We develop sophisticated text parsers for custom needs. They can find certain information in free text like addresses, prices, dates etc. Or, they can parse a file, structurize its content in certain manner and put into database. Our parsers use semantic analysis and self-learning algorithms that allow recognizing in unstructured data.
Below is a sample of one of our intelligent text parsers. This one goes through text looking for an address and if found, complete and normalize it. Being supplied with an appropriate database the parser can detect addresses of any country. Because of heuristic algorithm developed by us it can detect addresses that do not comply with any pattern like those written by a human.
This demo uses UK address database so it should be tried with text containing a real UK address. A test text can be gotten from some UK directory, e.g. real-estate web site. Because of sharing resources of the web hosting server, the parser may sometimes run a noticeable time. To reduce resource consumption the demo does not accept text longer than 1000 chars while the parser itself has no such restriction and can process much longer texts.
Some test examples (click any to fill in the box):
- The premise is located at eckington , coal aston, and can be easily found by the board.
- Superior holiday cottages in Chesterfield Road, Dronfield leading south from the suburb of Meersbrook to Woodseats.
- Desirable One, Two, Three Bed Apartments, Royal Artillery Quays, Erebus Drive in Range of £100,000 - £300,000
Paste a text that contains an address in UK and click the button:
