Traditional information retrieval (IR) models, such as keyword-based and vector-based techniques, have long been used in centralized systems. However, the Web’s re-decentralization, with its focus on data ownership and privacy, calls for a re-evaluation of these methods in these settings. While standards for decentralized search enhance privacy to some extent, they also introduce computational overhead, black-box decision-making, and infrastructure complexity. Despite these challenges, traditional IR techniques remain largely unexplored in such environments. This article presents an innovative application of traditional IR models in the decentralized Web by adapting them for Personal Online Data Stores (PODs), where search parties have varying access rights. We explore their role in source selection, document ranking, and result merging, extending them to meet decentralized search demands. Using Solid PODs and a synthetic medical dataset, we evaluate these models in a privacy-sensitive environment. Our findings demonstrate that extended IR methods provide an effective balance of performance, interpretability, and efficiency. These approaches hold strong potential as privacy-preserving alternatives for decentralized search on a re-decentralized Web. Notably, our top-performing model achieved competitive results in top-item retrieval compared to centralized search systems, maintaining high relevance scores under both limited and full data access conditions.

Data privacy and sovereignty are open challenges in today’s Web, which the Solid ( https://solidproject.org ) ecosystem aims to meet by providing personal online datastores (pods) where individuals can control access to their data. Solid allows developers to deploy applications with access to data stored in pods, subject to users’ permission. For the decentralised Web to succeed, the problem of search over pods with varying access permissions must be solved. The ESPRESSO framework takes the first step in exploring such a search architecture, enabling large-scale keyword search across Solid pods with varying access rights. This paper provides a comprehensive experimental evaluation of the performance and scalability of decentralised keyword search across pods on the current ESPRESSO prototype. The experiments specifically investigate how controllable experimental parameters influence search performance across a range of decentralised settings. This includes examining the impact of different text dataset sizes (0.5 MB to 50 MB per pod, divided into 1 to 10,000 files), different access control levels (10%, 25%, 50%, or 100% file access), and a range of configurations for Solid servers and pods (from 1 to 100 pods across 1 to 50 servers). The experimental results confirm the feasibility of deploying a decentralised search system to conduct keyword search at scale in a decentralised environment.

Abstract The increasing centralization of the Web raises serious concerns regarding privacy, security, and user autonomy. In response, there has been a renewed interest in the development of secure personal information management systems and a movement towards decentralization. Decentralized personal online data stores (pods) represent a revolutionary example within this movement, built on the W3C’s existing guidelines – an approach exemplified by initiatives such as ( https://solidproject.org ). In the Solid paradigm, individuals store their personal data in pods and have absolute discretion when choosing to grant access to different users and applications. A barrier to the adoption of the pod approach is the predominant reliance on centralized indexes for search functionality in current Web and Web-based systems. This paper introduces the framework, which is designed to facilitate this new paradigm of large-scale searches within personal data stores while respecting the individual pod owners’ data access governance. The current ESPRESSO prototype integrates access control within pod indexes to enhance distributed keyword-based search. ESPRESSO’s unique contribution not only enhances search capabilities on the decentralized Web but also paves the way for future explorations in decentralized search technologies.

Resource Description Framework datasets can be queried using the SPARQL language but are often irregularly structured and incomplete, which may make precise query formulation hard for users. The SPARQLARlanguage extends SPARQL 1.1 with two operators—APPROX and RELAX—to allow flexible querying over property paths. These operators encapsulate different dimensions of query flexibility, namely, approximation and generalisation, and they allow users to query complex, heterogeneous knowledge graphs without needing to know precisely how the data is structured. Earlier work has described the syntax, semantics, and complexity of SPARQLAR, has demonstrated its practical feasibility, but has also highlighted the need for improving the speed of query evaluation. In the present article, we focus on the design of two optimisation techniques targeted at speeding up the execution of SPARQLARqueries and on their empirical evaluation on three knowledge graphs: LUBM, DBpedia, and YAGO. We show that applying these optimisations can result in substantial improvements in the execution times of longer-running queries (sometimes by one or more orders of magnitude) without incurring significant performance penalties for fast queries.

Graph databases are well-suited to managing large, complex, dynamically evolving datasets. However, for data that is irregular and heterogeneous, it may be difficult to formulate queries that precisely capture a user's information seeking requirements. This points to the need for approximate query processing capabilities that can automatically make changes to a query so as to aid in the incremental discovery of relevant information. In this paper we motivate and explore techniques for providing such capabilities for the Cypher query language. This is the first time that query approximation has been investigated in the context of the property graph data model, which is becoming increasingly prevalent in research and industry.

We aim to improve the accuracy of path selectivity estimation in graph databases by intelligently ordering the domain of a histogram used for estimation. This problem has not, to our knowledge, received adequate attention in the research community. We present a novel framework for the systematic study of path ordering strategies in histogram construction and use. In this framework, we introduce new ordering strategies which we experimentally demonstrate lead to significant improvement of the accuracy of path selectivity estimation over current strategies. These positive results highlight the fundamental role that domain ordering plays in the design of effective histograms for efficient and scalable graph query processing.

We demonstrate the use of localized path indexes in generating efficient execution plans for regular path queries. This study is motivated by both the practicality of this class of queries and by the current dearth of scalable solutions for their evaluation. Our proposed solution leverages widely available relational database technology and is often orders of magnitude faster than currently known approaches. We aim in this hands-on demonstration to both highlight the promise of our approach and to stimulate further discussion and study of engineering solutions for this practical yet challenging class of graph queries.

We present a seed expansion based approach to classify named entities in web search queries. Previous approaches to this classification problem relied on contextual clues in the form of keywords surrounding a named entity in the query. Here we propose an alternative approach in the form of a Bag-of-Context-Words (BoCW) that is used to represent the context words as they appear in the snippets of the top search results for the query. This is particularly useful in the case where the query consists of only the named entity without any context words, since in the previous approaches no context is discovered. In order to construct the BoCW, we employ a novel algorithm, which iteratively expands a Class Vector that is created through expansion by gradually aggregating the BoCWs of similar named entities appearing in other queries. We provide comprehensive experimental evidence using a commercial query log showing that our approach is competitive with existing approaches.

The information extraction task of Named Entities Recognition (NER) has been recently applied to search engine queries, in order to better understand their semantics. Here we concentrate on the task prior to the classification of the named entities (NEs) into a set of categories, which is the problem of detecting candidate NEs via the subtask of query segmentation.We present a novel method for detecting candidate NEs using grammar annotation and query segmentation with the aid of top-n snippets from search engine results and a web n-gram model, to accurately identify NE boundaries. The proposed method addresses the problem of accurately setting boundaries of NEs and the detection of multiple NEs in queries.

What kind of metadata can be used for expressing the multiplicity of data models and the data transformation and integration processes in data warehousing environments? How can this metadata be further used for supporting other data warehouse activities? We examine how these questions are addressed by AutoMed, a system for expressing data transformation and integration processes in heterogeneous database environments.

We describe a new approach to data integration which subsumes the previous approaches of local as view (LAV) and global as view (GAV). Our method, which we term both as view (BAV), is based on the use of reversible schema transformation sequences. We show how LAV and GAV view definitions can be fully derived from BAV schema transformation sequences, and how BAV transformation sequences may be partially derived from LAV or GAV view definitions. We also show how BAV supports the evolution of both global and local schemas, and we discuss ongoing implementation of the BAV approach within the AutoMed project.

XML repositories are now a widespread means for storing and exchanging information on the Web. As these repositories become increasingly used in dynamic applications such as e-commerce, there is a rapidly growing need for a mechanism to incorporate reactive functionality in an XML setting. Event-condition-action (ECA) rules are a technology from active databases and are a natural method for supporting suchfunctionality. ECA rules can be used for activities such as automatically enforcing document constraints, maintaining repository statistics, and facilitating publish/subscribe applications. An important question associated with the use of a ECA rules is how to statically predict their run-time behaviour. In this paper, we define a language for ECA rules on XML repositories. We then investigate methods for analysing the behaviour of a set of ECA rules, a task which has added complexity in this XML setting compared with conventional active databases.

Hyperlog is a declarative, graph based language that supports database querying and update. It visualizes schema information, data, and query output as sets of nested graphs, which can be stored, browsed, and queried in a uniform way. Thus, the user need only be familiar with a very small set of syntactic constructs. Hyperlog queries consist of a set of graphs that are matched against the database. Database updates are supported by means of programs consisting of a set of rules. The paper discusses the formulation, evaluation, expressiveness, and optimization of Hyperlog queries and programs. We also describe a prototype implementation of the language and we compare and contrast our approach with work in a number of related areas, including visual database languages, graph based data models, database update languages, and production rule systems.

Three recent trends in database research are object-oriented and deductive databases and graph-based user interfaces. We draw these trends together in a data model we call the Hypernode Model. The single data structure of this model is the hypernode , a graph whose nodes can themselves be graphs. Hypernodes are typed, and types, too, are nested graphs. We give the theoretical foundations of hypernodes and types, and we show that type checking is tractable. We show also how conventional type-forming operators can be simulated by our graph types, including cyclic types. The Hypernode Model comes equipped with a rule-based query language called Hyperlog, which is complete with respect to computation and update. We define the operational semantics of Hyperlog and show that the evaluation can be performed efficiently. We discuss also the use of Hyperlog for supporting database browsing, an essential feature of Hypertext databases. We compare our work with other graph-based data models—unlike previous graph-based models, the Hypernode Model provides inherent support for data abstraction via its nesting of graphs. Finally, we briefly discuss the implementation of a DBMS based on the Hypernode Model.