Information Retrieval Perspective Research Articles

A case study on decompounding in Indian language IR

Abstract Decompounding is an essential preprocessing step in text-processing tasks such as machine translation, speech recognition, and information retrieval (IR). Here, the IR issues are explored from five viewpoints. (A) Does word decompounding impact the Indian language IR? If yes, to what extent? (B) Can corpus-based decompounding models be used in the Indian language IR? If yes, how? (C) Can machine learning and deep learning-based decompounding models be applied in the Indian language IR? If yes, how? (D) Among the different decompounding models (corpus-based, hybrid machine learning-based, and deep learning-based), which provides the best effectiveness in the IR domain? (E) Among the different IR models, which provides the best effectiveness from the IR perspective? This study proposes different corpus-based, hybrid machine learning-based, and deep learning-based decompounding models in Indian languages (Marathi, Hindi, and Sanskrit). Moreover, we evaluate the effectiveness of each activity from an IR perspective only. It is observed that the different decompounding models improve IR effectiveness. The deep learning-based decompounding models outperform the corpus-based and hybrid machine learning-based models in Indian language IR. Among the different deep learning-based models, the Bi-LSTM-A model performs best and improves mean average precision (MAP) by 28.02% in Marathi. Similarly, the Bi-RNN-A model improves MAP by 18.18% and 6.1% in Hindi and Sanskrit, respectively. Among the retrieval models, the In_expC2 model outperforms others in Marathi and Hindi, and the BB2 model outperforms others in Sanskrit.

Fast and Constrained Absent Keyphrase Generation by Prompt-Based Learning

Generating absent keyphrases, which do not appear in the input document, is challenging in the keyphrase prediction task. Most previous works treat the problem as an autoregressive sequence-to-sequence generation task, which demonstrates promising results for generating grammatically correct and fluent absent keyphrases. However, such an end-to-end process with a complete data-driven manner is unconstrained, which is prone to generate keyphrases inconsistent with the input document. In addition, the existing autoregressive decoding method makes the generation of keyphrases must be done from left to right, leading to slow speed during inference. In this paper, we propose a constrained absent keyphrase generation method in a prompt-based learning fashion. Specifically, the prompt will be created firstly based on the keywords, which are defined as the overlapping words between absent keyphrase and document. Then, a mask-predict decoder is used to complete the absent keyphrase on the constraint of prompt. Experiments on keyphrase generation benchmarks have demonstrated the effectiveness of our approach. In addition, we evaluate the performance of constrained absent keyphrases generation from an information retrieval perspective. The result shows that our approach can generate more consistent keyphrases, which can improve document retrieval performance. What’s more, with a non-autoregressive decoding manner, our model can speed up the absent keyphrase generation by 8.67× compared with the autoregressive method.

Atopic dermatitis or eczema? Consequences of ambiguity in disease name for biomedical literature mining.

Biomedical research increasingly relies on computational approaches to extract relevant information from large corpora of publications. To investigate the consequence of the ambiguity between the use of terms "Eczema" and "Atopic Dermatitis" (AD) from the Information Retrieval perspective, and its impact on meta-analyses, systematic reviews and text mining. Articles were retrieved by querying the PubMed using terms 'eczema' (D003876) and "dermatitis, atopic" (D004485). We used machine learning to investigate the differences between the contexts in which each term is used. We used a decision tree approach and trained model to predict if an article would be indexed with eczema or AD tags. We used text-mining tools to extract biological entities associated with eczema and AD, and investigated the discrepancy regarding the retrieval of key findings according to the terminology used. Atopic dermatitis query yielded more articles related to veterinary science, biochemistry, cellular and molecular biology; the eczema query linked to public health, infectious disease and respiratory system. Medical Subject Headings terms associated with "AD" or "Eczema" differed, with an agreement between the top 40lists of 52%. The presence of terms related to cellular mechanisms, especially allergies and inflammation, characterized AD literature. The metabolites mentioned more frequently than expected in articles with AD tag differed from those indexed with eczema. Fewer enriched genes were retrieved when using eczema compared to AD query. There is a considerable discrepancy when using text mining to extract bio-entities related to eczema or AD. Our results suggest that any systematic approach (particularly when looking for metabolites or genes related to the condition) should be performed using both terms jointly. We propose to use decision tree learning as a tool to spot and characterize ambiguity, and provide the source code for disambiguation at https://github.com/cfrainay/ResearchCodeBase.

Text‐based question answering from information retrieval and deep neural network perspectives: A survey

AbstractText‐based question answering (QA) is a challenging task which aims at finding short concrete answers for users' questions. This line of research has been widely studied with information retrieval (IR) techniques and has received increasing attention in recent years by considering deep neural network approaches. Deep learning (DL) approaches, which are the main focus of this paper, provide a powerful technique to learn multiple layers of representations and interaction between the questions and the answer sentences. In this paper, we provide a comprehensive overview of different models proposed for the QA task, including both a traditional IR perspective and a more recent deep neural network environment. We also introduce well‐known datasets for the task and present available results from the literature to have a comparison between different techniques.This article is categorized under: Algorithmic Development > Text Mining Technologies > Machine Learning

A survey on single server private information retrieval in a coding theory perspective

In this paper, we present a new perspective of single server private information retrieval (PIR) schemes by using the notion of linear error-correcting codes. Many of the known single server schemes are based on taking linear combinations between database elements and the query elements. Using the theory of linear codes, we develop a generic framework that formalizes all such PIR schemes. This generic framework provides an appropriate setup to analyze the security of such PIR schemes. In fact, we describe some known PIR schemes with respect to this code-based framework, and present the weaknesses of the broken PIR schemes in a unified point of view.

The Influence of Argument Quality towards Information Adoption in E-commerce Review Platform

The viral nature the content of the Web has transformed the landscape of e-Commerce review platforms to be in a state of constant growth. Similarly, the prominent features of these platforms have been recognized to be among the dominant factors in shaping online consumer behavior. Nonetheless, in this regard, if the review platform returns too many reviews, and the reviews are presented in non-relevant manner, in which this may be cumbersome and time-consuming for consumers. Therefore, identifying credible reviews that contain valuable information has becomes increasingly important for online businesses. The main research question to be addressed in this study is to determine on how can a model be developed to improve the argument quality perceptions in the adoption of online reviews across e-Commerce review platform. Subsequently, the main objective to be achieved is to develop a model of argument quality for review’s adoption in the e-Commerce review platform. The potential effects of consumer relevance judgment from information retrieval perspective have been considered, which include perceived informative and affective relevance in developing the research model by using Elaboration Likelihood Model (ELM). A quantitative research method has been applied to test and validate the propose research model. The response data from 238 valid respondents was analyzed using the Partial Least Square Structural Modelling (PLS-SEM) technique. The findings from the results indicate that content novelty, content topicality, content similarity, content tangibility and content sentimentality could positively influence the perception of argument quality which lead to information adoption behavior. Finally, the importance of information relevancy was also highlighted in this study, which reveals some appropriate features that can be utilized by e-Commerce practitioners to better refine their information search criteria in the online review platforms.

Source Code Plagiarism Detection in Academia with Information Retrieval: Dataset and the Observation

Source code plagiarism is an emerging issue in computer science education. As a result, a number of techniques have been proposed to handle this issue. However, comparing these techniques may be challenging, since they are evaluated with their own private dataset(s). This paper contributes in providing a public dataset for comparing these techniques. Specifically, the dataset is designed for evaluation with an Information Retrieval (IR) perspective. The dataset consists of 467 source code files, covering seven introductory programming assessment tasks. Unique to this dataset, both intention to plagiarise and advanced plagiarism attacks are considered in its construction. The dataset's characteristics were observed by comparing three IR-based detection techniques, and it is clear that most IR-based techniques are less effective than a baseline technique which relies on Running-Karp-Rabin Greedy-String-Tiling, even though some of them are far more time-efficient.

Extracting Emotion Causes Using Learning to Rank Methods From an Information Retrieval Perspective

Emotion cause extraction is a challenging task for the fine-grained emotion analysis. Even though a few studies have addressed the task using clause-level classification methods, most of them have partly ignored emotion-level context information. To comprehensively leverage the information, we propose a novel method based on learning to rank to identify emotion causes from an information retrieval perspective. Our method seeks to rank candidate clauses with respect to certain provoked emotions in analogy with query-level document ranking in information retrieval. To learn effective clause ranking models, we represent candidate clauses as feature vectors involving both emotion-independent features and emotion-dependent features. Emotion-independent features are extracted to capture the possibility that a clause is expected to provoke an emotion, and emotion-dependent features are extracted to capture the relevance between candidate cause clauses and their corresponding emotions. We investigate three approaches to learning to rank for emotion cause extraction in our method. We evaluate the performance of our method on an existing dataset for emotion cause extraction. The experimental results show that our method is effective in emotion cause extraction, significantly outperforming the state-of-the-art baseline methods in terms of the precision, recall, and F-measure.

Social Book Search: a survey

Social Book Search is a new area of social search. In the modern world everything is going to be amenable due to the web and social media. The web and social media give us access to a wealth of information, not only different in quantity but also in character. Traditional descriptions from professionals are now supplemented with user generated content. Although books have been the predominant source of information for centuries, the way we acquire, share, and publish information has changed and has been changing in fundamental ways due to the web. In the modern era, in order to purchase a book, the users are not only depend on the title, author name, publisher etc of the book available as controlled metadata but also on other aspects which include the reviews, editorial reviews etc available in different social media. Our primary focus in this survey is to describe the features of different social cataloging book sites as well as their recommendation based on books. How the online searching of books is useful to the user and up to what extent, and what are their aim are some of the issues we shall deal with. We will also discuss evolution of those techniques for Social Book Search presented over years at the Initiative for the Evaluation of XML Retrieval (INEX) and Conference and Labs of the Evaluation Forum (CLEF). To what extent the features and functionality of a social sharing platform influence the user behavior, is also discussed in the paper. This survey provides an overview of research done in the area of Social Book Search from perspective of Information Retrieval.

PubMed Phrases, an open set of coherent phrases for searching biomedical literature

In biomedicine, key concepts are often expressed by multiple words (e.g., ‘zinc finger protein’). Previous work has shown treating a sequence of words as a meaningful unit, where applicable, is not only important for human understanding but also beneficial for automatic information seeking. Here we present a collection of PubMed® Phrases that are beneficial for information retrieval and human comprehension. We define these phrases as coherent chunks that are logically connected. To collect the phrase set, we apply the hypergeometric test to detect segments of consecutive terms that are likely to appear together in PubMed. These text segments are then filtered using the BM25 ranking function to ensure that they are beneficial from an information retrieval perspective. Thus, we obtain a set of 705,915 PubMed Phrases. We evaluate the quality of the set by investigating PubMed user click data and manually annotating a sample of 500 randomly selected noun phrases. We also analyze and discuss the usage of these PubMed Phrases in literature search.

Construction of Corpus in Artificial Intelligence Age

As a kind of revolutionary technology, artificial intelligence marked an explosive transformation in many fields of study. Nowadays, much of the translation work used to be done by human has been undertaken by machines. The construction of corpus is a crucial step leading to successful machine translation. The paper aims to exploring the mode of corpus construction from the perspective of information mining, information retrieval and information processing. The retrieval system uses web crawlers to collect network information and automatic tagging technology to index the collected information, then applies corresponding language processing techniques to achieve correspondence between two languages and form an index database. In the age of artificial intelligence, machines can keep a track of many users’ searches, queries, so as to record, extract as well as to feed back on different translations to build a new corpus. In this way, machine translation is improving in its scope and accuracy in translation with the goal to take up the tedious work of human translation as well as to increase the speed and reduce the cost of it.

Modeling and Simulating the Web of Things from an Information Retrieval Perspective

Internet and Web technologies have changed our lives in ways we are not yet fully aware of. In the near future, Internet will interconnect more than 50 billion things in the real world, nodes will sense billions of features and properties of interest, and things will be represented by web-based, bi-directional services with highly dynamic content and real-time data. This is the new era of the Internet and the Web of Things . Since the emergence of such paradigms implies the evolution and integration of the systems with which they interact, it is essential to develop abstract models for representing and simulating the Web of Things in order to establish new approaches. This article describes a Web of Things model based on a structured XML representation. We also present a simulator whose ultimate goal is to encapsulate the expected dynamics of the Web of Things for the future development of information retrieval (IR) systems. The simulator generates a real-time collection of XML documents containing spatio-temporal contexts and textual and sensed information of highly dynamic dimensions. The simulator is characterized by its flexibility and versatility for representing real-world scenarios and offers a unique perspective for information retrieval. In this article, we evaluate and test the simulator in terms of its performance variables for computing resource consumption and present our experimentation with the simulator on three real scenarios by considering the generation variables for the IR document collection.

Retrieving and classifying instances of source code plagiarism

Automatic detection of source code plagiarism is an important research field for both the commercial software industry and within the research community. Existing methods of plagiarism detection primarily involve exhaustive pairwise document comparison, which does not scale well for large software collections. To achieve scalability, we approach the problem from an information retrieval (IR) perspective. We retrieve a ranked list of candidate documents in response to a pseudo-query representation constructed from each source code document in the collection. The challenge in source code document retrieval is that the standard bag-of-words (BoW) representation model for such documents is likely to result in many false positives being retrieved, because of the use of identical programming language specific constructs and keywords. To address this problem, we make use of an abstract syntax tree (AST) representation of the source code documents. While the IR approach is efficient, it is essentially unsupervised in nature. To further improve its effectiveness, we apply a supervised classifier (pre-trained with features extracted from sample plagiarized source code pairs) on the top ranked retrieved documents. We report experiments on the SOCO-2014 dataset comprising 12K Java source files with almost 1M lines of code. Our experiments confirm that the AST based approach produces significantly better retrieval effectiveness than a standard BoW representation, i.e., the AST based approach is able to identify a higher number of plagiarized source code documents at top ranks in response to a query source code document. The supervised classifier, trained on features extracted from sample plagiarized source code pairs, is shown to effectively filter and thus further improve the ranked list of retrieved candidate plagiarized documents.

Paraphrase-focused learning to rank for domain-specific frequently asked questions retrieval

A frequently asked questions (FAQ) retrieval system improves the access to information by allowing users to pose natural language queries over an FAQ collection. From an information retrieval perspective, FAQ retrieval is a challenging task, mainly because of the lexical gap that exists between a query and an FAQ pair, both of which are typically very short. In this work, we explore the use of supervised learning to rank to improve the performance of domain-specific FAQ retrieval. While supervised learning-to-rank models have been shown to yield effective retrieval performance, they require costly human-labeled training data in the form of document relevance judgments or question paraphrases. We investigate how this labeling effort can be reduced using a labeling strategy geared toward the manual creation of query paraphrases rather than the more time-consuming relevance judgments. In particular, we investigate two such strategies, and test them by applying supervised ranking models to two domain-specific FAQ retrieval data sets, showcasing typical FAQ retrieval scenarios. Our experiments show that supervised ranking models can yield significant improvements in the precision-at-rank-5 measure compared to unsupervised baselines. Furthermore, we show that a supervised model trained using data labeled via a low-effort paraphrase-focused strategy has the same performance as that of the same model trained using fully labeled data, indicating that the strategy is effective at reducing the labeling effort while retaining the performance gains of the supervised approach. To encourage further research on FAQ retrieval we make our FAQ retrieval data set publicly available.

Retrieving relevant time-course experiments: a study on Arabidopsis microarrays.

Understanding time-course regulation of genes in response to a stimulus is a major concern in current systems biology. The problem is usually approached by computational methods to model the gene behaviour or its networked interactions with the others by a set of latent parameters. The model parameters can be estimated through a meta-analysis of available data obtained from other relevant experiments. The key question here is how to find the relevant experiments which are potentially useful in analysing current data. In this study, the authors address this problem in the context of time-course gene expression experiments from an information retrieval perspective. To this end, they introduce a computational framework that takes a time-course experiment as a query and reports a list of relevant experiments retrieved from a given repository. These retrieved experiments can then be used to associate the environmental factors of query experiment with the findings previously reported. The model is tested using a set of time-course Arabidopsis microarrays. The experimental results show that relevant experiments can be successfully retrieved based on content similarity.

Science models for search: a study on combining scholarly information retrieval and scientometrics

Models of science address statistical properties and mechanisms of science. From the perspective of scholarly information retrieval (IR) science models may provide some potential to improve retrieval quality when operationalized as specific search strategies or used for rankings. From the science modeling perspective, on the other hand, scholarly IR can play the role of a validation model of science models. The paper studies the applicability and usefulness of two particular science models for re-ranking search results (Bradfordizing and author centrality). The paper provides a preliminary evaluation study that demonstrates the benefits of using science model driven ranking techniques, but also how different the quality of search results can be if different conceptualizations of science are used for ranking.

Information retrieval approach to meta-visualization

Visualization is crucial in the first steps of data analysis. In visual data exploration with scatter plots, no single plot is sufficient to analyze complicated high-dimensional data sets. Given numerous visualizations created with different features or methods, meta-visualization is needed to analyze the visualizations together. We solve how to arrange numerous visualizations onto a meta-visualization display, so that their similarities and differences can be analyzed. Visualization has recently been formalized as an information retrieval task; we extend this approach, and formalize meta-visualization as an information retrieval task whose performance can be rigorously quantified and optimized. We introduce a machine learning approach to optimize the meta-visualization, based on an information retrieval perspective: two visualizations are similar if the analyst would retrieve similar neighborhoods between data samples from either visualization. Based on the approach, we introduce a nonlinear embedding method for meta-visualization: it optimizes locations of visualizations on a display, so that visualizations giving similar information about data are close to each other. In experiments we show such meta-visualization outperforms alternatives, and yields insight into data in several case studies.

Analysis of FRBR User Tasks

FRBR, FRAD, and FRSAD models propose user tasks as a way to address and categorize functions that a catalog should support. The user tasks are not harmonized among these models, but to do that, they should first be fully understood and analyzed, especially “select” and “identify.” We decided to look at the FRBR user tasks from the perspective of interactive information retrieval (IIR). Several IIR models were reviewed and Ellis’ and Belkin's models were chosen for further analysis and interpretation of FRBR “select” and “identify” tasks.

Application of Text Summarization techniques to the Geographical Information Retrieval task

Automatic Text Summarization has been shown to be useful for Natural Language Processing tasks such as Question Answering or Text Classification and other related fields of computer science such as Information Retrieval. Since Geographical Information Retrieval can be considered as an extension of the Information Retrieval field, the generation of summaries could be integrated into these systems by acting as an intermediate stage, with the purpose of reducing the document length. In this manner, the access time for information searching will be improved, while at the same time relevant documents will be also retrieved. Therefore, in this paper we propose the generation of two types of summaries (generic and geographical) applying several compression rates in order to evaluate their effectiveness in the Geographical Information Retrieval task. The evaluation has been carried out using GeoCLEF as evaluation framework and following an Information Retrieval perspective without considering the geo-reranking phase commonly used in these systems. Although single-document summarization has not performed well in general, the slight improvements obtained for some types of the proposed summaries, particularly for those based on geographical information, made us believe that the integration of Text Summarization with Geographical Information Retrieval may be beneficial, and consequently, the experimental set-up developed in this research work serves as a basis for further investigations in this field.

Using SVM to Predict Stock Price Changes from Online Financial News

Many technical analysis use financial indices to predict stock price changes. In this paper, we present a different approach for prediction stock price fluctuations using financial news. Our method approaches the stock price prediction problem from an information retrieval perspective. We apply both text analysis and pattern classification techniques to search for important online news that are relevant for stock price changes. First, the online financial news and the corresponding stocks are extracted. Then we apply Support Vector Machine (SVM) to construct a model that predicts the price changes for the stocks. Finally, the stock changes prediction model is used to classify and extract upcoming important financial news. The experimental results demonstrate our method is effective for seeking the important financial news for stock price changes.