Output API

Essential API

The following are the essential elements of the API.

GetBufferedAIDRData

/channel/filter/{crisisCode}

Method GetBufferedAIDRData.getBufferedAIDRDataPostFilter() - load latest tweets

---

Additional API (possibly obsolete)

The following are additional elements of the API and/or obsolete documentation prior to 2015-04-28. These should be integrated into the documentation above or moved to the Javadoc-based documentation.

Base URI: http://localhost:port/aidr-output/rest

Get a JSONP list of labeled tweets from a specific channel

GET .../rest/crisis/fetch/channel/{crisisCode}?callback={callbackName}&count={count}

• crisisCode [mandatory]: Redis channel identifier to which to subscribe
• callback [optional]: name of the callback function for JSONP data
• count [optional]: the specified number of messages that have been buffered by the service. If unspecified or <= 0 or larger than the MAX_MESSAGES_COUNT field, the default number of messages are returned.

Return a list of active channels

GET .../rest/crisis/fetch/channels/list

Returns an HTML page listing the currently active channels that are being monitored by aidr-output

Get the latest labeled tweet across all channels as a JSONP object

GET .../rest/crisis/fetch/channels/latest?confidence={val}

When query parameter confidence is specified, then only a tweet which has at least one nominal label with confidence > val is returned, otherwise only a tweet which has at least one nominal label with confidence > 0.7 is returned (default behavior).

Stream jsonp data from Redis for a specific channel

GET ...rest/crisis/stream/filter/channel/{crisisCode}?callback={callback}&rate={rate}&duration={duration}

• crisisCode [mandatory]: Redis channel identifier to which to subscribe
• callback [optional]: name of the callback function for JSONP data
• rate [optional]: an upper bound on the rate at which to send messages to client, expressed as messages/min (a floating point number). If <= 0, then default rate is assumed.
• duration [optional]: time for which to subscribe (connection automatically closed after that). The allowed suffixes are: s (for seconds), m (for minutes), h (for hours) and d (for days). If nothing is specified, then by default, duration is disabled.

Test whether aidr-output is running

GET .../rest/manage/ping

Provides a method to test whether:

• Connection to Redis is available
• aidr-output services are running as expected

Get a JSONP list of labeled tweets from a specific channel, filtered by selection criteria

POST .../rest/crisis/fetch/channel/filter/{crisisCode}?callback={callbackName}&count={count}

• crisisCode [mandatory]: Redis channel identifier to which to subscribe
• callback [optional]: name of the callback function for JSONP data
• count [optional]: the specified number of messages that have been buffered by the service. If unspecified or <= 0 or larger than the MAX_MESSAGES_COUNT field, the default number of messages are returned.

POST request payload example: Request Headers: Content-Type: application/json Accept: application/json

Note: If no filtering required, then the payload will be as follows:

{ "constraints": [] }

Otherwise, if filtering required, then:

{ "constraints": [ { "queryType": "date_query", "comparator": "is_before", "timestamp": 1575339860 }, { "queryType": "date_query", "comparator": "is_after", "timestamp": 1272339860 }, { "queryType": "classifier_query", "classifier_code": "informative_pray_personal", "label_code": "praying", "comparator": "is", "min_confidence": 0.8 }, { "queryType": "classifier_query", "classifier_code": "informative_pray_personal", "label_code": "030_info", "comparator": "is_not" }, { "queryType": "classifier_query", "classifier_code": "informative_pray_personal", "label_code": null, "comparator": "has_confidence", "min_confidence": 0.5 } ] }

Note: Only those documents that satisfy ALL the constraints are returned.

Parameter details:

• queryType: indicates type of query. Currently can take only two values - "date_query" and "classifier_query".

• comparator: query predicate evaluation criterion. For date_query it can be either is_after or is_before depending on whether to filter documents that occurred after or before the specified timestamp value, respectively. For classifier_query it can be either is or is_not or has_confidence.

• timestamp: unix time specified in date_query as Java type long.

• classifier_code: corresponds to the attribute_code.

• label_code: corresponds to a valid label_code for a given classifier_code.

• min_confidence: Include only those documents for which the confidence of the specified classifier_code is greater than min_confidence. Is Java float type.