Đang chuẩn bị nút TẢI XUỐNG, xin hãy chờ
Tải xuống
Although ultimately intended functions include text generation (e.g., 4), present efforts focus on text analysis: developing the capability to take in essentially unconstrained business text and to output grammar and style critiques, on a sentence by sentence basis. Briefly, we use a large on-line dictionary and a bottom-up parser in connection with an Augmented Phrase Structure Grammar (5) to obtain an approximately correct structural description of the surface t e x t (e.g., we posit no transformations or recovery of deleted material to infer underlying "deep" structures). . | SOLUTIONS TO ISSUES DEPEND ON THE KNOWLEDGE REPRESENTATION Frederick B. Thompson California Institute of Technology Pasadena California In organizing this panel our Chairman Bob Moore expressed the view that too often discussion of natural language access to data bases has focused on what particular systems can or cannot do rather than on underlying issues. He then admirably proceeded to organize the panel around issues rather than systems. In responding I attempted to frame my remarks on each of his five issues in a general way that would not reflect My own parochial experience and interest. At one point I thought that I had succeeded quite well. However after taking a clearer eyed view it was apparent that my remarks reflected assumptions about knowledge represen tat 1 on that were by no means universal. This suggests a sixth issue which I would like to nominatei Are there really useful generalizations about computational linguistic Issues that are independent of assumptions concerning knowledge representation I will come back to this sixth issue after discussing the five chosen by our Chairman. Issue Aggregate.Functions and Quantity Quest Ions First let us cast this issue in a somewhat different way. In many data base situations there are classs of individuals all of whose Members share the vame attributes and thus from the point of view of the data base are Indistinguishable. Thus there is no need to add all of these individuals as separate entities. To use Bob Moore s example if a DEPARTMENT file has a field for NUMBER-OF-EMPLOYEES it stands to reason that the particular individuals who actually existed in the various departments would not be separately represented in the database for otherwise there would be a redundancy whose consistency would be hard to police . In such situations we need the notion of a collective namely a single data base object that takes the place of a number of individuals and which can carry their common attributes together with one .
