Multi-Document Summarization Service

Topic Name:

Depth:

Input data:


Service Documentation

Service Parameters

Parameter Name

Description

Data Type

Default Value

root

The name of the topic to which your multi-document summarization report will apply. If not specified or zero-length, the label "Untitled" is used.

String

"Untitled"

depth

The concept depth or density the multi-document summarizer should shoot for when constructing the summary. Use either 0 (more detail), 1 (typical detail), or 2 (less detail).

Integer

1

indata

A data structure comprising a concatenation of the individual instance data from which the multi-document summary will be created. Then multi-document summarizer returns a summary of the documents represented in this data structure. The Ultimate Research Assistant Information Extraction Service can be used to help construct this data structure. Please view some of the sample code on the API home page for more information. Each record in the expected data structure takes the following form (see example above):

Page URL or unique ID (Mandatory value, should be unique for each record)<FIELD>Text snippet (Mandatory, used to construct the actual text of the summary<FIELD>Space delimited list of concepts or tags (multi-word tags should use underscore rather than space). Mandatory value - these concepts or tags are used to select and order specific text snippets in the summary<RECORD>

String

Not applicable (required value)