Cypher 1.9, Ubiquity, etc

[Sherman Monroe]

Hi All,

It's been a very busy and productive August. Cypher 1.9 is now only a couple of days from release, and is a major step forward for the technology. This release is the result of a huge refactorization of the framework, so many of the enhancements you won't readily see as new features, per se. But I'd like to give a short list of what you can expect.

MetaLanguage Ontology (MLO)
Cypher datasets are now written in a fully RDF-compliant language called MLO (pronounced "mellow"). This allows Cypher to load datasets from any SPARQLEndpoint, which now opens us to the possiblities of (in the near future) building a directory of these endpoints, then allowing Cypher to load datasets on demand on a per-input basis by searching a directory of datasets, using the features of the input as lookup keys. An example is, if a high threshold of French characters or words are found in the input, Cypher could lookup datasets in the directory that deal with French phrase grammars and lexemes. The MLO also makes viewing grammar results a lot more powerful, because now you can use your favorite RDF browser to view the grammar trees.

Bug fixes
A slew of bugs have been fixed, so all of the problems you've reported about transcoder throwing errors and now loading properly have been resolved. You can expect a more predictable experience with starting the engine and loading input from flat-files and/or the Web Services API.

Ubiquity Command
A long standing vision for Cypher is the ability to just speak to your database and call up (or insert data) to and from your own personal database, and do really interesting things with the data, so in other words, creating meshed data using plain sentences and phrases. That vision is drawing nearer and nearer to fuition with the recent announcement of Ubiquity. When I saw Yahoo! Pipes, I instantly got excited about user-generated meshes and the possibilities there. I then tried to tie-in Cypher to it, but the framework wouldn't allow me to browse RESTful data (at the time I tried, they may have enhanced libs by now). When I got hold of Ubiquity and particularly jQuery, it was an answer to a prayer. You can see the results in the new cypher-transcoder Ubiquity command. Unless you already have Ubiquity installed, nothing happens when you go to the page, otherwise, you get prompted to install the command. This is only a sound-check version, no really interesting features yet. For now, it takes your input, sends it to the Cypher demo server, retrieves the result set (if one is produced), then displays it and optionally allows you to insert the resultset into the browser window. So you can do "the people who wrote books about Apollo 13", and insert the list of authors into your email.

I believe Ubiquity has the capacity of becoming a pillar of the Semantic Web. Expect frequent enhancements to the Cypher support for Ubiquity in the coming months. Installing the command this time automatically registers you for future updates, which your browser will consume automatically.

More Lexemes
Now that we have Cypher connected to the SPARQLEndpoint spec, the endpoints are starting to grow faster than we can write lexeme's for the ontologies they deal with. We're working on lexemes for dbpedia as well as UMBEL, so for now, you can do things like "the presidents who wrote books" or "the films Naomi Watts played in". These can all be previewed on the demo site (which runs the Complete Dataset).

--

Thanks,
-sherman

I pray that you may prosper in all things and be healthy, even as your soul prospers
(3 John 1:2)