A Web site architecture and GUI for UML models search

Abstract

The tools and the way that nowadays exists for documenting doesn’t get on well with all the different existing file formats and media files. There are many ways to document software: with a video, an audio, a word document, a PDF, diagrams in different formats, discussions via email or on a chat, Internet resources and references... And you sometimes need to document another document: i.e. some the meeting reports are a document of another document (the recorded videos) and this dissertation is all the documentation of this project piled up. UmlModels is trying to gather the main existing ways to document and embed them into an area/place easily accessible everywhere you can connect to the Internet. Many advantages can be taken from here such us no worries storing the versions, the storage, wrong references to renamed or moved documents and many, many others. Here, the door is open to a huge world of new revolutionary ideas. UmlModels is also a search portal with superior semantic content. UmlModels provides a search service of UML‐Models addressed to software community engineers. This provided service does not exist on the Internet and aims to be innovative. Unlike other search engines such as Google, UmlModels provides intrinsic information of the artifacts (see ‘An Artifact for UmlModels’) in addition to the context information. Full machines entirely dedicated to search software artifacts (crawlers) are day and night searching information on the Internet that can be relevant for the software community. The retrieved information is studied, indexed, organized and offered to the users of UmlModels into a representative Model. The work presented here is the first step of a constant evolution towards a new conception of the software documentation. Nowadays the portal is able to provide almost any kind of software artifact; however only images are automatically searched by the crawlers so only software diagrams are indexed. On the future, as soon as we give intelligence enough to our bots to extract semantic from other elements such as source code, packages, assemblies and many others, will be searched into source code files, zips, textual documents such as docs, power points, PDFs and so on.