Covering RSS and Data Integration on the Web
One of the very cool thing about Dapper is that it can take content from both normal public websites as well as protected websites. (For full details on how to use Dapper, see one of our earlier posts, but typing in Dapper into the search box.) When a secure websites is entered, it asks for your login information. This can allow Dapper to be use for corporate purposes or simply pulling information from websites that require a subscription.
Feed readers are becoming more popular as feeds have become more popular. However, feed readers have the following limitations.
Is this all that interesting to look at? Also, if I am interested in say articles from the Lancet, do I really want an interface to show me so many other topics? This is not a natural way that humans review and understand information…at least in our view.
For some reason we don’t think feed readers are the future. Instead technologies that all feeds to integrate with existing web pages and blogs makes much more sense to us.
In the previous post we looked at the state of real estate feeds. This time we turn our attention to medical feeds. One place we looked was at the New England Journal of Medicine. We found something similar to the feeds at company employment pages and realities. That is searchers could not be converted into RSS feeds (and thus would require a feed screen scraper). See the view below.
so we are able to find articles, however we can not create a feed from this page without using a third party product.
As for the feeds that the NEJM does have, they are very broad as one can see from reviewing what they have to offer.
Real estate is a perfect area that could really benefit from targeted RSS feeds. However, our review of various real estate sites and MLS services showed very little focus on RSS feeds, or any other feeds for that matter. One site that did have feeds ZipRealty, did not offer a way of filtering the feeds, thus an external filter would be necessary.
Why ZipRealty thinks that a person would want to see an unfiltered feed of an entire city of listings is beyond us. Likely they just don’t know and are new to feeds. Our conclusion from this and many more examples is that real estate is ripe for taking advantage of feed technology.
We mention creating a feed from an html screen in this post:
http://semanticwebs.wordpress.com/2008/12/02/dapper-create-feeds-from-web-pages/
What this means is that html based UIs now have a real advantage over non html based UIs. SAP is just such an application that does not have an html based UI. RSSBus excels at creating RSS feeds from things like SQL databases, but they do not yet have hooks into SAP (although they may be developing some). GuiXT on the other hand exposes SAP fields to an html page, making it possible to scrape this page using a tool like Dapper. This means that there is the possibility for using it for application integration.
Discussion with GuiXT
However, after speaking with a GuiXT representative we learned that they do not provide tempates or make templates for clients. Instead they sell a development environment. This would mean that a client would need to create the front end in GuiXT in order to scrape it and extract the feilds. Because of this, the client would need to already be committed to creating the front end for other purposes such as user interactions.
Technorati Tags: ERP integration, SAP integration, RSSBus, GuiXT, Dapper, Screen Scrapping
This was originally published in our blog http://wordpressideas.wordpress.com. However it applies just as well to this blog. Essentially we explain how a folder in Box.Net can be entered into WordPress and how users can then subscribed to a feed that is just for the files represented in this folder.
Article
_________________________________
This is functionality that eluded us for a while. We wanted to share different folders with the WordPress Box.net widget. All one has to do is perform a right mouse click on a folder and you will receive the dialog box with the embedded folder code.
You can then goto your widget page and enter the code.
Furthermore, this code can be entered right in WordPress as well. Here you can see what the output looks like.
Finally, your viewers can actually subscribe to an RSS feed from this file that is located in at Box.net.
The next area that we think will be a big focus area is improving the filtering of feeds. We do some of our filtering in DabbleDB, and of course many feeds, such as job boards come with sophisticated filtering capability in the pre-feed selection. However, many feeds do not. What seems more an more likely is that we will all have a filtering program siting between our web feeds and our eventual consuming software (web page, feed reader, dashboard, etc…) This book on Amazon seems to have the right idea.
“Product Description
The increasing amount of information on the internet makes it hard to filter out relevant parts. Every second hundreds of news-articles and blog-entries are written. A subscriber to this constant stream of information quickly looses overview – a typical needle-in-a-haystack problem. Especially in the field of news-reading, software can help to increase efficiency by personalized filtering mechanisms. This filtering can be predictive based on previous user-choices, but also heeds the power of social networks: Users help each other by marking relevant information. The author Ingo Schommer investigates the conceptual and practical development of a web-based news-reader with advanced filtering features. Research is conducted in the field of information visualization and filtering as well as in competitive products and their shortcomings. The solution will mainly target pro-users who are already accustomed with news-reading, the modern “information junkies”.”
Before seeing this service we never really thought of creating feeds from scratch. However, IceRocket allows you to create a free form feed. Here we create a feed representing our blog spplan.wordpress.com. There is not much content here, really just more of a test.
After we have set this up we can save and then select the XML button in order to be taken to a URL of our feed.
Here we can see both the URL and how the feed would show in a browser or in a feed reader.
This brings up interesting concepts in interweaving your feed with web pages. For instance, many web pages could now have a single feed, that is created in a custom manner. When the feed is changed, all the pages change at once, instead of having to go and update every page separately.
We found that Google Alerts (not affiliated with Google) does such a great job of scouring the net for specific search results, that we have given it our Gold Medal for Excellence in Feed Technology.
There are many feed creation services out there like Rocket News and Topix.net, however these excel at creating News Feeds. Google Alerts is more of a generalized search and feed creation tool. We were most impressed with its results.
As for the service offerings, we will let you do your own by digging into their site http://www.googlealerts.com. A few of the features we liked were:
You can see how specific the searches can be by selecting the advanced options on the search.
Finally, we want to point out that the RSS feeds created by Google Alerts are highly readable.
In our last post we discussed the feeds available off of Nike’s job website, and found them very limited and overly broad. Because of this, Nike’s feeds, while better than many companies that have no job feeds at all, are not all that useful.
FilterMyRSS
This service allows you to filter any feed by only removing the line items that do not contain what is entered in the keyword box. This is an important website, as so many RSS feeds simply export everything and so many companies are behind the curve on providing filtering at the source.
However, we have not gotten the feed to form properly in WordPress (and we are using an RSS plug in)
Feed Rinse
A second feed service is called Feed Rinse. Again, we paste our feed into the entry box.
What we notice is that the filter is a bit more clean and less free-form than FilterMyFeed.
You then grab the link by right mouseclicking the RSS button an selecting copy link.
Again, we push most of our feeds out to WordPress, and we had a problem getting FeedRinse feeds to show in WordPress. For this reason we prefer to perform our filtering inside of DabbleDB.
You can see our blog on DabbleDB if you are interested in learned more about this software. We don’t want to write much here has we have a separate blog for DabbleDB. http://dabbledb.wordpress.com
