What do we

Abstract

This paper describes preliminary work in the uncovering of mashup patterns in order to find new directions for the design of mashup tools. We conducted a qualitative survey of high-quality mashups, as nominated in two popular mashup directories, and examined how the mashups made use of existing websites or improved upon them, how data from multiple websites were combined, and what kinds of user tasks these mashups might be suitable for. We describe a set of patterns what we found in our sample of mashups.

Mining social tags to predict mashup patterns

“…Moreover, further analysis of the results proves that this approach enables predicting tag mashability across all different mashup categories from [15].…”

“…Their results are particularly useful for designers of Web mashup tools. In [15], the authors categorize mashups surveyed from ProgrammableWeb under the following categories, which are referred to later in this thesis:…”

“…Through further analysis of these results, it is found that the tag similarity approach used in experiment 0 is successful in discovering mashup patterns falling under the "Aggregation" mashup category [15], referring to mashups aggregating data from APIs offering similar functionality. Examples include "Only Apple Stories", a mashup aggregating Apple Inc. news queried from Digg [50] and Delicious APIs, both tagged with "bookmark", and "Google vs Yahoo Maps", a mashup displaying locations side-byside from Google Maps and Yahoo Maps [51] APIs, both tagged with "mapping".…”

“…This social aspect has attracted the attention of a number of researchers, and enabled the study of many interesting phenomena related to online user behavior and mashup creation trends [13] [14] [15].…”

End-User Programming for Web Mashups

“…The research efforts behind the design of EUP approaches for mashups (so-called "mashup tools") have resulted in the growth of this field as an interesting research topic spanning areas including Model-Driven Development (MDD) [6], programming languages [7], software and service composition [8], and HumanComputer Interaction (HCI) [9]. Existing mashup tools can be classified according to the EUP technique [10] they utilize as follows:…”