People Search with PeekYou

Peekyou is one of the new kind of people search services. It allows you to search people either by name (first and last) or username. One can indicate the search realm. It is also possible to add tags to a search query. The query result is presented either as a list, a grid or as a map. One can also refine the query by ###, age or location. I like this presentation a lot.

The results of a query are very US oriented, so it has limited appeal to me. Also here one finds integration with the Intelius-service.

Clicking on my name one can zoom in a bit further and see some (11) services, where I have registered. Also links to web-pages (6), where I am mentioned ar listed.

I tried to sign up for the service in order to correct some information, but never managed to get signed up. I gave up, a pity.

Categories/tags: MicroContent
PermaLink • Comments • TrackBacks

People Search Classic

The classic way to search for people is by using whitepages. This is a strictly local method for searching, i.e. each country has its own set of pages. Often the search possibilities are restricted. For instance one can only search by city and can not do a country wide search. The information obtained can be limited, for instance no first names.

The idea of whitepages has been pushed to include mobile phonenumbers and email-adresses. However it seems that this never caught on and the directories that used this remained limited.

In the US the idea of whitepages has been extended to include any public record. The Intelius service seems to have monopoly on this information, but much information is behind pay walls. The Intelius data is syndicated by many third parties and you see it on many sites. Interestingly each syndicator reveals something that others keep hidden.

Google also incorporates phonenumbers, but you have to use a keyword such as rphonebook:. And it only works for the US.

These are good solutions if one needs a phone number quickly, but remain a bit limited.

Categories/tags: MicroContent
PermaLink • (1) Comments • TrackBacks

People Search Images

A natural thing to do is to search for images of oneself. For this one could do a search on Google. The results for my name are pretty good. The first 40 images are either of me or by me. But looking further on things go wrong. I start to see images that have nothing to do with me. And this is due to the algorithm that Google uses to associated images on a web-page to text on that web-page.

So there must be better methods / services to search for my face. MicroContent search should come to the rescue here.

Categories/tags: MicroContent
PermaLink

People Search by free text

The most obvious way to search for People is by using a standard search engine, such as Google or Yahoo. So let’s see how well the search engines are doing in finding my name. I judge the results by two things: completeness and relevancy.

For completeness I look at the position of my most important ramblings on Internet: my blog. For me my accounts on other services should follow, but I guess those are not important to everyone. I should add some obscure page and see where it appears.

For relevancy I just locate the first irrelevant result, i.e. a web-page where my name does not occur. There are many pages that do not offer interesting information on me, but I will not discuss those.

Google results:

• arnaud leene : about 38.400 results, first irrelevant result at #189, blog #2;
• leene arnaud : about 38.500 results, first irrelevant result is at #71, blog #2;
• “arnaud leene” : about 2.060 results, can’t go beyond #199, no irrelevant results, blog #2;
• “leene, arnaud” : about 146 results, no irrelevant results, blog not listed;

One can repeat this with other search engines, but it does not change the overall picture. The main lesson is that it works very good, but do use quotation marks to indicated the entire name. I am lucky that I got a worldwide unique name, so there are very few irrelevant results. Irrelevant results arrive from changed pages and from spam sites, that inject names into the Google index.

However is this good enough? In this way you do not find my images, the music I listen to, my email addresses, telephone numbers, etc. I guess one could add specific keywords to get this information, but there are better solutions.

Categories/tags: MicroContent
PermaLink

People Search Requirement

So what do we want, when we say People Search? Well, I want to see every information on a person: his blog, his phone number, his pictures, quotes citing him, etc. So any web-page where his name occurs in the text, any MicroContent Item, where his name appears in a field or in the meta-data, etc.

So I want to do queries with for example: lastname=“leene” and firstname=“arnaud”. And that is where the problems begin. Search engines do not know the concepts of lastname and firstname. And even if they did know, the data is not encoded for these concepts. We are talking semantic search here.

I am still talking simple queries from the standpoint of a genealogical researcher. As a research I would like to add dates (birth, death, residence), places, family members, roles, etc, etc. Fortunately there are many genealogical services, where one can ask such queries. Unfortunately the search realm of these services is limited.

In the field of free text search there seems to be not much progress. Fortunately social network sites and MicroContent come (a bit) to the rescue. And it all centers on identity.

And one should judge the results of queries on their relevancy (what percentage of the results are relevant to the query?) and completeness (do the results include the answers that I need / expect?).

Categories/tags: MicroContent
PermaLink • Comments • TrackBacks

People Search intro

Last week I picked up my genealogy hobby. About once a year I do a Google search on my lastname, in order to see whether there is anything new in my family. Thus I am slowly building up a family tree with basic information.

This year I nosed around several social network services, such as LinkedIn, Hyves and Schoolbank. This provided me with a lot of new information.

It is getting more easier and easier to find information on persons. We are not here yet however. As this subject is related to semantic searching and MicroContent, I thought it worthwhile to write something about it.

Categories/tags: MicroContent
PermaLink • Comments • TrackBacks

BusySync

Another application, BusySync, that does syncing calendars came in the news today, thanks to the Eddy they just won. So an opportunity to try it out. This application does a bit more than the Calaboration application that I just looked at. BusyMac allows for syncing of iCal calendars on your local network and between different users on a single machine.

This application installs itself as a preference-pane, which seems second best location. The best would be an installation within iCal itself, but I guess iCal does not support these kind of extensions.

In the preference-pane the user can specify which of the calendars must be published (read and/or write access) and one can subscribe to published calendars on the network. It also allows for publishing and subscribing to Google calendars. Makes you wonder about interference with Calaboration.

And the preference pane allows you to reset the things, start afresh and resolve conflicts. Which shows that syncing remains a difficult business.

Categories/tags: MicroContent • client • application • type • event , macosx, busysync
PermaLink • Comments • TrackBacks

Calaboration

I just downloaded, installed and ran Calaboration. This small application creates a Caldev-link between the iCal application and Google Calendar. I am not a Google calendar user, but as far as I can see, calendars that have been defined in Google Calendar will be created in iCal as well.

Then one can create an event in either iCal or Google Calendar and the two calendars will be synchronised.

I do not know how well it works, but it made me thinking about replication, duplication and synchronisation of MicroContent. Should dive into this a bit.

Categories/tags: MicroContent • client • application • type • event
PermaLink • Comments • TrackBacks

OLE anyone?

Does this still exist? Is it used? On Windows?

All I know that it is not available on MacOSX. It was available on NeXTSTEP, but got lost in the transition. It was mashing avant le lettre in some sense.

Categories/tags: MicroContent • mashing
PermaLink • Comments • TrackBacks

Tarpipe

I have been playing a bit with Tarpipe. This service allows users to pipe together services from third parties. So send an email, have the attached image uploaded to Flickr and post the resulting picture URL to Twitter.

The idea is great, but it is not yet robust enough. I was only able to get an workflow run twice. And some other activities are in the pipeline (I hope). The Dropipe application ended with an error message. It can be that the service is now overloaded due to its exposure on ReadWrite web. So I will come back to it later.

The documentation still needs a lot of work. It is totally unclear what some services do.

Categories/tags: MicroContent , service, tarpipe
PermaLink • Comments • TrackBacks

Bento feedback 1

Since a week I am using Bento for two different projects. The application offers a great flexibility, which is great. For such a young application there are naturally still many improvements that can be made.

Bento is called a personal database system and can (and should) not be compared to a full fledged relational database application. And this is also the challenge for the developers: what functionality can be added without becoming to complex. I have the feeling that quite a lot can still be added without becoming to advanced. It should have not become more complex than one of Apple’s iApps.

In my first testcase I created a flat database, such as a collectioneur might use. In my case the collection consists op bottle caps. I have a version of this database online at Listphile. As you can see there, a record consists of a title, a description, an image, links, etc.

1. Tags - Bento does not know tags as a concept. It is possible to add tags, but then I lack facilities that I am used to, as in Sandvox or WebnoteHappy for instance. It is possible to use tags through a "related records list", but that is not as nice;
2. Links - Bento supports links, but in a limited fashion. Their approach is taken from Apple’s AddressBook. This implies that the title corresponding to an URL, is extreme limited. Out of the box the title can be ‘home’, ‘work’ or ‘other’. One can add other titles, but this is in no way a generic link approach;
3. Edit/Design - Bento allows the user to edit his data and change the presentation at the same time. It is very good that this is possible, but I noticed that I accidentally get in design-mode when I want to edit and vice-versa. I would like to see a locking system. This should allow me to lock-down a design when I am finished with it;
4. Limited design - Bento does not allow the user to change the design to much. I guess that this way to go is OK for this application, but as it is now it is not enough. Bento should take an approach similar to the iLife suite and introduce an Inspector. This Inspector could have multiple modes, which are now hidden under the menu of Bento;
   1. Field - allows to set and change field type and its attributes. Change the field name;
   2. Label - The label text defaults to the field name. However the user can also use another text without changing the field name. The appearance of the label (font, font size, color). The position of the label (above/below/left/right/none);
   3. Field content - This should allow to change the appearance of a field. This is a bit similar to the Form Tools, but more flexible: choice of font, font size, color of font, color of field box, etc;
   4. Page appearance - the Themes that Bento supports is extremely limited. Just look at the Keynote application for a better approach;
   5. Related List appearance - I would like to be apple to suppress the column and row header. And I certainly would like to suppress the controls at the bottom. And if I have just one 1 item in the list, I would like to see a presentation similar to the other fields;
   6. Object appearance - Now I can not change anything and yet there are many attributes that could be changeable;
5. Tabular layout - Bento forces to use tabular layout for elements (fields, related lists, objects) on a form. I guess this is OK in order to limit the complexity. I would like to see some more flexibility, where for instance a column divider does not span the whole page, but just a limited number of rows;
6. Objects - The number of objects is now a bit limited. I would like to see images for instance (my company logo!). And naturally the Inspector should allow to edit the attributes of the objects. Take the Keynote app as inspiration here;
7. Publishing - any modern data based application should allow to integrate with Internet. There are many possible ways to do this. At its basic Bento should allow to export records into HTML-files. At the next level integration with MobileMe could be possible. Take inspiration from MacGourmet here;
8. Spotlight - integration with Spotlight would be nice. Individual records should turn up in the Spotlight results;
9. Media browser - a media browser such as can be found in the iLife suite is nice. This would make adding media to records more easier. Just integrate the iMedia browser from Karelia;
10. Button bar I would like to see a standard MacOSX button bar. This would allow me to set the buttons that I need most and suppress the others. For me this is: Add record, change view (table/split/form), set form, next/previous record, show media browser;

It might seem that I have a lot on comments on Bento (and that I am not happy with it). The contrary is true. It just inspires me and I see many growth directions. The list here are just some ideas that I would use right away. It can become more complete without being much more complex. It does not have to be as simple as iTunes, but can be complex as Keynote. Anyway, I put my money where my mouth is and acquired the application.

Categories/tags: MicroContent • client • application , bento

PermaLink • Comments

Bento

Bento is a relatively new application under MacOSX. It recently got an update and is now called Bento 2. This application can be seen as a generic MicroContent Client. Bento is created by the makers of Filemaker, which is a true database application. Bento is fortunately much more simple and reminds much more of MicroContent. So let’s look at it in more detail.

Bento allows the user to create any MicroContent Type he likes. In Bento each MicroContent Type is called a library. The user can create as much types as he wants. Each Library consists of one or multiple fields. One can select from field types text, number, choice, textbox, media, time, date, duration, counting, rating, address, phone number, email address and URL. This shows the database origin of Bento. As nothing is preset, one can add any field, one can say that Bento supports Wild MicroContent.

The nice thing of Bento is that it can link to other MicroContent Types. These links are called Lists. Out of the box Bento recognizes File Lists, Message Lists, Address Lists, Event Lists and Task Lists. These Lists are by the way linked to the Finder, Mail, AddressBook and iCal applications. These fields are truly Lists as they can be linked to multiple Items.

Also nice is that one can export definitions of MicroContent Types, called Library Templates. So sharing of MicroContent definitions is thus easy for Bento users.

The main Bento screen is very straightforward and reminiscent of other MicroContent Clients. On the left one has the Libraries pane. Each MicroContent Type is indicated by booklike icon. The libraries such as Address Book, iCal Events and iCal Tasks are preset. For each Library it is possible to create handpicked and smart lists. Thus the small icon labelled ‘beer’ indicates a smart list from Capsules.

Below the libraries pane one can see a field pane. This pane allows to add and remove fields. The visibility of this field pane can be suppressed. Also the visibility of the entire library/field pane can be suppressed.

The top right pane is the standard Items pane in table format. One can drag columns around, set the sorting on a column. The nice thing of this implementation is that it works like a spreadsheet. And one can suppress the visibility of this pane.

This brings us to the view-pane. As expected this shows an individual Item. Each field is presented with its label. Note the Bookmarks and Tags Lists, which are presented as Item-panelets. Everything in Bento is editable. There is no separate edit-mode. Thus any field can be changed as needed. Some fields come with preset buttons, for external linking, etc. Below the view-pane one can see controls to add/delete/import/export/print Items.

Above the view-pane buttons allow to change forms or to change to items-pane. The layout of the view-pane is known as a form. And the user can define as many forms as he like. And this is another important feature of Bento. The user can design the view-pane form. Thus he can set which fields should appear on a form, where they should appear, etc. There is no separate design-mode either, anything can be changed, resized, etc. at will.

Naturally Bento can still be improved. I would like to seem integration with other applications, such as iPhoto, iTunes and iMovie. I would like to see some standard libraries for MicroContent Types such as bookmarks, blogs, recipes, etc. I would like to see support for tags as a field type.

Instead of the table format for the Items pane, I would like to see support other formats, such as a grid format for images. I have mixed feelings on the instant editability and designability. I would like to see some locking mechanisms to prevent errors.

The import and export facilities are reminiscent of a true database: csv and tab-delimited files. I miss XML support, such as RSS and OPML and a Internet publishing facility.

All in all a very good application for those who like to keep lists or have the need for a simple database. I guess I will be buy it.

Categories/tags: MicroContent • client • application
PermaLink • Comments • TrackBacks

Twine

I started playing with Twine. It is not very obvious from the beginning. I must find and invest some more time in it.

Categories/tags: MicroContent • client • service
PermaLink • Comments • TrackBacks

Blip.fm

I created a profile on blip.fm. It took me a long time to figure out how it worked. I hoped it would detect which songs I listened to in iTunes. I did set up audioscrobbler after all. So it is much easier to just look at my last.fm page.

So what does blip.fm add. Basically it just allows to add a microblog comment to each song that is played. I already record some of these things on various microblogs, so this does not add much.

So what is left, is the social component, the DJ. I wonder whether this adds something over last.fm. So I can’t be enthusiastic at the moment about this service.

Categories/tags: MicroContent • type • audio
PermaLink • Comments • TrackBacks

NetNewsWire Drops Microformats Support

NetNewsWire drops support for microformat. I am afraid that is a right decision. NJobody seems to use microformats within feeds/blog posts. So deleting this code will speed up NNW, which is of greater importance.

[Inspiration Chris Casciano]

Categories/tags: MicroContent • client • application • format • hformat • type • blog , netnewswire
PermaLink • Comments • TrackBacks