Project

Back to overview

A Market-based Approach for Querying the Web of Data

English title A Market-based Approach for Querying the Web of Data
Applicant Bernstein Abraham
Number 153598
Funding scheme Project funding (Div. I-III)
Research institution Institut für Informatik Universität Zürich
Institution of higher education University of Zurich - ZH
Main discipline Information Technology
Start/End 01.09.2014 - 31.08.2019
Approved amount 448'428.00
Show all

All Disciplines (2)

Discipline
Information Technology
Economics

Keywords (9)

Combinatorial Auction; Market Mechanisms; Semantic Web; Economic Methods for Computational Problems; Web of Data; Distributed Query Optimization ; Game Theory; Federated Databases; Mechanism Design

Lay Summary (German)

Lead
Immer mehr Daten werden heutzutage im sogenannten Web of Data (WoD) publiziert. Im Gegensatz zum traditionellen Web existiert die Mehrheit dieser Daten nur aufgrund von Subventionen von Forschungsprojekten oder der öffentlichen Hand. Dies führt zu mehreren Problemen. Erstens, wenn ein Forschungsprojekt zu Ende ist, dann verschwinden die Daten oft oder werden nicht mehr aktuell gehalten. Zweitens sind viele der Daten oft veraltet. Drittens werden viele der Daten nicht in einem einfach zu verarbeitenden Format angeboten, da die Anreize zur Konversion fehlen. Der Grund für diese Probleme ist einfach: Werbung funktioniert im WoD als Finanzierungsquelle nicht, da die Daten meist von Programmen “gelesen” werden, welche die Werbung einfach filtern. Daher stellt sich die Frage, wie sich das WoD langfristig von Subventionen als Finanzierungsbasis lösen lässt.
Lay summary

In diesem Projekt untersuchen wir den Entwurf eines globalen Marktes für die Beantwortung von Anfragen auf dem WoD. Ziel ist es, den Markt so zu entwerfen, dass Anreize geschaffen werden, um Daten zu publizieren und zu unterhalten. Wir untersuchen dabei sowohl den Einfluss des für die Anfragenden generierten Mehrwertes wie auch die Kosten für die Datenproduktion und den Datenunterhalt. Für die Entwicklung des Marktes benutzen wir Mechanismus Design Theorie, welche eine formale Analyse der Markteffizienz und der Anreizeigenschaften ermöglicht.

Das Projekt wird eine neue Sichtweise auf die ökonomischen Aspekte des Semantischen Webs eröffnen. Zusätzlich leistet das Projekt Pionierarbeit in der Untersuchung von Marktmechanismen für grosse und komplexe Informationsdienstleistungen welche von zentraler Wichtigkeit für Big Data und Data Science, und damit auch essentiell für die Zukunft der Informationsgesellschaft sind.

 

Direct link to Lay Summary Last update: 21.08.2014

Responsible applicant and co-applicants

Employees

Publications

Publication
Data Markets with Dynamic Arrival of Buyers and Sellers
MoorDmitry (2019), Data Markets with Dynamic Arrival of Buyers and Sellers, in Proceedings of the 14th Workshop on the Economics of Networks, Systems and Computation (NetEcon), ACM, New York.
Financing the Web of Data with Delayed-Answer Auctions
Grubenmann Tobias, Bernstein Abraham, Moor Dmitry, Seuken Sven (2018), Financing the Web of Data with Delayed-Answer Auctions, in WWW 2018: The 2018 Web Conference, LyonACM, New York.
Monetization Strategies for the Web of Data
Grubenmann Tobias (2018), Monetization Strategies for the Web of Data, in WWW '18 Companion: The 2018 Web Conference Companion, LyonACM, New York.
Challenges of source selection in the {WoD}
Grubenmann Tobias, Bernstein Abraham, Moor Dmitry, Seuken Sven (2017), Challenges of source selection in the {WoD}, in Proceedings of the International Semantic Web Conference {ISWC} '17, ViennaSpringer, Heidelberg.
Decentralizing the {Semantic Web:} Who will pay to realize it?
Grubenmann Tobias, Dell'Aglio Daniele, Bernstein Abraham, Moor Dmitry, Seuken Sven (2017), Decentralizing the {Semantic Web:} Who will pay to realize it?, in Proceedings of the Workshop on Decentralizing the Semantic Web (DeSemWeb), ViennaCEUR - WS, AAchen.
Core-selecting payment rules for combinatorial auctions with uncertain availability of goods
Moor Dmitry, Seuken Sven, Grubenmann Tobias, Bernstein Abraham (2016), Core-selecting payment rules for combinatorial auctions with uncertain availability of goods, in Twenty-Fifth International Joint Conference on Artificial Intelligence, 424-432, AAAI Press, Palo Alto, CA424-432.
A Double Auction for Querying the Web of Data
Moor Dmitry, Grubenmann Tobias, Seuken Sven, Bernstein Abraham (2015), A Double Auction for Querying the Web of Data, in The Third Conference on Auctions, Market Mechanisms and Their Applications, European Alliance for Innovation, Ghent, Belgium.
Collaborative Streaming: Trust Requirements for Price Sharing
GrubenmannTobias, Dell'AglioDaniele, BernsteinAbraham, Collaborative Streaming: Trust Requirements for Price Sharing, in 4th Workshop on Real-time & Stream Analytics in Big Data & Stream Data Management, IEEE Computer Society Press, Washingron DC.

Scientific events

Active participation

Title Type of contribution Title of article or contribution Date Place Persons involved
The 14th Workshop on the Economics of Networks, Systems and Computation (NetEcon) Talk given at a conference Data Markets with Dynamic Arrival of Buyers and Sellers 28.06.2019 Phoenix, USA, United States of America Moor Dimitrii;
The Web Conference 2018 – PhD Symposium Talk given at a conference Monetization Strategies for the Web of Data 25.04.2018 Lyon, France Grubenmann Tobias;
The Web Conference 2018 Talk given at a conference Financing the Web of Data with Delayed-Answer Auctions 23.04.2018 Lyon, France Bernstein Abraham; Grubenmann Tobias;
DeSemWeb: Decentralizing the Semantic Web Talk given at a conference Decentralizing the Semantic Web: Who will pay to realize it? 22.10.2017 Vienna, Austria Bernstein Abraham; Grubenmann Tobias;
The 16th International Semantic Web Conference Talk given at a conference Challenges of source selection in the WoD 21.10.2017 Vienna, Austria Grubenmann Tobias; Bernstein Abraham;
25th International Joint Conference on Artificial Intelligence (IJCAI) Talk given at a conference Core-Selecting Payment Rules for Combinatorial Auctions with Uncertain Availability of Goods 09.07.2016 New York, NY, United States of America Seuken Sven; Moor Dimitrii;
Third Conference on Auctions, Market Mechanisms and Their Applications (AMMA) Talk given at a conference A Double Auction for Querying the Web of Data 08.08.2015 Chicago, United States of America Seuken Sven; Grubenmann Tobias;


Associated projects

Number Title Start Funding scheme
167303 PIG DATA: Health Analytics for the Swiss Swine Industry 01.06.2017 NRP 75 Big Data

Abstract

The Semantic Web is to computers what the traditional Web is to humans. The goal is to expose data in a way such that machines can easily find the information they are looking for. If implemented properly, a semantic search would lead to the desired results much faster than using a traditional search engine. To enable this vision, the data needs to be exposed in a machine-readable format, and RDF has to be adopted as the de-facto representation of Semantic Web documents. RDF documents co-exist with HTML documents and also link to other RDF documents, leading to the so-called Web of Data (WoD).In contrast to the traditional Web, the majority of WoD datasets only exist due to subsidies from research projects or governments. This leads to a number of problems. First, when a research project ends, it is often the case that an endpoint simply disappears (because servers are costly). Second, RDF data is often out of date or has otherwise low data quality (because maintaining data is costly). Third, and most importantly, the majority of data that exists somewhere in some form is simply not available as RDF documents (because creating the RDF dataset and making it available is costly). The reason for these problems is simple: on the WoD, datasets are usually queried by algorithms rather than viewed by people, which removes the traditional incentives for publishing and maintaining data. In particular, advertising, the main source of income for search engines, does not work in this environment where machines process the data and automatically filter out unwanted information. This raises the question: How can the WoD free itself from purely relying on government subsidies and overcome the problems listed above?We propose the use of a global market for querying the WoD, where queriers must pay for receiving results, and data providers get paid for answering queries. This creates a financial incentive for data providers to publish and maintain data. Additionally, this also enables the economically efficient allocation of resources in the WoD. Our approach takes into account the value from querying service-providers (or end-users), as well as the data providers costs. In contrast to prior work, we use market mechanisms that handle the “combinatorial aspects” of the problem in a principled way. Instead of using a “large-market” argument or “market equilibrium prices,” our market mechanisms are grounded in mechanism design theory which allows for a formal analysis of the market’s efficiency and incentive properties.Our explorations break new ground in large-scale data organization. While prior work has investigated the technical means for the Semantic Web and Linked Data for the past 12 years, these explorations have never included the economic viability; it was always assumed that money would come in some way. The scientific impact of our work, therefore, lies in spearheading new investigations into the economic viability of the Semantic Web vision. Furthermore, this project pioneers investigations in pricing truly large, diverse, and complex information services a topic of utmost importance with the advent of big data and data science.
-