Theia SDK Search and Data Extraction
Project description
qtcurate-sdk-python
Python SKD for Search and Data extraction
Theia documentation; Overview and Definitions
Theia is a managed solution to mine and retrieve Entities* or Typed Entities** out of unstructured data. In the following we cover details of configuring and submitting data mining jobs via Python SDK
*Entity : An entity is a word or phrase and can be associated with a meaning. For example, "AWS" is an entity and it can be associated with "Amazon Inc."
**Typed Entity : An entity that is meaningful only if it co-occurs with a number, date or another entity. For example, "Revenue" is considered a Typed Entity only if it is found in a context where the actual dollar value is reported:
Amazon Inc reported $68.8 billion in revenue
Table of content
- Installation
- Authentication
- Data Dictionaries
- Create a New Dictionary
- Upload a TSV Dictionary File
- Update an Existing Dictionary
- Delete a Dictionary
- List Available Dictionaries
- Fetch dictionary
- Extraction and Mining
- Mining Content Files
- Mining Web URLs
- Mining Data Streams
- Extracting Typed Entities
- Status Monitoring
- Searching in the Results
- Exporting the Results
- Exporting in Excel Format
- Exporting in JSON
- Clear temporary data
Installation
You can install it with pip. You have to have Python 3.5 or above
pip install qtcurate
Authentication
Valid API key is required for all operations.
Request
If API key is valid, than the response will be HTTP 200
containing user profile data.
If API key is missing or not valid, the endpoint will return HTTP 401
:
Response
{
"error": "unauthorized",
"error_description": "Full authentication is required to access this resource"
}
"Net Revenue" is a Typed Entity:
"Apple reported net revenue of $53.3 billion in 2018."
"Net Revenue" is not a Typed Entity:
"Tim Cook will go over net revenue of Apple in the upcoming conference call."
Entity Dictionaries
Entity dictionaries are lists of entities that can be used in a data mining task. Each item of a entity dictionary has a key
and a value
. The key
is the associated or normalized phrase and the value
is the actual word or phrase that represents the entity.
Theia scans all input utterances for every word or phrase in the dictionary and if found, map it to the associated value:
Dictionary (one item):
M&A => merger and acquisition
Input utterance:
Merger and acquisitions report published in 2019.
The above will be labeled with M&A
Theia uses various strategies for matching of dictionary values allowing users to configure the fuzziness of search. User can also provide a list of synonyms and stop phrases for the value matching. For example, you can only have "Apple Inc" as one entity in your dictionary and provide "inc", "corp", "corporation" and "company" as synonyms, allowing you to find all occurrences of "Apple the Company", "Apple Corporations" and "Apple Corp" in the content.
Entity dictionaries can be created in two ways:
- By providing dictionary entries in the request payload
- By uploading a TSV file containing dictionary entries in the format of
key<TAB>value
.
Data Dictionaries
Authentication with API KEY
dic = QtDict(api_key='1234567')
Create A New QtDict
Create data dictionaries with create() function First we prepare data Name is required.
dic.name("Name of QtDict)
Two ways to create entries
key = "some key"
value = "some value"
dic.add_entry(key, value)
or
tmp_dictionary = {"key": "some key", "value": "some value"}
dic.entries(tmp_dictionary)
where method entries take a Python dictionary with 2 elements where KEYS have string value "key" and "value" After that we can create our QtDict
dic.create()
Response
{
"id": "58608b1f-a0ff-45d0-b12a-2fb93af1a9ad",
"key": "user-example-com/58608b1f-a0ff-45d0-b12a-2fb93af1a9ad.csv.gz",
"name": "Name of dictionary",
"global": false,
"entries": [
{
"key": "some key",
"value": "some value"
}
]
}
Upload a TSV Dictionary File
Upload TSV data dictionaries with upload function:
dic.upload("file-path.tsv", "Name of dictionary")
TSV is required file format. Response is the same as create function response.
Update an Existing QtDict
To update an existing QtDict:
dic.update("58608b1f-a0ff-45d0-b12a-2fb93af1a9ad")
We use ID as argument of function.
Delete a QtDict
To delete an existing QtDict:
dic.delete("58608b1f-a0ff-45d0-b12a-2fb93af1a9ad")
List Available Dictionaries:
To list all of your existing dictionaries:
dic.list()
Response is a list of dictionaries
[
{
"id": "58608b1f-a0ff-45d0-b12a-2fb93af1a9ad",
"key": "user-example-com/58608b1f-a0ff-45d0-b12a-2fb93af1a9ad.csv.gz",
"name": "My dictionary",
"global": false,
"entries": []
}
{
...
}
]
Fetch QtDict:
To fetch data from the QtDict
dic.fetch("58608b1f-a0ff-45d0-b12a-2fb93af1a9ad")
Response
{
"id": "58608b1f-a0ff-45d0-b12a-2fb93af1a9ad",
"key": "user-example-com/58608b1f-a0ff-45d0-b12a-2fb93af1a9ad.csv.gz",
"name": "My dictionary",
"global": false,
"entries": []
}
Extraction and Mining
Data extraction and mining is the process of identifying entities found in the unstructured content using entity dictionaries and formatting them it into structured format. Unstructured data can be streamed from content files, data APIs or directly from public URLs.
Authenticate with API_KEY
tag = DataProcess(api_key="1234")
Mining Content Files
First, upload all content files for DataProcess:
tag.upload("file.pdf")
PDF, TXT and HTML formats are supported.
Response
{
"uuid": "c351283c-330c-418b-8fb7-44cf3c7a09d5",
"fileName": "file.pdf",
"link": "http://portal.document.quantxt.amazonaws.com/user@example.com/c351283c-330c-418b-8fb7-44cf3c7a09d5",
"date": "2019-10-25T20:14:41.925+02:00",
"contentType": "application/pdf",
"source": "file.pdf"
}
Now you can mine data via dictionaries. First you have to prepare additional options.
title
is optional but it is highly recommended for easier distinction between different tagging jobs.
vocabValueType
(Optional): can have one of the following values: NONE
, STRING
, DOUBLE
, DATETIME
. If set, the engine will extract only entities that are associated with an entity of this type.
vocabPath
(Required): The path to entity QtDict. Path is returned either after creation of a new QtDict or via listing of existing dictionaries.There is no limit on the number of files and dictionaries that can be processed via tagging_files function
tag.files("c351283c-330c-418b-8fb7-44cf3c7a09d5")
tag.title("My data mining with files and dictionaries")
tag.search_rule("user-example-com/58608b1f-a0ff-45d0-b12a-2fb93af1a9ad.csv.gz", DictionaryType.NONE)
tag.tagging_files()
Response
{
"index": "puvqrjfhqq",
"title": "My data mining with files and dictionaries",
"get_phrases": true,
"maxTokenPerUtt": 35,
"minTokenPerUtt": 6,
"excludeUttWithoutEntities": false,
"stitle": null,
"files": ["c351283c-330c-418b-8fb7-44cf3c7a09d5"],
"searchDictionaries": [
{
"vocabPath": "user-example-com/58608b1f-a0ff-45d0-b12a-2fb93af1a9ad.csv.gz",
"vocabValueType": "NONE"
}
]
}
index
represents the unique identification for the container that holds output labeled data.
You can set index, autotag, maxTokenPerUtt, minTokenPerUtt, excludeUttWithoutEntities with following methods:
tag.index("some index")
tag.autotag(True)
tag.exclude_utt_without_entities(True)
tag.max_token_per_utt(101)
tag.min_token_per_utt(101)
tag.stitle("some stitle string")
Request parameters:
chunk
(Optional, string) can be SENTENCE
or PARAGRAPH
or NONE
. This will result in splitting of data into semantic chunks before processing. For example, this allows user to split the content of an article in semantic sentences and apply entity dictionaries at sentence level.
excludeUttWithoutEntities
(Optional, boolean) if true
the output will only include chunks that have at least one label from the input dictionaries.
get_phrases
(Optional, boolean) if true
auto tagging will be performed.
minTokenPerUtt
(Optional, int)
maxTokenPerUtt
(Optional, int)
stitle
(Optional, string) Override command.
To delete a data container:
tag.delete("puvqrjfhqq")
Mining Web URLs:
Mining can be performed on a list of URLs. All parameters used in tagging files are applicable here.
tag.mining_url()
Theia can process both static and dynamic web pages. However, a number of websites use mechanisms to block webpage scrapping. Theia built-in Web parser is not designed to bypass such blocking mechanisms
Mining Data Streams
Tagging data from data streams such as third party APIs is supported. Please contact support@quantxt.com for details.
Extracting Typed Entities
Entity dictionaries allow the user to quickly search and label thousands of phrases in unstructured content. There are cases when users want to label a keyword or phrase as an entity only if it is associated with a value. For example:
A "release => "Manufactured" dictionary item will label both of the following utterances:
The first automobile in the US released by Ford. The first automobile in the US released in 1908 by Ford.
However, if the user is looking for only release year of the car makers, the first utterance won't have much use for him. He can only label the second utterance using a Typed Entities.
Supported types for associated entities are:
REGEX
, DOUBLE
, DATETIME
types are passed via vocabValueType
parameter.
If a type is set, a dictionary item will be extracted only if a type is found in its close proximity.
In the above example, user can set vocabValueType in the request to DOUBLE
to identify 1908 only if it is associated with the entity released
Status Monitoring
The progress endpoint allows user to check the progress of a submitted data mining job:
Request
tag.progress()
The search result is an array of active data mining jobs:
[
{
"index": "cjaejhvtao",
"progress": 36,
"progress_msg": "Collecting data..."
}
]
index
Unique ID of the running job
progress
Progress in % (a number between 0 to 100).
progress_msg
(Optional) Progress message.
It is also possible to check the progress of a specific data mining job:
Request
tag.progress("cjaejhvtao")
Response
{
"index": "cjaejhvtao",
"progress": 36,
"progress_msg": "Collecting data..."
}
Searching in the Results
The search endpoint allows user to run full-text and faceted search in the extracted data.
Request
tag.search("puvqrjfhqq", param_from, size, f1, f2)
Request parameters:
q
(Optional, string) Search query that filters the main content title
field. It supports boolean OR
, AND
and NOT
parameters.
f
(Optional, string) Query filters must be used in pairs. Filters are created for each input dictionary. For example, to include results that have one or more label from Vehicle
dictionary the request should look like: &f=Vehicle&f=*
. To include results that are labeled with Ford
or BMW
from the Vehicle
dictionary, the request would be &f=Vehicle&f=BMW&f=Vehicle&f=Ford
from
(Optional, int) Offset for paging results. Defaults to 0. Each page contains 20 items.
size
(Optional, int) Number of results to return. Maximum is 200.
Response
{
"Total": 2610,
"results": [
{
"title": "The Federal Reserve Bank of New York provides gold custody to several central banks, governments and official international organizations on behalf of the Federal Reserve System.",
"id": "Wv8fBG4Bc3WI8L9MbaO2",
"link": "https://www.hamilton.edu/news/story/hamilton-nyc-program-tours-federal-reserve-museum",
"score": 0.10268514747565982,
"source": "abc15.com",
"date": "2018-05-24T00:00:00.000Z",
"tags": [
"Federal Reserve",
"New York"
]
}
],
"aggs": {
"Tag": [{
"key": "Central bank",
"count": 878
}, {
"key": "Gold",
"count": 523
}
}
}
Total
: Number of results.
result []
: The array that contains result items.
aggs
: Facets over the results with count of items for each facet.
Exporting the Results
Results can be exported in XLSX or JSON format:
Exporting in Excel Format
tag.report_to_xlsx("puvqrjfhqq", "name-of-output-file.xlsx")
Exporting in JSON
tag.report_to_json("puvqrjfhqq", "name-of-output-file.json")
Clear temporary data
Both classes QtDict and DataProcess have a method clear() to delete all used variable.
dic.clear()
tag.clear()
The export output is limited to 5000. All /search
parameters can be passed here to export the desired slice of the data.
For technical questions please contact support@quantxt.com
Project details
Release history Release notifications | RSS feed
Download files
Download the file for your platform. If you're not sure which to choose, learn more about installing packages.