. The notion of relevance is taken as the key concept in the theory of information retrieval and a comparative concept of relevance is explicated in terms of the theory of probability. The resulting technique called “Probabilistic Indexing,” allows a computing machine, given a request for information, to make a statistical inference and derive a number (called the “relevance number”) for each document, which is a measure of the probability that the document will satisfy the given request. The result of a search is an ordered list of those documents which satisfy the request ranked according to their probable relevance. The paper goes on to show that whereas in a conventional library system the cross-referencing (“see” and “see also”) is based solely on the “semantical closeness” between index terms, statistical measures of closeness between index terms can be defined and computed. Thus, given an arbitrary request consisting of one (or many) index term(s), a machine can elaborate on it to increase the probability of selecting relevant documents that would not otherwise have been selected. Finally, the paper suggests an interpretation of the whole library problem as one where the request is considered as a clue on the basis of which the library system makes a concatenated statistical inference in order to provide as an output an ordered list of those documents which most probably satisfy the information needs of the user.

View PDF Chat

RELEVANCE: A review of and a framework for the thinking on the notion in information science

1975-11-01

Tefko Saračević

Abstract Information science emerged as the third subject, along with logic and philosophy, to deal with relevance‐an elusive, human notion. The concern with relevance, as a key notion in information … Abstract Information science emerged as the third subject, along with logic and philosophy, to deal with relevance‐an elusive, human notion. The concern with relevance, as a key notion in information science, is traced to the problems of scientific communication. Relevance is considered as a measure of the effectiveness of a contact between a source and a destination in a communication process. The different views of relevance that emerged are interpreted and related within a framework of communication of knowledge. Different views arose because relevance was considered at a number of different points in the process of knowledge communication. It is suggested that there exists an interlocking, interplaying cycle of various systems of relevances.

An Investigation of the Relationships between Beliefs about an Object and the Attitude toward that Object

1963-08-01

Martin Fishbein

Probabilistic models of information retrieval based on measuring the divergence from randomness

2002-10-01

Giambattista Amati , Cornelis J. van Rijsbergen

We introduce and create a framework for deriving probabilistic models of Information Retrieval. The models are nonparametric models of IR obtained in the language model approach. We derive term-weighting models … We introduce and create a framework for deriving probabilistic models of Information Retrieval. The models are nonparametric models of IR obtained in the language model approach. We derive term-weighting models by measuring the divergence of the actual term distribution from that obtained under a random process. Among the random processes we study the binomial distribution and Bose--Einstein statistics. We define two types of term frequency normalization for tuning term weights in the document--query matching process. The first normalization assumes that documents have the same length and measures the information gain with the observed term once it has been accepted as a good descriptor of the observed document. The second normalization is related to the document length and to other statistics. These two normalization methods are applied to the basic models in succession to obtain weighting formulae. Results show that our framework produces different nonparametric models forming baseline alternatives to the standard tf-idf model.

View PDF Chat

Accurate methods for the statistics of surprise and coincidence

1993-03-01

Ted Dunning

Much work has been done on the statistical analysis of text. In some cases reported in the literature, inappropriate statistical methods have been used, and statistical significance of results have … Much work has been done on the statistical analysis of text. In some cases reported in the literature, inappropriate statistical methods have been used, and statistical significance of results have not been addressed. In particular, asymptotic normality assumptions have often been used unjustifiably, leading to flawed results.This assumption of normal distribution limits the ability to analyze rare events. Unfortunately rare events do make up a large fraction of real text.However, more applicable methods based on likelihood ratio tests are available that yield good results with relatively small samples. These tests can be implemented efficiently, and have been used for the detection of composite terms and for the determination of domain-specific terms. In some cases, these measures perform much better than the methods previously used. In cases where traditional contingency table methods work well, the likelihood ratio tests described here are nearly identical.This paper describes the basis of a measure based on likelihood ratios that can be applied to the analysis of text.

Untangling text data mining

1999-01-01

Marti A. Hearst

The possibilities for data mining from large text collections are virtually untapped. Text expresses a vast, rich range of information, but encodes this information in a form that is difficult … The possibilities for data mining from large text collections are virtually untapped. Text expresses a vast, rich range of information, but encodes this information in a form that is difficult to decipher automatically. Perhaps for this reason, there has been little work in text data mining to date, and most people who have talked about it have either conflated it with information access or have not made use of text directly to discover heretofore unknown information.

View PDF Chat

Automatic Keyword Extraction from Individual Documents

2010-03-04

Stuart Rose , Dave Engel , Nick Cramer +1 more

Keywords are widely used to define queries within information retrieval (IR) systems as they are easy to define, revise, remember, and share. This chapter describes the rapid automatic keyword extraction … Keywords are widely used to define queries within information retrieval (IR) systems as they are easy to define, revise, remember, and share. This chapter describes the rapid automatic keyword extraction (RAKE), an unsupervised, domain-independent, and language-independent method for extracting keywords from individual documents. It provides details of the algorithm and its configuration parameters, and present results on a benchmark dataset of technical abstracts, showing that RAKE is more computationally efficient than TextRank while achieving higher precision and comparable recall scores. The chapter then describes a novel method for generating stoplists, which is used to configure RAKE for specific domains and corpora. Finally, it applies RAKE to a corpus of news articles and defines metrics for evaluating the exclusivity, essentiality, and generality of extracted keywords, enabling a system to identify keywords that are essential or general to documents in the absence of manual annotations. Controlled Vocabulary Terms benchmark polls

View PDF Chat

Lexical storage and retrieval of prefixed words

1975-12-01

Marcus Taft , Kenneth I. Forster

Inverted files for text search engines

2006-07-25

Justin Zobel , Alistair Moffat

The technology underlying text search engines has advanced dramatically in the past decade. The development of a family of new index representations has led to a wide range of innovations … The technology underlying text search engines has advanced dramatically in the past decade. The development of a family of new index representations has led to a wide range of innovations in index storage, index construction, and query evaluation. While some of these developments have been consolidated in textbooks, many specific techniques are not widely known or the textbook descriptions are out of date. In this tutorial, we introduce the key techniques in the area, describing both a core implementation and how the core can be enhanced through a range of extensions. We conclude with a comprehensive bibliography of text indexing literature.

Short-term Memory for Word Sequences as a Function of Acoustic, Semantic and Formal Similarity

1966-11-01

Alan Baddeley

A STATISTICAL INTERPRETATION OF TERM SPECIFICITY AND ITS APPLICATION IN RETRIEVAL

1972-01-01

Karen Spärck Jones

Abstract The exhaustivity of document descriptions and the specificity of index terms are usually regarded as independent. It is suggested that specificity should be interpreted statistically, as a function of … Abstract The exhaustivity of document descriptions and the specificity of index terms are usually regarded as independent. It is suggested that specificity should be interpreted statistically, as a function of term use rather than of term meaning. The effects on retrieval of variations in term specificity are examined, experiments with three test collections showing in particular that frequently‐occurring terms are required for good overall performance. It is argued that terms should be weighted according to collection frequency, so that matches on less frequent, more specific, terms are of greater value than matches on frequent terms. Results for the test collections show that considerable improvements in performance are obtained with this very simple procedure.

Indexing by latent semantic analysis

1990-09-01

Scott Deerwester , Susan Dumais , George W. Furnas +2 more

A new method for automatic indexing and retrieval is described. The approach is to take advantage of implicit higher-order structure in the association of terms with documents ("semantic structure") in … A new method for automatic indexing and retrieval is described. The approach is to take advantage of implicit higher-order structure in the association of terms with documents ("semantic structure") in order to improve the detection of relevant documents on the basis of terms found in queries. The particular technique used is singular-value decomposition, in which a large term by document matrix is decomposed into a set of ca. 100 orthogonal factors from which the original matrix can be approximated by linear combination. Documents are represented by ca. 100 item vectors of factor weights. Queries are represented as pseudo-document vectors formed from weighted combinations of terms, and documents with supra-threshold cosine values are returned. Initial tests find this completely automatic method for retrieval to be promising. © 1990 John Wiley & Sons, Inc.

View PDF Chat

Factors In The Natural Language Of Personality: Re-Analysis, Comparison, And Interpretation Of Six Major Studies

1981-04-01

John M. Digman , Naomi K. Takemoto-Chock

Psychometric studies of the organization of the "natural language of personality" have typically employed rating scales as measurement medium and factor analysis as statistical technique. The results of such investigations … Psychometric studies of the organization of the "natural language of personality" have typically employed rating scales as measurement medium and factor analysis as statistical technique. The results of such investigations over the past 30 years have varied greatly, both with respect to number of factors and with respect to the constructs generated. Re-analysis of the correlations of six studies, including the classical work of Cattell, indicated that the domain appears to be well described by five factors, with some suggestion of a sixth. The five factors were related across studies, using the Kaiser-Hunka-Bianchini method. Generally, the factors were highly related, with most indices of relatedness exceeding .90. The five-factor model was tested by the multiple-group method, used to factor a large-scale study of teachers' ratings of children. With slight modification of the originally hypothesized structure, the five-factor model accounted for the observed relationships quite well. The five constructs suggested by the factors appear to be domains of research effort and theoretical concern which have long been of interest to psychologists.

Role of Product-Related Conversations in the Diffusion of a New Product

1967-08-01

Johan Arndt

This article reports an experiment designed to investigate the short-term sales effects of product-related conversations. The results show that exposure to favorable comments aids acceptance of a new product, while … This article reports an experiment designed to investigate the short-term sales effects of product-related conversations. The results show that exposure to favorable comments aids acceptance of a new product, while unfavorable comments hinder it.

A Parsimonious Estimating Technique for Interaction and Quadratic Latent Variables

1995-08-01

Robert A. Ping

The author proposes an alternative estimation technique for quadratic and interaction latent variables in structural equation models using LISREL, EQS, and CALIS. The technique specifies these variables with single indicants. … The author proposes an alternative estimation technique for quadratic and interaction latent variables in structural equation models using LISREL, EQS, and CALIS. The technique specifies these variables with single indicants. The loading and error terms for the single indicants can be specified as constants in the structural model. The author's technique is shown to perform adequately using synthetic data sets.

Probabilistic Topic Models

2015-02-17

Mark Steyvers , Thomas L. Griffiths

Many chapters in this book illustrate that applying a statistical method such as latent semantic analysis (LSA; Landauer & Dumais, 1997; Landauer, Foltz, & Laham, 1998) to large databases can … Many chapters in this book illustrate that applying a statistical method such as latent semantic analysis (LSA; Landauer & Dumais, 1997; Landauer, Foltz, & Laham, 1998) to large databases can yield insight into human cognition. The LSA approach makes three claims: that semantic information can be derived from a word-document co-occurrence matrix; that dimensionality reduction is an essential part of this derivation; and that words and documents can be represented as points in Euclidean space. This chapter pursues an approach that is consistent with the first two of these claims, but differs in the third, describing a class of statistical models in which the semantic properties of words and documents are expressed in terms of probabilistic topics.

Estimation issues with PLS and CBSEM: Where the bias lies!

2016-06-25

Marko Sarstedt , Joseph F. Hair , Christian M. Ringle +2 more

Discussions concerning different structural equation modeling methods draw on an increasing array of concepts and related terminology. As a consequence, misconceptions about the meaning of terms such as reflective measurement … Discussions concerning different structural equation modeling methods draw on an increasing array of concepts and related terminology. As a consequence, misconceptions about the meaning of terms such as reflective measurement and common factor models as well as formative measurement and composite models have emerged. By distinguishing conceptual variables and their measurement model operationalization from the estimation perspective, we disentangle the confusion between the terminologies and develop a unifying framework. Results from a simulation study substantiate our conceptual considerations, highlighting the biases that occur when using (1) composite-based partial least squares path modeling to estimate common factor models, and (2) common factor-based covariance-based structural equation modeling to estimate composite models. The results show that the use of PLS is preferable, particularly when it is unknown whether the data's nature is common factor- or composite-based.

View PDF Chat

Variance-Based Structural Equation Modeling

2012-01-01

José L. Roldán , Manuel J. Sánchez‐Franco

Partial Least Squares (PLS) is an efficient statistical technique that is highly suited for Information Systems research. In this chapter, the authors propose both the theory underlying PLS and a … Partial Least Squares (PLS) is an efficient statistical technique that is highly suited for Information Systems research. In this chapter, the authors propose both the theory underlying PLS and a discussion of the key differences between covariance-based SEM and variance-based SEM, i.e., PLS. In particular, authors: (a) provide an analysis of the origin, development, and features of PLS, and (b) discuss analysis problems as diverse as the nature of epistemic relationships and sample size requirements. In this regard, the authors present basic guidelines for the applying of PLS as well as an explanation of the different steps implied for the assessment of the measurement model and the structural model. Finally, the authors present two examples of Information Systems models in which they have put previous recommendations into effect.

Minimum sample size estimation in PLS‐SEM: The inverse square root and gamma‐exponential methods

2016-11-29

Ned Kock , Pierre Hadaya

Abstract Partial least squares‐based structural equation modelling (PLS‐SEM) is extensively used in the field of information systems, as well as in many other fields where multivariate statistical methods are used. … Abstract Partial least squares‐based structural equation modelling (PLS‐SEM) is extensively used in the field of information systems, as well as in many other fields where multivariate statistical methods are used. One of the most fundamental issues in PLS‐SEM is that of minimum sample size estimation. The ‘10‐times rule’ has been a favourite because of its simplicity of application, even though it tends to yield imprecise estimates. We propose two related methods, based on mathematical equations, as alternatives for minimum sample size estimation in PLS‐SEM: the inverse square root method, and the gamma‐exponential method. Based on three Monte Carlo experiments, we demonstrate that both methods are fairly accurate. The inverse square root method is particularly attractive in terms of its simplicity of application. © 2016 John Wiley & Sons Ltd

quanteda: An R package for the quantitative analysis of textual data

2018-10-06

Kenneth Benoit , Kohei Watanabe , H. P. Wang +4 more

quanteda is an R package providing a comprehensive workflow and toolkit for natural language processing tasks such as corpus management, tokenization, analysis, and visualization.It has extensive functions for applying dictionary … quanteda is an R package providing a comprehensive workflow and toolkit for natural language processing tasks such as corpus management, tokenization, analysis, and visualization.It has extensive functions for applying dictionary analysis, exploring texts using keywords-in-context, computing document and feature similarities, and discovering multi-word expressions through collocation scoring.Based entirely on sparse operations, it provides highly efficient methods for compiling document-feature matrices and for manipulating these or using them in further quantitative analysis.Using C++ and multithreading extensively, quanteda is also considerably faster and more efficient than other R and Python packages in processing large textual data.

View PDF Chat

Using Information Content to Evaluate Semantic Similarity in a Taxonomy

1995-11-29

Philip Resnik

This paper presents a new measure of semantic similarity in an IS-A taxonomy, based on the notion of information content. Experimental evaluation suggests that the measure performs encouragingly well (a … This paper presents a new measure of semantic similarity in an IS-A taxonomy, based on the notion of information content. Experimental evaluation suggests that the measure performs encouragingly well (a correlation of r = 0.79 with a benchmark set of human similarity judgments, with an upper bound of r = 0.90 for human subjects performing the same task), and significantly better than the traditional edge counting approach (r = 0.66).

The structure and dynamics of cocitation clusters: A multiple‐perspective cocitation analysis

2010-03-20

Chaomei Chen , Fidelia Ibekwe-Sanjuan , Jianhua Hou

A multiple-perspective co-citation analysis method is introduced for characterizing and interpreting the structure and dynamics of co-citation clusters. The method facilitates analytic and sense making tasks by integrating network visualization, … A multiple-perspective co-citation analysis method is introduced for characterizing and interpreting the structure and dynamics of co-citation clusters. The method facilitates analytic and sense making tasks by integrating network visualization, spectral clustering, automatic cluster labeling, and text summarization. Co-citation networks are decomposed into co-citation clusters. The interpretation of these clusters is augmented by automatic cluster labeling and summarization. The method focuses on the interrelations between a co-citation cluster's members and their citers. The generic method is applied to a three-part analysis of the field of Information Science as defined by 12 journals published between 1996 and 2008: 1) a comparative author co-citation analysis (ACA), 2) a progressive ACA of a time series of co-citation networks, and 3) a progressive document co-citation analysis (DCA). Results show that the multiple-perspective method increases the interpretability and accountability of both ACA and DCA networks.

View PDF Chat

Lateral Collinearity and Misleading Results in Variance-Based SEM: An Illustration and Recommendations

2012-07-01

Ned Kock , Gary S. Lynn

Variance-based structural equation modeling is extensively used in information systems research, and many related findings may have been distorted by hidden collinearity. This is a problem that may extend to … Variance-based structural equation modeling is extensively used in information systems research, and many related findings may have been distorted by hidden collinearity. This is a problem that may extend to multivariate analyses, in general, in the field of information systems as well as in many other fields. In multivariate analyses, collinearity is usually assessed as a predictor-predictor relationship phenomenon, where two or more predictors are checked for redundancy. This type of assessment addresses vertical, or “classic”, collinearity. However, another type of collinearity may also exist, here called “lateral” collinearity. It refers to predictor-criterion collinearity. Lateral collinearity problems are exemplified based on an illustrative variance-based structural equation modeling analysis. The analysis employs WarpPLS 2.0, with the results double-checked with other statistical analysis software tools. It is shown that standard validity and reliability tests do not properly capture lateral collinearity. A new approach for the assessment of both vertical and lateral collinearity in variance-based structural equation modeling is proposed and demonstrated in the context of the illustrative analysis.

View PDF Chat

Scatter/Gather: a cluster-based approach to browsing large document collections

1992-01-01

Douglass R. Cutting , David R. Karger , Jan Pedersen +1 more

Document clustering has not been well received as an information retrieval tool. Objections to its use fall into two main categories: first, that clustering is too slow for large corpora … Document clustering has not been well received as an information retrieval tool. Objections to its use fall into two main categories: first, that clustering is too slow for large corpora (with running time often quadratic in the number of documents); and second, that clustering does not appreciably improve retrieval.

View PDF Chat

On the evaluation of structural equation models

1988-03-01

Richard P. Bagozzi , Youjae Yi

Data Classification: Algorithms and Applications

2014-07-25

Charų C. Aggarwal

Comprehensive Coverage of the Entire Area of Classification Research on the problem of classification tends to be fragmented across such areas as pattern recognition, database, data mining, and machine learning. … Comprehensive Coverage of the Entire Area of Classification Research on the problem of classification tends to be fragmented across such areas as pattern recognition, database, data mining, and machine learning. Addressing the work of these different communities in a unified way, Data Classification: Algorithms and Applications explores the underlying algorithms of classification as well as applications of classification in a variety of problem domains, including text, multimedia, social network, and biological data. This comprehensive book focuses on three primary aspects of data classification: Methods-The book first describes common techniques used for classification, including probabilistic methods, decision trees, rule-based methods, instance-based methods, support vector machine methods, and neural networks. Domains-The book then examines specific methods used for data domains such as multimedia, text, time-series, network, discrete sequence, and uncertain data. It also covers large data sets and data streams due to the recent importance of the big data paradigm. Variations-The book concludes with insight on variations of the classification process. It discusses ensembles, rare-class learning, distance function learning, active learning, visual learning, transfer learning, and semi-supervised learning as well as evaluation aspects of classifiers.

Linear Sequential Recommendation Models using Textual Side Information

2025-06-24

Dongcheol Lee , Minjin Choi , Jongwuk Lee | Journal of KIISE

Exploring Multivariate Data: Models Based on Latent Variables

2025-06-24

M.S. Sánchez , M.C. Ortiz , L.A. Sarabia | Essential textbooks in chemistry

The visual, the textual, and the one‐dimensional: An exploration of the visual elements of bibliographic classification schemes

2025-06-20

Deborah Lee | Journal of the Association for Information Science and Technology

Abstract Classification schemes are a key way of organizing bibliographic knowledge, yet the way that classification schemes communicate their information to classifiers receives little attention. This article takes a novel … Abstract Classification schemes are a key way of organizing bibliographic knowledge, yet the way that classification schemes communicate their information to classifiers receives little attention. This article takes a novel approach by exploring the visual aspects contained within classification schemes. The research uses a classification scheme analysis methodology. Three different classification scheme phenomena are discussed in terms of their visualization: hierarchy, notation, and notes. Indentation is found to be a significant—and implicit—method of communicating hierarchy to classifiers and offers intriguing solutions to the issues of transmuting from two dimensions into one. The visual elements of notation reveal a strong separation between notation and class, while the visual elements of notes illuminate a varying narrative around the position of notes in the classification scheme . A categorization system for visual elements in classification schemes is presented. Model 1 proffers visual elements as a fourth plane of classification, which extends and remodels Ranganathan's Three Planes of Work . Model 2 shows how visual elements could fit into classification scheme versioning. Ultimately, looking at visual aspects of classification schemes is a novel way of thinking about knowledge organization and can help us to better understand—and ultimately, to better use—classification schemes.

Quantum-like evidence networks decision-making model

2025-06-20

Xiaozhuan Gao , Huijun Yang , Lipeng Pan +1 more | Engineering Applications of Artificial Intelligence

Abstractive Text Summarization in English Using Swin Transformer Encoder based Recurrent Neural Network

2025-06-20

| International journal of intelligent engineering and systems

The Risks of Declining Theory Construction and Theorisation and Rising Empiricism in the Discipline of Information Science: An Overview

2025-06-19

Waseem Afzal | Journal of the Australian Library and Information Association

A Multimodal Semantic Fusion Network with Cross-Modal Alignment for Multimodal Sentiment Analysis

2025-06-17

Shunxiang Zhang , Jiajia Liu , Yuyang Jiao +3 more | ACM Transactions on Multimedia Computing Communications and Applications

User-generated multimodal data can provide powerful sentiment clues for sentiment analysis task. Existing works have aligned common sentiment features in different modalities through various multimodal fusion methods. However, these works … User-generated multimodal data can provide powerful sentiment clues for sentiment analysis task. Existing works have aligned common sentiment features in different modalities through various multimodal fusion methods. However, these works have certain limitations: (1) Previous research works only align common sentiment features between image and text, without fully exploring interactions among these features, leading to suboptimal analysis results. (2) Redundant noise in image and text increases the risk of feature misalignment during cross-modal alignment. To address these issues, we propose a multimodal semantic fusion network (MSFN) to deeply explore the semantic relationship between image and text for Multimodal Sentiment Analysis (MSA). Specifically, we align image region and text word features related to sentiment by using a gated attention mechanism. Subsequently, we employ graph convolutional networks to model the interactions among these features to obtain explicit sentiment semantics. The proposed gated attention mechanism corrects potential feature misalignment during cross-modal alignment using a gating mechanism. Moreover, considering not all image-text pairs have explicit corresponding sentiment features, we integrate implicit sentiment semantics to our model for enhancing reliability in analysis. Experimental results on benchmark datasets demonstrate the effectiveness of our proposed model compared to baselines.

Section Recommendation of Online Medical Platform Based on Keyword Expansion with Self-Adaptive-Attention-Prompt-BERT-RCNN Modeling

2025-06-16

Tianbao Xie , Yuqi Han , Ganglong Duan +3 more | Applied Sciences

Background: Implementing automatic classification of short texts in online healthcare platforms is crucial to increase the efficiency of their services and improve the user experience. A short text classification method … Background: Implementing automatic classification of short texts in online healthcare platforms is crucial to increase the efficiency of their services and improve the user experience. A short text classification method combining the keyword expansion technique and a deep learning model is constructed to solve the problems of feature sparsity and semantic ambiguity in short text classification. Methods: First, we use web crawlers to obtain patient data from the online medical platform “Good Doctor”; then, we use TF-IWF to weight the keyword importance and Word2vec to calculate the keyword similarity to expand the short text features; and then we integrate the cue learning and deep learning models to construct a self-adaptive attention model to solve the problem of sparse features and unclear semantics in short text classification in the adaptive-attention-Prompt-BERT-RCNN model to realize effective classification of medical short texts. Results: Empirical studies show that the classification effect after keyword expansion is significantly higher than that before expansion, the accuracy of the model in classifying medical short texts after expansion is as high as 97.84%, and the model performs well in different categories of medical short texts. Conclusions: The short text expansion methods of TF-IWF and Word2vec make up for the shortcomings of not taking into account the keyword rarity and the contextual information of the subwords, and the model can achieve effective classification of medical short texts by combining them. The model further improves the classification accuracy of short text by integrating Prompt’s bootstrapping, self-adaptive attention’s keyword weight weighting, BERT’s deep semantic understanding, and RCNN’s region awareness and feature extraction; however, the model’s accuracy in individual topics still needs to be improved. The results show that the recommender system can effectively improve the efficiency of patient consultation and support the development of online healthcare.

Neurosymbolic Information Extraction from Transactional Documents

2025-06-16

Arthur Hemmer , Mickaël Coustaty , Nicola Bartolo +1 more | International Journal on Document Analysis and Recognition (IJDAR)

Citation structural diversity: a novel metric combining structure and semantics for literature evaluation

2025-06-16

Mingyue Kong , Yinglong Zhang , Sheng Li-kun +1 more | Scientometrics

Newspaper Summarizer using NLP and ML

2025-06-16

Ojasvi Sanjay More | INTERANTIONAL JOURNAL OF SCIENTIFIC RESEARCH IN ENGINEERING AND MANAGEMENT

Abstract—In the era of digital information, users are inundated with news articles from numerous sources, resulting in information overload and an overwhelming user experience. This research presents an advanced, real-time … Abstract—In the era of digital information, users are inundated with news articles from numerous sources, resulting in information overload and an overwhelming user experience. This research presents an advanced, real-time Newspaper Aggregator that utilizes Natural Language Processing (NLP) and Machine Learning (ML) techniques to collect, process, and personalize news articles from diverse sources in real-time. The aggregator’s architecture integrates several NLP models to achieve comprehensive news handling: topic modeling categorizes articles into predefined topics such as Politics, Sports, and Technology using Latent Dirichlet Allocation (LDA), while sentiment analysis, powered by BERT, classifies public sentiment as Positive, Negative, or Neutral, capturing nuanced perspectives. The system’s summarization module leverages PEGASUS and Text Rank to deliver coherent, concise summaries, improving information accessibility and reducing reading time. Additionally, the recommendation engine employs a hybrid filtering approach, combining collaborative and content-based filtering, to provide personalized news recommendations based on user history and article characteristics. Our methodology includes systematic data collection, text pre-processing, topic categorization, sentiment classification, summarization, and real-time recommendation, followed by rigorous evaluation. The aggregator achieves high accuracy across tasks: BERT-driven sentiment analysis achieves 92% accuracy, LDA models yield coherent topic clusters, and summarization evaluations produce a ROUGE-L score of 0.75, all of which underscore the system's reliability in managing dynamic news content. Performance testing indicates that this Newspaper Aggregator offers a significant improvement in user relevance and engagement compared to traditional keyword-based systems. Overall, this study establishes a foundation for intelligent, real-time news aggregation, providing users with a streamlined, personalized news experience. Keywords— Real-time news aggregation, Natural Language Processing (NLP), Machine Learning (ML), topic modeling, sentiment analysis, BERT, Latent Dirichlet Allocation (LDA), text summarization, PEGASUS, Text Rank, recommendation systems, collaborative filtering, content-based filtering, personalized news, information overload, news categorization, user relevance, article classification, hybrid recommendation model.

Makine Öğrenmesi ve Derin Öğrenmeye Dayalı Duygu Analizinde Metin Temsil Yöntemlerinin Sınıflandırma Başarımına Etkisinin İncelenmesi

2025-06-14

Aleyna Kahraman , Durmuş Özkan Şahın , Dilara Bıyıklı +2 more | Karadeniz Fen Bilimleri Dergisi

Bu çalışmada, iki farklı veri seti üzerinde çeşitli metin temsil yöntemleri kullanılarak ikili ve üçlü sınıflandırma işlemleri gerçekleştirilmiştir. Metin temsil yöntemleri olarak TF-IDF, GloVe, Word2Vec, FastText ve Bag of Words … Bu çalışmada, iki farklı veri seti üzerinde çeşitli metin temsil yöntemleri kullanılarak ikili ve üçlü sınıflandırma işlemleri gerçekleştirilmiştir. Metin temsil yöntemleri olarak TF-IDF, GloVe, Word2Vec, FastText ve Bag of Words kullanılmıştır. Makine öğrenimi algoritmalarından Naive Bayes, Lojistik Regresyon, Destek Vektör Makineleri, Rastgele Orman, Yapay Sinir Ağı, En Yakın Komşu Algoritması, Karar Ağacı, XGBoost ve LightGBM uygulanmıştır. Derin öğrenme algoritmaları olarak ise Evrişimli Sinir Ağı, Tekrarlayan Sinir Ağı ve Uzun Kısa Süreli Bellek kullanılmıştır. Elde edilen sonuçlarla, kullanılan metin temsil yöntemleri ve algoritmaların performansları karşılaştırılmıştır. Amazon veri setinde, makine öğrenimi yöntemleri arasında en yüksek doğruluk oranı LightGBM algoritması, derin öğrenme yöntemleri arasında ise TF-IDF ve FastText kullanan LSTM algoritması tarafından elde edilmiştir. IMDb veri setinde, makine öğrenimi yöntemleri arasında en yüksek doğruluk oranı Lojistik Regresyon algoritması, derin öğrenme yöntemleri arasında ise FastText kullanan LSTM algoritması tarafından elde edilmiştir.

Utilization Of TF-IDF Weighting In Song Search System Based On Spotify Lyrics

2025-06-14

Nelti Juliana Sahera , Eviriawan , Hikma +1 more | Journal of Artificial Intelligence and Engineering Applications (JAIEA)

In the rapidly developing digital era, the need for an efficient information retrieval system is increasing. Spotify, as one of the largest music streaming platforms, faces challenges in providing a … In the rapidly developing digital era, the need for an efficient information retrieval system is increasing. Spotify, as one of the largest music streaming platforms, faces challenges in providing a fast and accurate song search system. Improving user experience in searching for song titles based on lyrics is the main focus in developing a search system on the music streaming platform. like Spotify. Study This explore use method weighting using TF-IDF (Term Frequency- Inverse Document Frequency) to optimize the search for song titles through lyrics. By applying TF-IDF, system can assess and weighting words in lyrics based on the frequency in One song and its uniqueness in gathering song data in overall. As for the data that used in this study totaling 30 entries. The methods used include system design, preprocessing (data cleaning, tokenization, filtering, and stemming), and TF-IDF weighting. The test results show that this approach significantly improves the relevance and accuracy of search results, making it easier for users to find the appropriate song title. with lyrics Which they remember. System Which proposed This expected can repair quality search services on Spotify and provide a more satisfying experience for users.

A type-theoretical approach to categorical interaction and complex system adaptation: an empirical study based on register classification

2025-06-13

Renkui Hou , Chu‐Ren Huang , Kathleen Ahrens | Humanities and Social Sciences Communications

A rhetorical structure theory and inference-aware graph network for emotion-cause pair extraction

2025-06-13

Qi Zhang , Botao Wang , Peican Zhu +2 more | Complex & Intelligent Systems

Unsupervised Discovery of Salient Design Features of Websites

2025-06-13

Thisaranie Kaluarachchi , Dushan Dissanayake , Manjusri Wickramasinghe | International Journal on Advances in ICT for Emerging Regions (ICTer)

Web development is one of the fastest-growing fields in the IT industry. User Interface (UI) design of a website is critical in attracting new users, which helps businesses increase sales … Web development is one of the fastest-growing fields in the IT industry. User Interface (UI) design of a website is critical in attracting new users, which helps businesses increase sales and revenue. A unique website design will encourage user interaction among website visitors and ensure that the time and resources spent on a webpage are worthwhile. Web designers create websites either by using pre-existing templates or by building them from scratch. The web designer's design skills heavily influence the overall appearance of a website. However, such websites do not always meet the client's expectations. As a result of these challenges and the ever-changing web development trends, the automatic website generation concept has emerged, which generates websites without relying on human interaction. In this concept, it is useful to understand how to classify websites based on their appearance and how to identify design features that distinguish websites. This study aims to develop a classification system for websites on the internet based on their salient design features.

Utilizing Association Rules in Knowledge Graphs for Enhanced News Summarization

2025-06-13

Thilini Lakshika , Amitha Caldera | International Journal on Advances in ICT for Emerging Regions (ICTer)

The rapid progress in web news articles has led to an abundance of text content, often than needed, and consequently, misleading readers. Recent Knowledge Graph (KG) based approaches have proven … The rapid progress in web news articles has led to an abundance of text content, often than needed, and consequently, misleading readers. Recent Knowledge Graph (KG) based approaches have proven successful in abstract summary generation due to their ability to represent structured and interconnected knowledge with semantic context. The KG ranking algorithm responsible for selecting graph data for inclusion in the abstract still relies on traditional ranking algorithms which lack the consideration for semantic relationships between graph nodes, and are associated with high memory consumption, processing times, and increased complexity. Knowledge discovery plays a crucial role in improving the quality of summarization by uncovering hidden patterns and enhancing contextual understanding. Therefore, our study centers on introducing a novel KG ranking algorithm, aimed at a statistically significant enhancement in abstract generation by integrating knowledge discovery techniques. The suggested ranking algorithm considers the semantic and topological graph properties and interesting relationships, patterns, and features in text data using Association Rule Mining techniques to identify the most significant graph information for generating abstracts. The experiments conducted using the DUC-2002 dataset indicate that the suggested KG ranking algorithm is effective in producing detailed and accurate abstracts for a collection of web news articles.

Automation of Systematic Reviews with Large Language Models

2025-06-13

Christian Cao , Rohit Arora , Paul Cento +7 more | medRxiv (Cold Spring Harbor Laboratory)

Systematic reviews (SRs) inform evidence-based decision making. Yet, they take over a year to complete, are prone to human error, and face challenges with reproducibility; limiting access to timely and … Systematic reviews (SRs) inform evidence-based decision making. Yet, they take over a year to complete, are prone to human error, and face challenges with reproducibility; limiting access to timely and reliable information. We developed otto-SR, an end-to-end agentic workflow using large language models (LLMs) to support and automate the SR workflow from initial search to analysis. We found that otto-SR outperformed traditional dual human workflows in SR screening (otto-SR: 96.7% sensitivity, 97.9% specificity; human: 81.7% sensitivity, 98.1% specificity) and data extraction (otto-SR: 93.1% accuracy; human: 79.7% accuracy). Using otto-SR, we reproduced and updated an entire issue of Cochrane reviews (n=12) in two days, representing approximately 12 work-years of traditional systematic review work. Across Cochrane reviews, otto-SR incorrectly excluded a median of 0 studies (IQR 0 to 0.25), and found a median of 2.0 (IQR 1 to 6.5) eligible studies likely missed by the original authors. Meta-analyses revealed that otto-SR generated newly statistically significant conclusions in 2 reviews and negated significance in 1 review. These findings demonstrate that LLMs can autonomously conduct and update systematic reviews with superhuman performance, laying the foundation for automated, scalable, and reliable evidence synthesis.

Accurate short-read alignment through<i>r</i>-index-based pangenome indexing

2025-06-12

Rahul Varki , Massimiliano Rossi , Eddie Ferro +4 more | Genome Research

Aligning to a linear reference genome can result in a higher percentage of reads going unmapped or being incorrectly mapped owing to variations not captured by the reference, otherwise known … Aligning to a linear reference genome can result in a higher percentage of reads going unmapped or being incorrectly mapped owing to variations not captured by the reference, otherwise known as reference bias. Recently, in efforts to mitigate reference bias, there has been a movement to switch to using pangenomes, a collection of genomes, as the reference. In this paper, we introduce Moni-align, the first short-read pangenome aligner built on the r -index, a variation of the classical FM-index that can index collections of genomes in O( r )-space, where r is the number of runs in the Burrows–Wheeler transform. Moni-align uses a seed-and-extend strategy for aligning reads, utilizing maximal exact matches as seeds, which can be efficiently obtained with the r -index. Using both simulated and real short-read data sets, we demonstrate that Moni-align achieves alignment accuracy comparable to vg map and vg giraffe, the leading pangenome aligners. Although currently best suited for aligning to localized pangenomes owing to computational constraints, Moni-align offers a robust foundation for future optimizations that could further broaden its applicability.

Beyond performance: How design choices shape chemical language models

2025-06-12

Inken Fender , Jannik Andrian Gut , Thomas Lemmin | Research Square (Research Square)

<title>Abstract</title> Chemical language models (CLMs) have shown strong performance in molecular property prediction and generation tasks. However, the impact of design choices, such as molecular representation format, tokenization strategy, and … <title>Abstract</title> Chemical language models (CLMs) have shown strong performance in molecular property prediction and generation tasks. However, the impact of design choices, such as molecular representation format, tokenization strategy, and model architecture, on both performance and chemical interpretability remains underexplored. In this study, we systematically evaluate how these factors influence CLM performance and chemical understanding. We evaluated models through fine-tuning on downstream tasks and probing the structure of their latent spaces using simple classifiers and dimensionality reduction techniques.Despite similar performance on downstream tasks across model configurations, we observed substantial differences in the structure and interpretability of their internal representations. SMILES molecular representation format with atomwise tokenization strategy consistently produced more chemically meaningful embeddings, while models based on BART and RoBERTa architectures yielded comparably interpretable representations. These findings highlight that design choices meaningfully shape how chemical information is represented, even when external metrics appear unchanged. This insight can inform future model development, encouraging more chemically grounded and interpretable CLMs.

LLM-Cure: LLM-based Competitor User Review Analysis for Feature Enhancement

2025-06-12

Maram Assi , Safwat Hassan , Ying Zou | ACM Transactions on Software Engineering and Methodology

The exponential growth of the mobile app market underscores the importance of constant innovation and rapid response to user demands. As user satisfaction is paramount to the success of a … The exponential growth of the mobile app market underscores the importance of constant innovation and rapid response to user demands. As user satisfaction is paramount to the success of a mobile application (app), developers typically rely on user reviews, which represent user feedback that includes ratings and comments to identify areas for improvement. However, the sheer volume of user reviews poses challenges in manual analysis, necessitating automated approaches. Existing automated approaches either analyze only the target app's reviews, neglecting the comparison of similar features to competitors or fail to provide suggestions for feature enhancement. To address these gaps, we propose a Large Language Model (LLM) -based C ompetitive U ser R eview Analysis for Feature E nhancement) ( LLM-Cure ), an approach powered by LLMs to automatically generate suggestions for mobile app feature improvements. More specifically, LLM-Cure identifies and categorizes features within reviews by applying LLMs. When provided with a complaint in a user review, LLM-Cure curates highly rated (4 and 5 stars) reviews in competing apps related to the complaint and proposes potential improvements tailored to the target application. We evaluate LLM-Cure on 1,056,739 reviews of 70 popular Android apps. Our evaluation demonstrates that LLM-Cure significantly outperforms the state-of-the-art approaches in assigning features to reviews by up to 13% in F1-score, up to 16% in recall and up to 11% in precision. Additionally, LLM-Cure demonstrates its capability to provide suggestions for resolving user complaints. We verify the suggestions using the release notes that reflect the changes of features in the target mobile app. LLM-Cure achieves a promising average of 73% of the implementation of the provided suggestions, demonstrating its potential for competitive feature enhancement.

WordMap: Text Mining Application of Enhanced Corpus Segmentation and Semantic Topic Recognition

2025-06-12

Zhijian Wei , Huiwen Zou , Patrick Cheong-Iao Pang +2 more | Applied Sciences

This study presents WordMap, an integrated text mining application developed to enhance the efficiency and usability of text analysis over a network. As unstructured text data continues to grow across … This study presents WordMap, an integrated text mining application developed to enhance the efficiency and usability of text analysis over a network. As unstructured text data continues to grow across domains, effective tools for segmentation and topic modeling have become increasingly essential for extracting insightful information. However, most existing solutions depend on multiple disconnected tools, and these often compromise workflow efficiency and user experience. Unlike traditional tools, WordMap combines corpus segmentation, topic modeling, and result visualization into a unified workflow for both Chinese and English languages, thereby reducing workflow fragmentation and lowering the user threshold. To assess usability and user acceptance, this research adopts the Technology Acceptance Model (TAM). WordMap employs PKUSEG and NLTK for bilingual corpus segmentation, utilizes BERTopic for dynamic topic modeling, and integrates interactive visualization to enable intuitive analysis. The PLS-SEM result shows that the perceived ease of use (PEOU) has a significant impact on both perceived usefulness (PU) and user attitude (ATT), while ATT strongly predicts behavioral intention (BI) (β = 0.674, p < 0.001). The results indicate that integrating core text mining processes into a user-centered design significantly boosts user satisfaction and adoption. By combining key processes and empirically validating user perceptions, the proposed framework facilitates the development of efficient and accessible text mining tools. It offers both theoretical and practical insights for future advancement and deployment in the field of text mining.

Examining Lower Secondary School Students’ Misunderstandings of Decision Trees: A Distractor Analysis

2025-06-10

Katherine S. Moore , Beatriz Perret , Baiyu Zhang +1 more | Proceedings.

Scientific retrieval: an effective heuristic-aided multi-scale adaptive transformer network for information retrieval process from scientific publications

2025-06-07

K. R. Sabarmathi , M. Kalamani | Knowledge and Information Systems

Integrating Deep Learning Techniques in Information Retrieval: A Hybrid Approach to Relevance Optimization

2025-06-03

Anant Manish Singh | Journal of Information Systems Engineering & Management

Traditional keyword-based information retrieval (IR) systems while effective for exact term matching, often fail to capture semantic meaning, leading to suboptimal relevance especially for complex queries. Studies show that conventional … Traditional keyword-based information retrieval (IR) systems while effective for exact term matching, often fail to capture semantic meaning, leading to suboptimal relevance especially for complex queries. Studies show that conventional models typically achieve 85–90% accuracy whereas deep learning methods like BERT and DeepCT have reached up to 98.6% accuracy in text retrieval tasks. However, many current implementations do not fully exploit the complementary strengths of neural and lexical techniques. This research addresses that gap by proposing a hybrid IR framework that integrates BM25 with neural embeddings using transformer models and contextual weighting. Using MS-MARCO and TREC-CAR datasets, the methodology includes training neural ranking models, implementing Learning to Rank (LTR) and pseudo-relevance feedback (PRF) and evaluating performance via metrics such as mean average precision (MAP), nDCG and MRR. The hybrid system outperformed traditional models with a 25–30% improvement in recall and a 12% gain in MAP; user satisfaction scores were also 15–20% higher particularly for ambiguous or domain-specific queries. These findings suggest that combining lexical and semantic signals significantly enhances retrieval relevance and user experience. The model's applicability spans enterprise, academic and web search contexts with systems like Vertex AI and Elasticsearch already demonstrating similar performance gains. Future work will explore reducing model complexity for real-time scalability, enhancing interpretability and developing adaptive algorithms that incorporate continuous user feedback for iterative optimization.

Anime popularity prediction before huge investments: a multimodal approach using deep learning

2025-06-02

Jesús Armenta-Segura , Grigori Sidorov

In the Japanese anime industry, predicting whether an upcoming product will be popular is crucial. This article introduces one of the most comprehensive free datasets for predicting anime popularity using … In the Japanese anime industry, predicting whether an upcoming product will be popular is crucial. This article introduces one of the most comprehensive free datasets for predicting anime popularity using only features accessible before huge investments, relying solely on freely available internet data and adhering to rigorous standards based on real-life experiences. To explore this dataset and its potential, a deep neural network architecture incorporating GPT-2 and ResNet-50 is proposed. The model achieved a best mean squared error (MSE) of 0.012, significantly surpassing a benchmark with traditional methods of 0.415, and a best R-square (R2) score of 0.142, outperforming the benchmark of −37.591. The aim of this study is to explore the scope and impact of features available before huge investments in relation to anime popularity. For that reason, and complementing the MSE and R2 metrics, Pearson and Spearman correlation coefficients are used. The best results, with Pearson at 0.382 and Spearman at 0.362, along with a well-fitted learning curves, suggests that while these features are relevant, they are not decisive for determining anime popularity and they likely interacts with additional features accessible after further investments. This is one of the first multimodal approaches to address this kind of tasks, aiming to support an entertainment industry by helping to avoid financial failures and guide successful production strategies.

View PDF Chat

<scp>Text2SQL</scp> Business Intelligence System Based on Retrieval‐Augmented Generation (<scp>RAG</scp>)

2025-06-01

Jie Liu , Shiwei Chu | Engineering Reports

ABSTRACT Modern enterprises increasingly depend on data‐driven decision‐making, yet traditional SQL queries require technical expertise, limiting accessibility for nonspecialists. Advances in natural language processing, particularly deep learning generative models, have … ABSTRACT Modern enterprises increasingly depend on data‐driven decision‐making, yet traditional SQL queries require technical expertise, limiting accessibility for nonspecialists. Advances in natural language processing, particularly deep learning generative models, have enabled text‐to‐SQL (text2SQL) conversion, making database interaction more intuitive. Retrieval‐Augmented Generation (RAG) enhances this by integrating retrieval and generation for greater accuracy and relevance. This article proposes a text2SQL business intelligence system based on RAG, allowing enterprise users to extract insights from complex databases using natural language queries. By streamlining data retrieval and lowering technical barriers, the system achieves state‐of‐the‐art performance in generating SQL queries for complex tasks. It leverages the BERT (Bidirectional Encoder Representations from Transformers) model for vectorized retrieval, Generative Pretrained Transformer 4 (GPT‐4) for query generation, and Graph Neural Networks (GNNs) for modeling database structures. User interaction and feedback mechanisms further refine semantic understanding and query accuracy. Experimental results demonstrate the system's effectiveness. For multitable joins, query matching accuracy using BERT + GPT‐4 + GNN reaches 52.3% and 55.1% with beam widths of 1 and 10, respectively. For nested queries involving multitable joins, accuracy increases to 60.2% and 61.9% under the same conditions. Additionally, the system achieves the highest user satisfaction scores, validating its practical utility. By enhancing the ability to handle complex queries and reducing data access barriers, the proposed RAG‐based text2SQL system provides enterprise users with an efficient, user‐friendly tool for database interaction, significantly improving decision‐making processes.

Guest editorial of the IPM special issue on information science in human-centered AI

2025-06-01

Dan Wu , Daqing He , Preben Hansen +1 more | Information Processing & Management