SOAP API Reference
- Pascal Himmel (Unlicensed)
- HelpdeskJV
SOAP API Reference
JoinVision offers a SOAP 1.1 interface for accessing the semantic extraction services. This page describes the different SOAP methods and explains their parameters and output in detail.
Service Location
Address: https://cvlizer.joinvision.com/cvlizer/exservicesoap
WSDL: https://cvlizer.joinvision.com/cvlizer/exservicesoap?wsdlService Name: {http://servlets.iex.jv.com/}SemanticExtractionService
Port Name: {http://servlets.iex.jv.com/}SemanticExtractionPortMethod Descriptions
For each method you will need to provide your product token. How to find the product token is explained here.
Get Contingent
Get the current contingent (e.g. number of conversions) left for a specific product.
Signature
Integer getContingent(String username, String token)Request Parameters
Parameter | Type | Description |
|---|---|---|
username | String | An empty string |
token | String | Your product token |
Response
If your request is successful, the API returns an Integer that describes the remaining contingent.
Error Handling
Returns a SOAPException in case of authentication issues.
Get XML Domains
Get the most recent domain value lists as XML.
Signature
String getXMLDomainsFor(String language)Request Parameters
Parameter | Type | Description |
|---|---|---|
language | String | The preferred Target Language in which the long-values of the domains should be returned. |
Response
If your request is successful, the API returns a XML object with the most recent domain values:
<domains>
<domain name="ISOCountry">
<pair code="PR" value="Puerto Rico"/>
<pair code="PS" value="Palestina"/>
...
</domain>
...
</domains>Error Handling
Returns a SOAPException in case the provided language is invalid.
Extract To XML
This method returns a structured XML containing all extracted information for a provided input file. Based on the selected Extraction Model, the returned XML validates against the respective XML Schema for the applied extraction mode.
The Target Language indicates in which language the long-values for the extracted abstraction codes (like skills, operation areas, education level, …) are returned. It does NOT define target or source language of the whole document. Text translation is not performed.
Extracting from an URL
It is possible to pass an URL instead of a binary document. In this case, the parameter “filetype” is set to “url” and the parameter “document” has to contain the UTF-8 encoded URL itself
Signature
String extractToXML(String username, String token, String language, String model, byte[] inputdata, String inputdatatype)Request Parameters
Parameter | Type | Description |
|---|---|---|
username | String | An empty string |
token | String | Your product token |
language | String | Preferred Target Language |
model | String | |
inputdata | byte[] | Document to be semantically extracted and transformed. Depending on the programming language in use (for instance Ruby and JavaScript) the byte array needs to be converted to a String and base64 encoded before submitting. |
inputdatatype | String | Filetype of the document (document postfix), e.g. 'docx' or the whole filename, e.g. 'document.pdf'. |
If the filename entered in parameter “inputdatatype” does not contain a postfix, the extraction will fail
Response
If your request is successful, the API returns a XML object as String, following the respective XML schema based on the chosen Extraction Model. More information on the returned XML can be found in the JoinVision XML 3.0 Schema Description.
Error Handling
Returns a SOAPException for any kind of extraction problems, containing a detailed error description.
Extracting a LinkedIn/XING profile
To extract data from a LinkedIn or XING profile, “inputdatatype” has to be set to “json-linkedin” or “json-xing” respectively, and “inputdata” has to contain the base64 encoded .json of the profile.
Categorize
JoinVision offers a document classification web service (called “Classifier”) which accepts a set of documents as input, converts each page of the document to PDF, determines the category of each page of the document, extracts the document's plain text and returns a set of categorized documents. The categorization service supports packaged files, like MIME-and MSG-Format (E-mails) and ZIP, also one enclosed in another (like an E-mail with ZIP-attachment). In case of mails the mail body is classified and returned as well.
Signature
OutputDoc[] categorize(InputDoc[] inputDocs, String domain, String username, String token)Request Parameters
The required input consists of an array of InputDocs, whereas an inputDoc contains the name of the file (in order to determine the file type) and the file itself as byte array.
Parameter | Type | Description |
|---|---|---|
inputDocs | inputDoc[] | An array of documents in the form of InputDocs |
domain | String | For classifications in the HR-domain, the domain value is always “hr” |
username | String | An empty string |
token | String | Your product token |
Response
If your request is successful, the API returns an array of OutputDocuments, which contains the binary document as byte array (always “pdf”), the amount of pages, whether or not the document has been OCR converted, the document's original filename, the document's plaintext, and the detected document's category.
Error Handling
Returns a SOAPException for any kinds of errors
Merge To XML
This method returns a structured XML containing all extracted information for one or more provided input files of the simple type InputDoc (one of the files is required to be a CV). All provided files are classified, splitted if required, semantically extracted and merged into the CV-XML. Based on the selected extration mode, the returned XML validates against the respective XML Schema for the applied extraction mode. The target language String accords to ISO 639-1 (upper-case).
Signature
String mergeToXML(String username, String token, String language, String model, InputDoc[] inputDocs)Request Parameters
The required input consists of an array of InputDocs, whereas an inputDoc contains the name of the file (in order to determine the file type) and the file itself as byte array.
Parameter | Type | Description |
|---|---|---|
username | String | An empty string |
token | String | Your product token |
language | String | Preferred Target Language |
model | String | Merging currently only supports model “cvlizer_3_0” |
inputDocs | inputDoc[] | An array of documents in the form of InputDocs One of the files is required to be a CV |
Response
If your request is successful, the API returns a XML object as String, following the respective XML schema based on the chosen Extraction Model. More information on the returned XML can be found in the JoinVision XML 3.0 Schema Description.
Error Handling
Returns a SOAPException for any kind of extraction problems containing a detailed error description.