|
Text Data Mining |
| The links below this paragraph relate to the generic Robot for WWW exploration. For those seeking
the specific Government Grant Robot, GrantBot,
review the same material on this page for a similar overview of goals and operation. If this style of application
meets your interest, then simply download GrantBot.jar. Be aware that it is in a beta status, but even then offers a high degree of robustness and suitability. For those seeking a search Robot for businesses in the greater Seattle region (Puget Sound), BizBot, review this page for overview of goals and operation. If this style of application is suitable, then simply download BizBot.jar. As with GrantBot, this applicaton is in a beta status, but as these are both of similar generation design, it also exhibits a high degree of robustenss. Graphical presentation of personal Web Robots. |
|
Lexical Analysis This design work moves along several fronts. It contains a robust search engine that performs brute force pattern matching to simple user entered expressions, much like the format popularized by AltaVista. It also contains specialized bounding equations for a search, expressed in terms of linear word distance between textual material of interest that is far more flexible. The design also encompass a Rule Base such that any query may become a Rule and be named, stored, restored, combined with other named Rules and called within a query. |
|
Semantic Analysis Further design elaborations relate to the Natural Language Parsing issues that still inhibit a natural machine to person dialog. In the matter of automatic reading of textual material on the Net, the parser is far more prone to elaborate writing examples - this sentence being a deliberate example. I have three examples of difficult sentence constructions illustrated in a form you put behind you in grade school - sentence diagrams. It should come as no surprise that the difficulty of analysis is in determining what is called the essential components, especially if the example contains four phrases. My designs parse every one of these examples easily. |
|
Fuzzyfication This topic is outside of the usual canon. Basically it is the preparation of the text through the application of weight based meaning to words, phrases and terms (already identified in the Lexical Analysis) to produce Fuzzy Text. In the larger scope would be found meaning networks, a mapping of their attribute/weights to contexts. |
|
Fuzzy Analysis Fuzzy Text allows a different sort of pattern matching paradigm. Rather than matching to specific data or knowledge, it responds to general data and knowledge. This form of analysis has proven to be robust and flexible, an unique combination in an otherwise crystalline world of rigid logic. |
| download jar download read-me file |
|
Running Java Applications Introduction Operation Problems ? Create a shortcut to a DOS session in the same directory that contains the download jar file. Run that shortcut, or simply open a DOS session in that directory. Enter the following call at the DOS prompt: java -jar GrantBot.jar or java -jar BizBot.jar If you encounter error messages, these relate to the same error issues that prevented "Java Web Start" from successfully performing. In that case, consult the Java Runtime Environment installation instructions, or repeat its installation, or visit the same site offered above. |