Memorized Question Answering System

This is a System that used memorized queries to simulate a question answering system.

Configuration

Before you can start the system you have to configure the following properties in the application.properties file.

Tripestore

qado.resultset.maxrows=10000

• qado.resultset.maxrows: The maximum number of rows that the system will return. If the database can only handle less than 10.000 rows you should change this value.

Stardog

stardog.url=
stardog.database=
stardog.username=
stardog.password=

• stardog.url: The url of the triplestore. For example: http://localhost:7200

• stardog.database: The name of the database in the triplestore. For example: qado

• stardog.username: The username of the triplestore. For example: admin

• stardog.password: The password of the triplestore. For example: admin

Virtuoso

virtuoso.url=
virtuoso.graph=
virtuoso.username=
virtuoso.password=

• virtuoso.url: The url of the triplestore. For example: http://localhost:7200

• virtuoso.graph: The name of the graph in the triplestore. For example: urn:bulkload:general

• virtuoso.username: The username of the triplestore. For example: admin

• virtuoso.password: The password of the triplestore. For example: admin

more options

You can change the following properties in the application.properties file if necessary.

qado.question.knowledgebase=wikidata
qado.question.user=open

• qado.question.knowledgebase: The knowledgebase that is used to answer the questions. For example: wikidata

• qado.question.user: The user that is used to answer the questions. For example: open

How to ask a question

You can ask a question by sending a POST request to the following url: http://demos.swe.htwk-leipzig.de:40165/mem-qa-system. The body of the request should contain a JSON object with the following properties:

• question: (required) The question that you want to ask. It musst be a question in the QADO dataset.

• language: (required) The language of the question. It musst be a language in the QADO dataset

• number_of_results_items: The number of results that you want to receive. If this field is 0 or not in the request the system will return all results.

• dataset: The dataset that you want to use. It musst be a dataset in the QADO database. If this field is not in the request the system will use datasets.

BE CAREFUL: If you send a request without number_of_results_items and dataset (Minimum Example) the system will respond with all possible results. This can take a very long time.

Full example

{
    "question": "What is the revenue of IBM?",
    "language": "en",
    "number_of_results_items": 5,
    "dataset": "QALD-9-plus-test-wikidata"
}

Minimum Example

{
    "question": "What is the revenue of IBM?",
    "language": "en"
}

Example without dataset

{
    "question": "What is the revenue of IBM?",
    "language": "en",
    "number_of_results_items": 5
}

Example without number of results items

{
    "question": "What is the revenue of IBM?",
    "language": "en",
    "dataset": "QALD-9-plus-test-wikidata"
}

Swagger UI

The system has a Swagger UI. You can use it to test the system and to see the documentation of the endpoints. The Swagger UI is available at the following url: http://demos.swe.htwk-leipzig.de:40165/swagger-ui/index.html

More information

Swagger UI is available at the following url: http://demos.swe.htwk-leipzig.de:40165/swagger-ui.html