Ranking Of Search Results Research Articles

Personalized and Diversified: Ranking Search Results in an Integrated Way

Ambiguity in queries is a common problem in information retrieval. There are currently two solutions: search result personalization and diversification. The former aims to tailor results for different users based on their preferences, but the limitations are redundant results and incomplete capture of user intents. The goal of the latter is to return results that cover as many aspects related to the query as possible. It improves diversity yet loses personality and cannot return the exact results the user wants. Intuitively, such two solutions can complement each other and bring more satisfactory reranking results. In this article, we propose a novel framework, namely, PnD , to integrate personalization and diversification reasonably. We employ the degree of refinding to determine the weight of personalization dynamically. Moreover, to improve the diversity and relevance of reranked results simultaneously, we design a reset RNN structure (RRNN) with the “reset gate” to measure the influence of the newly selected document on novelty. Besides, we devise a “subtopic learning layer” to learn the virtual subtopics, which can yield fine-grained representations of queries, documents, and user profiles. Experimental results illustrate that our model can significantly outperform existing search result personalization and diversification methods.

Semantic Search Engine Optimization in the News Media Industry: Challenges and Impact on Media Outlets and Journalism Practice in Greece

In today’s digital media environment, search engines constitute a significant gateway to news and an important source of traffic for online news outlets worldwide, identifying the significance of top rankings in search results. In this context, search engine optimization (SEO) has entered newsrooms as an essential element of disseminating the content of online media organizations, creating new practices and challenges for media professionals. However, the transition to the Semantic Web (SW) and the development of major search engines in recent years have led to the emergence of Semantic SEO as a new approach to old SEO practices, creating at the same time new challenges for the news media industry. This study investigates the application of Semantic SEO inside newsrooms and its impact on journalism and news media organizations. Drawing upon a series of semi-structured interviews with Greek SEO experts and a systematic review of the notion of semantic search and the corresponding semantic SEO technologies, the objective of this work is to present an analysis of how Semantic SEO affects news media and journalism content as well as to identify some of the technological practices that could contribute to enhancing the discoverability of news content in the changing context of online search.

Branching processes reveal influential nodes in social networks

Branching processes are discrete-time stochastic processes which have been largely employed to model and simulate information diffusion processes over large online social networks such as Twitter and Reddit. Here we show that a variant of the branching process model enables the prediction of the popularity of user-generated content and thus can serve as a method for ranking search results or suggestions displayed to users. The proposed branching-process variant is able to evaluate the importance of an agent in a social network and, thus we propose a novel centrality index, called the Stochastic Potential Gain (SPG). The SPG is the first centrality index which combines the knowledge of the network topology with a dynamic process taking place on it which we call a graph-driven branching process. SPG generalises a range of popular network centrality metrics such as Katz' and Subgraph. We formulate a Monte Carlo algorithm (called MCPG) to compute the SPG and prove that it is convergent and correct. Experiments on two real datasets drawn from Facebook and GitHub demonstrate that MCPG traverses only a small fraction of nodes to produce its result, thus making the Stochastic Potential Gain an appealing option to compute node centrality measure for Online social networks.

A pipeline for medical literature search and its evaluation

One database commonly used by clinicians for searching the medical literature and practicing evidence-based medicine is PubMed. As the literature grows, it has become challenging for users to find the relevant material quickly because most of the time the relevant results are not on the top. In this article, we propose a search and ranking pipeline to improve the search results based on relevancy. We first propose an ensemble model consisting of three classifiers: bidirectional long-short-term memory conditional random field (bi-LSTM-CRF), support vector machine and naive Bayes to recognise PICO (patient, intervention, comparison, outcome) elements from abstracts. The ensemble was trained on an annotated corpus consisting of 5000 abstracts split into 4000 and 1000 training and testing data, respectively. The ensemble recorded an accuracy of 93%. We then retrieved around 927,000 articles from PubMed for the years 2017–2021 (access date 16 April 2021). For every abstract, we extracted and grouped all P, I and O terms, and stored these groups along with the article ID in a separate database. During the search, every P, I and O term of the query is searched only in its corresponding group of every abstract. The scoring method simply counts the number of matches between the query’s P, I and O elements and the words in P, I and O groups, respectively. The abstracts are sorted by the number of matches and the top five abstracts are listed using their pre-stored abstract IDs. A comprehensive user study was conducted where 60 different queries were formulated and used to generate ranked search results using both PubMed and our proposed model. Five medical professionals assessed the ranked search results and marked every item as relevant or non-relevant. Both models were compared using precision@K and mean-average-precision@K metrics where K is 5. For most of the queries, our model produced higher precision@K values than PubMed. The mean-average-precision@K value of our model is also higher than PubMed (0.83 versus 0.67).

The concept of new information system of SB RAS State Public Scientific and Technological Library based on IRBIS64+

The authors review the principles and technologies of information retrieval systems implemented in Russia and abroad, with special attention to the function of e-library and interactive selective information dissemination. The main project aim is to build single comfortable information retrieval space integrating full-text abstracts, bibliographic resources legally acquired, structured and offered by SB RAS State Public Scientific and Technological Library.Based on the analysis of user search behavior of the Library’s information retrieval system based on IRBIS WEB-module and of the functional distinctions between IRBIS64+ and IRBIS WEB-module, the authors review the demands to contemporary e-catalogs and e-libraries, parameters of new information retrieval system, and stages of its implementation. The new information retrieval system based on IRBIS64+ WEB-module will provide access to all resources comprising literature in-flow to the Library with significant search potential, and guarantee their comfortable use by various user groups. IRBIS64+ WEB-module in the Library’s information retrieval system will provide searching in full-texts, and enable to make queries using Boolean logics, to setup distance between words, to rank search results, and to specify them with faceted navigation. The users will be also offered to use thematic search by navigators (GRNTI, thematic navigators of databases). The e-library is to provide open access to information resources of RAS SB SPSTL, eLIBRARY, and other open resources. The digital collections are to preserve printed original copies, which is important for old and rare publications.

Search algorithm, repetitive information, and sales on online platforms

A prominent feature of online sales is that buyers rely on the search tools offered by platforms to process information when searching for products. We develop a model that captures how the search algorithm affects buyers’ search processes, which further influences market equilibrium and welfare. If a platform adopts a highly unequal search algorithm, buyers are likely to obtain repetitive information about a small group of sellers, which causes buyers to consider fewer options and suppresses competition. By using data from food delivery platforms, we provide empirical evidence that markets with less equal distributions of store rankings in search results have higher average prices and more concentrated sales. We suggest that regulators should restrict search algorithms from showing repetitive information.

FE-TCM: Filter-Enhanced Transformer Click Model for Web Search

Constructing click models and extracting implicit relevance feedback information from the interaction between users and search engines are very important to improve the ranking of search results. Using neural network to model users’ click behaviors has become one of the effective methods to construct click models. In this paper, We propose a new Filter-Enhanced Transformer Click Model (FE-TCM) for web search. The model uses Transformer as the backbone network of feature extraction and add filter layer innovatively. Firstly, in order to reduce the influence of noise on user behavior data, we use the learnable filters to filter log noise. Secondly, following the examination hypothesis, we model the attraction estimator and examination predictor respectively to output the attractiveness scores and examination probabilities. A novel transformer model is used to learn the deeper representation among different features. Finally, we apply the combination functions to integrate attractiveness scores and examination probabilities into the click prediction. From our experiments on two real-world session datasets, it is proved that FE-TCM outperforms the existing click models for the click prediction.

Interactive Search on the Web: The Story So Far

Search on the web, specifically fetching of the relevant content, has been paid attention to since the advent of the web and particularly in recent years due to the tremendous growth in the volume of data and web pages. This paper categorizes the search services from the early days of the web to the present into keyword search engines, semantic search engines, question answering systems, dialogue systems and chatbots. As the first generation of search engines, keyword search engines have adopted keyword-based techniques to find the web pages containing the query keywords and ranking search results. In contrast, semantic search engines try to find meaningful and accurate results on the meaning and relations of things. Question-answering systems aim to find precise answers to natural language questions rather than returning a ranked list of relevant sources. As a subset of question answering systems, dialogue systems target to interact with human users through a dialog expressed in natural language. As a subset of dialogue systems, chatbots try to simulate human-like conversations. The paper provides an overview of the typical aspects of the studied search services, including process models, data preparation and presentation, common methodologies and categories.

Functional capabilities of higher education institutions’ repositories - members of the program “Priority-2030”

In order to study the composition and contentof repositories of scientific publications, to identify their functionality, 106 websites of domestic universities – members of the program«Priority-2030» have been monitored. The study has found that only 38% of the total quantity of institutions sites maintain publication repositories. At the same time, most of them include a small number of objects, or limited by the chronological framework.The search capabilities of repositories have been covered, as well as tools for managing search results. It has been determined that repositories mostly provide simple and advanced options for information retrieval, with the ability to use Boolean operators, endings shortening, faceted navigation, ranking search results. It has been found that not all repositories are equipped with instructions and recommendations on searching in them, which makes it difficult for inexperienced users to operate them.It has been revealed that the repository texts are provided with bibliographic metadata: as a rule, they are the author, the title, the imprint, and the source. However, some of them present more extended information: the detailed annotation, the abstract, the picture of the cover. The repositories are not equipped with linguistic tools, which prevents effective information search.It has been noted that the repositories studied are difficult to be found for a remote user. It is necessary to ensure their visibility, to provide information about existing repositories to potential users.In the conclusion, the benefits of creating repositories are enumerated: strengthening the authority of the institution, increasing the demand for library information resources, ensuring the visibility and accessibility of scientific and educational materials, and accelerating the knowledge exchange.

To Brush or Not to Brush: Product Rankings, Consumer Search, and Fake Orders

Brushing—online merchants placing fake orders of their own products—has been a widespread phenomenon on major e-commerce platforms. One key reason why merchants brush is that it boosts their rankings in search results. Products with higher sales volume are more likely to rank higher. Additionally, rankings matter because consumers face search frictions and narrow their attention to only the few products that show up at the top. Thus, fake orders can affect consumer choice. In our paper, we find that if brushing gets more costly for merchants (e.g., due to stricter platform policies), it may sometimes surprisingly harm consumers as it may only blunt brushing by the merchant who sells a more popular product but intensify brushing by the merchant selling a less popular product. If search is less costly for consumers (e.g., due to improved search technologies), it may not always benefit consumers, either. Moreover, the design of the ranking algorithm is critical: placing more weight on sales-volume-related factors may trigger a nonmonotone change in consumer welfare; tracking recent sales only as opposed to cumulative sales does not always dial down brushing and, in fact, may sometimes cause the merchant selling a less popular product to brush more.

Topic-independent modeling of user knowledge in informational search sessions

Web search is among the most frequent online activities. In this context, widespread informational queries entail user intentions to obtain knowledge with respect to a particular topic or domain. To serve learning needs better, recent research in the field of interactive information retrieval has advocated the importance of moving beyond relevance ranking of search results and considering a user’s knowledge state within learning oriented search sessions. Prior work has investigated the use of supervised models to predict a user’s knowledge gain and knowledge state from user interactions during a search session. However, the characteristics of the resources that a user interacts with have neither been sufficiently explored, nor exploited in this task. In this work, we introduce a novel set of resource-centric features and demonstrate their capacity to significantly improve supervised models for the task of predicting knowledge gain and knowledge state of users in Web search sessions. We make important contributions, given that reliable training data for such tasks is sparse and costly to obtain. We introduce various feature selection strategies geared towards selecting a limited subset of effective and generalizable features.

B+-Tree Based Multi-Keyword Ranked Similarity Search Scheme Over Encrypted Cloud Data

With the sustained evolution and expeditious popularization of cloud computing, an ever-increasing number of individuals and enterprises are encouraged to outsource data to cloud servers for reducing management overhead and ease of access. Privacy requirements demand encryption of sensitive information before outsourcing, which, on the other hand, diminishes the usability of data and makes considerable efficient keyword search techniques used on plaintext inapplicable. In this paper, we propose a secure multi-keyword ranked search scheme based on document similarity to work out the problem. In order to achieve the goals of multi-keyword search and ranking search results, we adopt the vector space model and TF-IDF model to generate index and query vectors. By introducing the secure kNN computation, index and query vectors can be encrypted to prevent cloud servers from obtaining sensitive frequency information. For the need of efficiency advancement, we adopt the $B^{+}$ -tree as the basic structure to build the index and construct a similar document collection for each document. Due to the use of our unique index structure, compared to linear search, the search efficiency is more exceptional. Extensive experiments on the real-world document collection are conducted to demonstrate the feasibility and efficiency of the proposed solution.

Multi-user Boolean searchable encryption supporting fast ranking in mobile clouds

Mobile cloud computing has become an important technology for mobile services. It overcomes physical limitations of mobile devices towards flexible and scalable mobile services. When using mobile clouds, people usually upload and store their data in cloud servers, and access the data through mobile devices from anywhere. As the privacy of the sensitive data is the major concern of users, the data is often encrypted before storing in cloud servers. However, the flexibility using the data is thereby affected, such as a Boolean search over the encrypted data, or ranking search results. To address these problems, in this paper we first propose a novel multi-user Boolean keyword search scheme (MBKSS) to achieve a rapid Boolean query outcome, while eliminating query interactions between users and the data owner. To rank search results and protect the privacy of relevance scores between files and keywords, we design a new homomorphic cryptosystem with partial decryption, which could be used as a basis for constructing a fast ranking search protocol (FRSP). A comprehensive security analysis shows that the proposed MBKSS and FRSP can achieve secure Boolean search and ranking. The experimental results demonstrate that the proposed MBKSS and FRSP are efficient and suitable for mobile devices.

SEO inside Newsrooms: Reports from the Field

The journalism profession has changed dramatically in the digital age as the internet, and new technologies, in general, have created new working conditions in the media environment. Concurrently, journalists and media professionals need to be aware and possess a new set of skills connected to web technologies, as well as respond to new reading tendencies and information consumption habits. A number of studies have shown that search engines are an important source of the traffic to news websites around the world, identifying the significance of high rankings in search results. Journalists are writing to be read, and that means ensuring that their news content is found, also, by search engines. In this context, this paper represents an exploratory study on the use of search engine optimization (SEO) in news websites. A series of semi-structured, in-depth interviews with professionals at four Greek media organizations uncover trends and address issues, such as how SEO policy is operationalized and applied inside newsrooms, which are the most common optimization practices, as well as the impact on journalism and news content. Today, news publishers have embraced the use of SEO practices, something that is clear also from this study. However, the absence of a distinct SEO culture was evident in newsrooms under study. Finally, according to results, SEO strategy seems to depend on factors, such as ownership and market orientation, editorial priorities or organizational structures.

Abstract 3451: Interpreting gene lists from -omics experiments

Abstract Understanding the mechanisms responsible for a cellular behavior often begins with observations of genes and gene products. Depending on the type of experiment, the number of resulting genes can be small, but increasingly, researchers are faced with many thousands of measurements, as in the case of transcriptomic or protein-DNA binding observations. Here, we describe ways to pair experimental results consisting of one or more genes with analysis tools with the overall aim being to make results more biologically interpretable. In certain cases, experimental approaches such as screens for essential genes can generate one or a few ‘genes of interest’ and there is a desire to understand their relationship to one another as well as discover links to additional, interesting genes. To this end, ‘GeneMANIA’ is a web tool that accepts gene names and returns a network visualization of related genes based on similarity in expression, localization, protein domains and those involved in physical interactions. Likewise, ‘PCViz’ is a web tool that displays a network of interactions drawn from Pathway Commons, a web resource for pathway and interaction knowledge. In cases where experiments generate a lengthy list of genes, for instance, transcriptomic measurements, there is a desire to understand their relevance to a phenotype of interest. Pathway enrichment analysis methods aim to summarize gene lists as pathways, which have a closer link to cell function. An online ‘Guide’ by Pathway Commons includes workflows that illustrate how to chain together software tools to identify pathways from the corresponding gene-level data then organize and summarize the pathway-level results in an interactive visualization known as an Enrichment Map. For those wishing to drill-down to individual pathways, Pathway Commons offers a set of web apps, including ‘Search’ that enables users to query by keyword and visualize ranked search results. Ongoing development of web apps aims to enhance the accessibility to pathways and integrate support for analysis and visualization of experimental data. The full complement of data, tools and resources offered by Pathway Commons in support of pathway analysis are described. Citation Format: Augustin Luna, Jeffrey V. Wong, Emek Demir, Igor Rodchenkov, Özgün Babur, Chris Sander, Gary D. Bader. Interpreting gene lists from -omics experiments [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2019; 2019 Mar 29-Apr 3; Atlanta, GA. Philadelphia (PA): AACR; Cancer Res 2019;79(13 Suppl):Abstract nr 3451.

Search Result Reranking with Visual and Structure Information Sources

Relevance estimation is among the most important tasks in the ranking of search results. Current methodologies mainly concentrate on text matching, link analysis, and user behavior models. However, users judge the relevance of search results directly from Search Engine Result Pages (SERPs), which provide valuable signals for reranking. In this article, we propose two different approaches to aggregate the visual, structure, as well as textual information sources of search results in relevance estimation. The first one is a late-fusion framework named Joint Relevance Estimation model (JRE). JRE estimates the relevance independently from screenshots, textual contents, and HTML source codes of search results and jointly makes the final decision through an inter-modality attention mechanism. The second one is an early-fusion framework named Tree-based Deep Neural Network (TreeNN), which embeds the texts and images into the HTML parse tree through a recursive process. To evaluate the performance of the proposed models, we construct a large-scale practical Search Result Relevance (SRR) dataset that consists of multiple information sources and relevance labels of over 60,000 search results. Experimental results show that the proposed two models achieve better performance than state-of-the-art ranking solutions as well as the original rankings of commercial search engines.

Does Diversity Affect User Satisfaction in Image Search

Diversity has been taken into consideration by existing Web image search engines in ranking search results. However, there is no thorough investigation of how diversity affects user satisfaction in image search. In this article, we address the following questions: (1) How do different factors, such as content and visual presentations, affect users’ perception of diversity? (2) How does search result diversity affect user satisfaction with different search intents? To answer those questions, we conduct a set of laboratory user studies to collect users’ perceived diversity annotations and search satisfaction. We find that the existence of nearly duplicated image results has the largest impact on users’ perceived diversity, followed by the similarity in content and visual presentations. Besides these findings, we also investigate the relationship between diversity and satisfaction in image search. Specifically, we find that users’ preference for diversity varies across different search intents. When users want to collect information or save images for further usage (the Locate search tasks), more diversified result lists lead to higher satisfaction levels. The insights may help commercial image search engines to design better result ranking strategies and evaluation metrics.

Content-Based Model of Web Search Behavior: An Application to TV Show Search

We develop a flexible content-based search model that links the content preferences of search engine users to query search volume and click-through rates. Content preferences are defined over latent topics that describe the content of search queries and search result descriptions. Moreover, our model allows content preferences to vary systematically based on the context of the search. To facilitate efficient and scalable inference, we develop a full Bayesian variational inference algorithm. We evaluate our modelling framework using real-world data on searches for TV shows on the Bing search engine. We illustrate how our model can quantify the content preferences associated with each query, and how these preferences vary systematically based on whether the query is observed before, during, or after a TV show is aired. We also show that our model can help the search engine improve its ranking of search results. In addition to search engines, advertisers can apply our model to understand their search advertising campaign data and gain insights into which search ads should be targeted to which users, when, and on which devices.

Didactic metadata informing teachers’ selection of learning resources: boundary crossing in professional development

With the emergence of e-textbooks, along with expectations to integrate technology in instruction, teachers are becoming significant co-designers of the curriculum. Informed selection and sequencing of learning resources requires sensitivity to didactic nuance, and tools to support development and application of such sensitivity. Teachers’ practices are constrained by the aspects of resources that are “searchable,” which are limited using standard search engines, and the selection of tasks is influenced by the engines’ ranking of search results, reflecting their popularity. We are developing and researching a coupled pair of tools to support mathematics teachers in making informed curricular decisions—a tool for tagging learning resources with prescribed categories of didactic metadata and a dashboard for browsing collections of resources according to this tagged metadata. In this article, we investigate affordances of these tools for the professional development of mathematics teachers—both practicing and pre-service teacher candidates. Viewing the dashboard, along with the metadata that it encodes, as a boundary object between the teachers’ and the researchers’ perspectives on curricular design, we show how teachers learned through acts of boundary crossing, conceived as transitions and interactions between the two communities’ curricular discourses. We show how using the dashboard in a task-selection assignment encouraged teachers to reflect on their practice—making explicit the tacit considerations that they apply to curricular decisions and articulating them from the researchers’ perspective. We also describe the emergence of “hybrid” search strategies, integrating multiple perspectives to create practices that are both didactically informed and practically relevant for instruction.

To Brush or Not to Brush: Product Rankings, Customer Search and Fake Orders

Brushing---online merchants placing fake orders of their own products---has been a widespread phenomenon on major e-commerce platforms. One key reason why merchants brush is that it boosts their rankings in search results. Products with higher sales volume are more likely to rank higher. Additionally, rankings matter because consumers face search frictions and narrow their attention to only the few products that show up at the top. Thus, fake orders can affect consumer choice. We focus on this search-ranking aspect of brushing and build a stylized model to understand merchants’ strategic brushing behavior as well as how it affects consumers. We consider a high-type merchant (who sells a more popular product) and a low-type merchant (who sells a less popular product) competing on an e-commerce platform where product rankings evolve over time. We find that if brushing gets more costly for merchants (e.g., due to stricter platform policies), it may sometimes surprisingly harm consumers as it may only blunt brushing by the high-type merchant but intensify brushing by the low-type merchant. If search is less costly for consumers (e.g., due to improved search technologies), it may not always benefit consumers, either. Moreover, the design of the ranking algorithm is critical: placing more weight on sales volume-related factors may trigger a non-monotone change in consumer welfare; tracking recent sales only as opposed to cumulative sales does not always dial down brushing and, in fact, may sometimes cause the low-type merchant to brush more.