Prof. Dr. Cesare Pautasso

Mashup Development with Web Liquid Streams

Andrea Gallidabino, Masiar Babazadeh, Cesare Pautasso

1st International Rapid Mashup Challenge (RMC 2015), Rotterdam, The Netherlands, pp. 98-117

2016

Abstract

Web services such as Twitter and Facebook provide direct access to their streaming APIs. The data generated by all of their users is forwarded in quasi-real-time to any external client requesting it: this continuous feed opens up new ways to create mashups that differ from existing data aggregation approaches, which focus on presenting with multiple widgets an integrated view of the data that is pulled from multiple sources. Streaming data flows directly into the mashup without the need to fetch it in advance, making it possible to exchange data between mashup components through streaming channels. In this challenge submission we show how streaming APIs can be integrated using a stream processing framework. Mashup components can be seen as stream operators, while the mashup can be defined by building a streaming topology. The mashup is built with Web Liquid Streams, a dynamic streaming framework that takes advantage of standard Web protocols to deploy stream topologies both on Web servers and Web browsers.

Download

PDF: ▼wls-icwe-rmc2015-mashup_challenge_web_liquid_streams.pdf (6MB)

Citation

Bibtex

@inproceedings{icwe:2015:rmc:wls,
	author = {Andrea Gallidabino and Masiar Babazadeh and Cesare Pautasso},
	title = {Mashup Development with Web Liquid Streams},
	booktitle = {1st International Rapid Mashup Challenge (RMC 2015)},
	year = {2016},
	pages = {98-117},
	publisher = {Springer},
	address = {Rotterdam, The Netherlands},
	abstract = {Web services such as Twitter and Facebook provide direct access to their streaming APIs. The data generated by all of their users is forwarded in quasi-real-time to any external client requesting it: this continuous feed opens up new ways to create mashups that differ from existing data aggregation approaches, which focus on presenting with multiple widgets an integrated view of the data that is pulled from multiple sources.
Streaming data flows directly into the mashup without the need to fetch it in advance, making it possible to exchange data between mashup components through streaming channels. In this challenge submission we show how streaming APIs can be integrated using a stream processing framework. Mashup components can be seen as stream operators, while the mashup can be defined by building a streaming topology. The mashup is built with Web Liquid Streams, a dynamic streaming framework that takes advantage of standard Web protocols to deploy stream topologies both on Web servers and Web browsers.},
	keywords = {mashup development, rapid mashup challenge, web liquid streams}
}