This project has received funding from the European Union’s Seventh
Framework Programme for research, technological development and
demonstration under grant agreement no 261712.
Even if KAF is the true OpeNER data representation format, it is not very friendly for humans. It is also difficult to parse and work with it, specially for new users that are not familiarized with KAF. This is part of the feedback we got from the OpeNER hackathon held in Amsterdam during first days of July. To ease the post-processing of the results by any potential user, we are designing an alternative JSON format to output translate KAF, and turn its complexity into a much more usable format. Although, the specification for this JSON is currently work in progress, we have already a version that covers the most relevant parts from the KAF representation.
This software is part of a larger collection of natural language processing tools known as “the OpeNER project”. You can find more information about the project at the OpeNER portal. There you can also find references to terms like KAF (an XML standard to represent linguistic annotations in texts), component, cores, scenario’s and pipelines.
Installing the kaf2json can be done by executing:
gem install opener-kaf2json
Please bare in mind that all components in OpeNER take KAF as an input and output KAF by default.
You should now be able to call the kaf2json as a regular shell command: by its name. Once installed the gem normally sits in your path so you can call it directly from anywhere.
This aplication reads a text from standard input in order to convert it to JSON.
cat some_kind_of_kaf_file.kaf | kaf2json
This will output (JSON Format):
{
"text" : "Beatrix Wilhelmina Armgard van Oranje -Nassau (Baarn , 31 januari 1938 ) is sinds 30 april 1980 koningin van het Koninkrijk der Nederlandén ",
"language" : "nl",
"terms" : {},
"sentiments": [],
"entities" : {},
"opinions" : {}
}
You can launch a webservice by executing:
kaf2json-server
This will launch a mini webserver with the webservice. It defaults to port 9292, so you can access it at http://localhost:9292.
To launch it on a different port provide the -p [port-number] option like this:
kaf2json-server -p 1234
It then launches at http://localhost:1234
Documentation on the Webservice is provided by surfing to the urls provided above. For more information on how to launch a webservice run the command with the -h option.
Last but not least the kaf2json comes shipped with a daemon that can read jobs (and write) jobs to and from Amazon SQS queues. For more information type:
kaf2json-daemon -h
This component runs best if you run it in an environment suited for OpeNER components. You can find an installation guide and helper tools in the OpeNER installer and an installation guide on the Opener Website
At least you need the following system setup:
The component is a fat wrapper around the actual language technology core. You can find the core technolies in the following repositories:
If you encounter problems, please email support@opener-project.eu or leave an issue tracker.
git checkout -b my-new-feature)git commit -am 'Add some feature')git push origin my-new-feature)This aplication reads a text from standard input in order to convert it to JSON.
cat some_kind_of_kaf_file.kaf | kaf2json
This will output (JSON Format):
{
"text" : "Beatrix Wilhelmina Armgard van Oranje -Nassau (Baarn , 31 januari 1938 ) is sinds 30 april 1980 koningin van het Koninkrijk der Nederlandén ",
"language" : "nl",
"terms" : {},
"sentiments": [],
"entities" : {},
"opinions" : {}
}
You can launch a webservice by executing:
kaf2json-server
After launching the server, you can reach the webservice at http://localhost:9292.
The webservice takes several options that get passed along to Puma, the webserver used by the component. The options are:
-b, --bind URI URI to bind to (tcp://, unix://, ssl://)
-C, --config PATH Load PATH as a config file
--control URL The bind url to use for the control server
Use 'auto' to use temp unix server
--control-token TOKEN The token to use as authentication for the control server
-d, --daemon Daemonize the server into the background
--debug Log lowlevel debugging information
--dir DIR Change to DIR before starting
-e, --environment ENVIRONMENT The environment to run the Rack app on (default development)
-I, --include PATH Specify $LOAD_PATH directories
-p, --port PORT Define the TCP port to bind to
Use -b for more advanced options
--pidfile PATH Use PATH as a pidfile
--preload Preload the app. Cluster mode only
--prune-bundler Prune out the bundler env if possible
-q, --quiet Quiet down the output
-R, --restart-cmd CMD The puma command to run during a hot restart
Default: inferred
-S, --state PATH Where to store the state details
-t, --threads INT min:max threads to use (default 0:16)
--tcp-mode Run the app in raw TCP mode instead of HTTP mode
-V, --version Print the version information
-w, --workers COUNT Activate cluster mode: How many worker processes to create
--tag NAME Additional text to display in process listing
-h, --help Show help
The daemon has the default OpeNER daemon options. Being:
Usage: kaf2json-daemon <start|stop|restart> [options]
When calling kaf2json without <start|stop|restart> the daemon will start as a foreground process
Daemon options:
-i, --input QUEUE_NAME Input queue name
-o, --output QUEUE_NAME Output queue name
--batch-size COUNT Request x messages at once where x is between 1 and 10
--buffer-size COUNT Size of input and output buffer. Defaults to 4 * batch-size
--sleep-interval SECONDS The interval to sleep when the queue is empty (seconds)
-r, --readers COUNT number of reader threads
-w, --workers COUNT number of worker thread
-p, --writers COUNT number of writer / pusher threads
-l, --logfile, --log FILENAME Filename and path of logfile. Defaults to STDOUT
-P, --pidfile, --pid FILENAME Filename and path of pidfile. Defaults to /var/run/tokenizer.pid
--pidpath DIRNAME Directory where to put the PID file. Is Overwritten by --pid if that option is present
--debug Turn on debug log level
--relentless Be relentless, fail fast, fail hard, do not continue processing when encountering component errors
These daemons make use of Amazon SQS queues and other Amazon services. The access to these services and other environment variables can be configured using a .opener-daemons-env file in the home directory of the current user.
It is also possible to provide the environment variables directly to the deamon.
For example:
AWS_REGION='eu-west-1' kaf2json start [other options]
We advise to have the following environment variables available:
This project has received funding from the European Union’s Seventh
Framework Programme for research, technological development and
demonstration under grant agreement no 261712.