Automatically assessing structured data for decision making
A computerized system and methods are provided for the automated extraction of contextually relevant information, and the automatic processing of actionable information from generic document sets. More specifically, automated systems and techniques for the extraction and processing of opportunity documents, are provided, which avoid inaccuracies and inefficiencies resulting from conventional and/or human-based document processing techniques.
1. A computer-implemented method comprising, by a first computer system:
obtaining a first document, the first document comprising a combination of prose, semi-structured text, and standard clauses used in a plurality of similar other documents and defining at least one task to be performed by a user;
extracting structure information from the first document and, based upon the structure information, classifying the first document into a first document type;
extracting a first well known object from the first document, the first well known object comprising information previously extracted from one or more of the similar other documents;
determining a substantive effect of the first well known object on the task to be performed, wherein the substantive effect comprises an effect of a standard term in the first document on the task;
extracting a first requirement of the task from the first document; and
automatically presenting to a user information comprising: an indication of the structure information, the substantive effect, the first requirement, the contents of the document, or a combination thereof.
2. The method of claim 1 , further comprising:
receiving a query from a first user for information about the task;
automatically providing a response to the query, the response comprising information automatically extracted from the first document.
3. The method of claim 1 , wherein obtaining the first document comprises scraping information from a second computer system that is separate from the first computer system.
4. The method of claim 1 , further comprising:
executing a plurality of extractors to automatically obtain information from the first document, the plurality of extractors comprising a first extractor configured to extract the structure information and a second extract configured to extract the well known object.
5. A system comprising:
a computer-readable medium storing a first plurality of documents and data extracted from a second plurality of documents, the second plurality including at least one document not included in the first plurality; and
a processor configured to:
obtain a first document of the first plurality of documents, the first document comprising a combination of prose, semi-structured text, and standard clauses used in a plurality of similar other documents and defining at least one task to be performed by a user;
extract structure information from the first document and, based upon the structure information, classifying the first document into a first document type;
extract a first well known object from the first document, the first well known object comprising information previously extracted from one or more of the similar other documents;
determine a substantive effect of the first well known object on the task to be performed, wherein the substantive effect comprises an effect of a standard term in the first document on the task;
extract a first requirement of the task from the first document; and
automatically present to a user information comprising: an indication of the structure information, the substantive effect, the first requirement, the contents of the document, or a combination thereof.
6. The system of claim 5 , further configured to:
receive a query from a first user for information about the task;
automatically provide a response to the query, the response comprising information automatically extracted from the first document.
7. The system of claim 5 , wherein obtaining the first document comprises scraping information from a second computer system that is separate from the first computer system.
8. The system of claim 5 , further configured to:
execute a plurality of extractors to automatically obtain information from the first document, the plurality of extractors comprising a first extractor configured to extract the structure information and a second extract configured to extract the well known object.
9. A computer-implemented method comprising, by a first computer system:
obtaining a first opportunity document, the first opportunity document listing requirements for participating in an opportunity;
extracting structure information from the first opportunity document and, based upon the structure information, classifying the first opportunity document into a first opportunity document type;
extracting a first well known object from the first opportunity document, the first well known object comprising information previously extracted from multiple other opportunity documents; and
automatically presenting to a user a structured representation of opportunity information for the opportunity, the opportunity information comprising: a unique identifier for the opportunity, a title, a description, an indication of a source of the opportunity, and a due date.
10. The method of claim 9 , wherein the opportunity information further comprises one or more of:
an indication of a customer associated with the opportunity, a posting date for the opportunity document; a link or other reference to a file associated with the opportunity; and one or more options for sharing information about the opportunity.