One of the major problems of web sites is to keep their content ... web page is published it's content becomes obsolete. ... , to describe.
International Science 2.0 Conference EEXCESS Final Conference
Content Injection for Web-Browser
Content Injection One of the major problems of web sites is to keep their content up-to-date. Soon after a web page is published it’s content becomes obsolete. Using content injection technologies can help to reduce this issue. Content injection is a technique used to include additional data into a web page at the time it is displayed. The traditional way for content injection is to embed an HTML-link into a web page. Clicking on the link loads the content associated with it and displays it to the user. A more advanced way is based on database queries. Averagely complex websites and well-defined dynamic data, e.g. data of a weather forecast, can use an automated request to a database to fetch the content and integrate it into the web page. All these methods, to add additional content, are based on information that becomes obsolete quickly, because it had been integrated by the website author when writing the original content of the HTML page. Thus, fixing the information related to the content description at the time of generating the web page can be described as an early binding of information. Obviously this kind of content injection shares the same problem as the HTML-content of a web page, it rapidly becomes obsolete!
Self embedding Characteristic Hyperlinks To overcome the problems of content injection based on early binding the SeCH– web-browser is based on late binding. Whereas early binding uses a predefined description of the content, e.g. the HTTPaddress of the web page, late binding relies on a symbolic description to define the information that is going to be injected. To ensure the quality of the injected content, the symbolic description is integrated by the website-author into the web page. Currently three different tags are used: 1. , to describe the overall context of the web page. 2. , to describe the context of a self-contained text entity. 3. , to define the place and specific content of the content injection At runtime, right before displaying the content of a page, the SeCH–web-browser extracts the symbolic description and creates a query for the EEXCESS-recommendation service for every . As a result, every time the SeCH-webbrowser displays a web page these symbolic descriptions generate an up-to-date content injection. In contrast to fully automated content injections based on text-mining algorithms, where the website author looses control over the injected information, the SeCH– approach still allows the author to take care of the quality of the content injections.
SeCH-Prototype The SeCH-prototype is an ongoing research project at the University of Applied Sciences Hof. The iOS implementation is being developed by 8 bachelor students studying „Mobile Computing“. Furthermore an OS X based query-tool for the EEXCESS recommendation service has been developed. Besides further improvements of the software architecture additional software elements will be included: 1. As a supplement to the EEXCESSrecommendation service we integrate the search engine DuckDuckGo to supply answers for the search-queries. 2. The “force-touch” technology of the iPad Pro will be used to develop a new way to display the injected content. Within the scope of further research the following topics will be explored: • Link-boosting as a technique to generate SeARCH-Links on the fly using already existing HTML-link information. • Enhancing the queries for different search engines to improve the quality of the results. • Re-ranking the results of search engines to incorporate a user profile. • Integration of an explanation-module for the re-ranked results. Contributors: • Implementation: Andreas Netsch, Alexander Pöhlmann, Burak Erol, Brian Mairhörmann, Gottfried von Recum, Lothar Mödl, Philipp Winterholler, Andreas Ziemer • EEXCESS project contact: Dr. Christin Seifert
