|
UIMA-compatible RCO annotators extract semantic and factographic data from texts.
RCO annotators extend IBM OmniFind ability to process unstructured texts and enable semantic search.
IBM OmniFind corporate search system involves UIMA (Unstructured Information Management Architecture) to link up additional text processing units (so-called "annotators"). Semantic data extracted by annotator within Omnifind framework may be used both for semantic search by attributes, so as for further analysis by external BI-systems.
RCO for IBM Omnifind includes the following annotators:
| Person annotator |
Extracts personal information from text.
Normalizes wordforms and abbreviations. Extracts and links additional attributes and facts with the individual mentioned:
- Date of birth;
- Place of birth;
- Citizenship;
- Addresses;
- Individual numbers;
- Passport details;
- Relatives;
- Education;
- Occupation;
- etc.
|
| Org annotator |
Extracts information on corporate body from text.
Normalizes wordforms and abbreviations. Extracts and links additional attributes and facts with the mentioned organization:
- Addresses;
- Bank details;
- Registration details(individual codes and numbers);
- Founders;
- Executives;
- Trade marks;
- Representative offices;
- etc.
|
| Geo annotator |
Extracts geographical and address information components. |
| Entity annotator |
Extracts and normalizes special entities:
- Numerals;
- Dates;
- Artefacts;
- Phone numbers;
- Vehicle numbers and models;
- etc.
|
| Themes annotator |
Based on lexical profiles describing specified domain, returns the list of themes related to the document. |
| Business facts annotator |
Extracts business related events and facts with definition of participants:
- Agreements;
- Shares;
- Merges and acquisitions;
- Ownership of property;
- Production output;
- Rendering of serbices;
- Partnership;
- IPO.
|
| Social facts annotator |
Business facts annotator analogue for social and political thematics, including:
- Elections;
- Meetings;
- Appeals;
- Travellings;
- Conflicts;
- Investigations and litigations;
- Events.
|
RCO annotators may be connected as extentions by means of standard IBM Omnifind interface.
|