Mashup Camp IV: Speed Geeking 1

Taking a break from the sessions just before lunch; there’s nothing that I want to go to right now, and this gives me a chance to catch up on my notes from yesterday’s speed geeking before today’s session starts after lunch.

Keep in mind that each session was a maximum of six minutes longer (shorter if the presenter didn’t actually restart when the siren went off), so there are likely to be errors and omissions here.

Here’s what I saw yesterday:

Voice + TWiki: This is an integration of LignUp‘s VOIP service with TWiki to allow for the integration of voice information into the wiki environment. Although the author claimed that this was for non-technical people (he actually used the phrase “over 40”) to be able to add content, Twiki is an easy enough environment that I don’t think that this is really necessary, but one of the use cases that he showed is more compelling reason: someone is out in the field without internet access and wants to add a comment to a particular page related to that field location. For example, someone at a building site sending in a verbal report of what’s going on; I can also imagine insurance adjusters doing something similar. If you’re on the wiki and doing it interactively, clicking a button causes LignUp to call your phone and prompts you through the recording process. If you’re in the field, you can call a number, enter a PIN and the desired project (page) ID, and record. The next step for this type of integration would be to convert the voice to text, although I’m not sure if LignUp offers that sort of service; I suspect in any sort of inspection scenario, you wouldn’t want to trust an automated conversion but would still want the voice recording.

5thbar: Not the place that you go after the 4th bar, 5thbar is a mashup of information about mobile phones and accessories that provides a good one-stop site. It combines blogs and news via feeds, videos via the YouTube API, listings of the device for sale via the eBay API, and allows signed-in users to add tags and reviews. Although some carriers provide a subset of this information for the phones that they carry, 5thbar covers information for a wide range of phones across multiple US and Canadian carriers. It doesn’t yet provide comparison charts between devices, although the author sounded like he’s considering that.

Fast Mash: This is a demonstration coming from a university research project, showing how unstructured data can be extracted from sites and served up in a more structured format by the use of a reference data set in the same domain of knowledge. The example that he showed was taking “cars for sale” listings from Craigslist and creating a structured, searchable data set, more like a relational/tabular form, for further filtering and display (potentially as a data feed to other mashups). He did this in part by using Edmunds car buying guide as a reference data set in order to understand the particular tokens that might occur in the unstructured data; this is the key to the technique.

ClubStumbler: Finally, someone’s doing something useful with Google Maps: a way to plot the optimal route between a number of different bars/clubs that you select in your area. 🙂 This uses the Eventful API to find out what’s going on at the clubs as well as Google and Flickr APIs, plus the author’s own API for finding the best route that he’s developing and commercializing. Seriously, though, this could be used in a number of commercial and business applications, such as optimizing delivery routes or real estate tours.

Plaxo Pulse: This mashup discovers information about all of your Plaxo contacts and feeds it back to you in a feed reader-type style (or it can be subscribed to in a regular feed reader). This includes any information found about the people from blogs, Flickr, Digg and other locations: sort of an automated identity aggregator based purely on feeds and RSS APIs from these sources. Although I don’t recall hearing the details, I suspect that the auto-discovery is done on some combination of what information that’s in the Plaxo contact record plus searches on the various content sites based on name and email address, since many allow the positive identification of a person if you know their email address. There’s also a new version of Plaxo out that looks less virus-like than the predecessors, I might give it a look. This is essentially a playlist of web-based video extracted from multiple sources (e.g., YouTube, Google Video) using the media RSS format, and played as a continuous stream. There are several pre-made “channels” of content of certain types, or you can create your own channels for private use or sharing. The Find And Play feature allows you to enter either a website URL that contains video or search terms, and get back a continuous-playing playlist of matching videos so that you don’t have to click your way through each result sequentially. This is a social money website that allows you to collect money from friends, family, roommates, etc. for specific events or purposes, such as “mom and dad’s anniversary present” or “this month’s electricity bill”. You organize a group around whatever the event is that requires payment, then send out invoices to people. They can pay by eCheck or credit card, even if they don’t have a PayPal account, and the person organizing the group doesn’t have to have a credit card merchant account to accept payment by credit card. There is a 3% fee, of course, since those transfers have to be paid for in some way through the various financial services organizations that billhighway uses to process payments behind the scenes. Payments to the group are organized and reported as such, so that you know how much money that you received by group/event rather than just by person, and you can get RSS feeds of some of the data on the site such as transactions with your contacts. $US only at this time. Although that domain’s not actually used, that name is on the site that was demonstrated so I’ve used it for naming purposes. I found this mashup particularly interesting since I’ve wanted to do something like this with Toronto’s library system ever since I started attending Mashup Camp last year. This uses an Amazon wish list to hold books that you want to read (since the Amazon API can retrieve the wish list using only your email address), then looks up the books on several different library sources (OCLC and LibraryThing were mentioned) to generalize the search by mapping the single ISBN from the Amazon wish list to all ISBN’s for that item, such as paperback, hardcover, audio and video versions. Metadata is also extracted from Amazon and the other library sources for display, and the local library systems (in this case, the author lives at the intersection of four overlapping library systems so searches all of them) to present a list of what’s available at the branches and allow holds to be placed on items. I so want this for myself in Toronto!

LignUp: I’ve already seen one mashup created using LignUp, and this was another one by someone from LignUp showing how to add voice alerts and SMS messages on a system administration functionality mashed up with the Intel Web 2.0 TDK. For example, a part fails or storage space starts to run low, and the designated system administrator receives an automated call or voice message, then can enter their current location by zip code and receive an SMS message with the address of the closest location to buy replacement parts. I moved on to the next mashup and found that it was also LignUp, showing voice mashups within Facebook: voice annotation of pictures by calling in, or web calling out either for 2 or 3-person calls or an automated broadcast message to an entire group. Another application, ReachMeRules, handled inbound calls and voice mails through the use of a cut-out number that masks your real phone number, and can provide specific responses to specific inbound numbers. This latter sounds like some similar functionality to what’s on Iotum, except that LignUp is offering a development platform that allows this to be created, not a specific fixed service.

SnapLogic: SnapLogic showed a mashup that highlighted the functionality of their data duration framework (available as open source under GPL), extracting data from Quickbooks and, matching it up and displaying it as a combined table. Although the methods of extracting from these two data sources are quite different, the SnapDojo mashup demo showed how they can be wrapped and then consumed in a consistent fashion using SnapLogic’s APIs via JSON. The mashup by showed Google Street View on a mobile phone, using techniques originally discovered during WhereCamp. They have support for Sprint phones — I saw a Samsung demonstrated — for an extra $2.99 on your monthly bill, and unofficial support for the Blackberry. I believe that it’s all running in a mobile browser rather than downloading an app to the phone, although I’m not completely sure.

Twitterlicious and Bookmark Cleaner: Twitterlicious is a mobile app that allows you to capture your tweets (if you don’t know what Twitter or tweets are, move on) as or Ma.gnolia bookmarks with the tag “tweet” so that you can review them later, especially useful when a tweet contains a URL that you can’t open on your mobile device. They’re stored as private bookmarks, so I’m not sure if they’ll get picked up in the automatic Links blog posts that I create from my bookmarks; obviously, I wouldn’t want that to happen. The use of to store the tweets for later review then necessitated the Bookmark Cleaner, a utility to delete all bookmarks with a certain tag; he’s also extended this to do some other useful cleanups on your bookmarks, such as finding and deleting all bookmarks that point to a 404 (page not found), and finding all bookmarks that appear to point to a malware or phishing site, which can then be previewed using and deleted as required. In the future, he also want to find 302’s (permanent redirect) and replace the bookmark URL with the redirected URL.

Today’s speed geeking is just about to start, I’m out of here…

Leave a Reply