《1. Introduction Natural Language Access to Data Bases Interpreting Update Requests 1》由会员分享，可在线阅读，更多相关《1. Introduction Natural Language Access to Data Bases Interpreting Update Requests 1（12页珍藏版）》请在金锄头文库上搜索。
1、Natural Language Access to Data Bases Interpreting Update Requests 1 James Davidson and S. Jerrold Kaplan 2 Computer Science Department Stanford University Stanford, CA 94305 For natural language data base systems to operate effectively in practical domains, they must have the capabilities required
2、by real applications. One such capability is understand- ing and performing update requests. The processing of natural language updates raises problems not encountered in the processing of queries. These difficulties stem from the fact that the user will naturally phrase requests with respect to his
3、 conception of the domain, which may be a considerable simplification of the actual underlying data base structure. Updates that are meaningful and unambiguous from the users standpoint may not translate into reasonable changes to the underlying data base. Update requests may be impossible (cannot b
4、e performed in any way), ambiguous (can be performed in several ways), or pathological (can be performed only in ways that cause undesirable side effects). Drawing on work in linguistics and philosophy of language, we have developed a domain- transparent approach to identifying and performing reason
5、able changes in response to a users update request, using only knowledge sources typically present in existing data base systems. A simple notion of user model and explanation with respect to the users state of knowledge are central to the design. This paper describes a prototype system PIQUE (Progr
6、am for Interpretation of QueryUpdate in English), which implements this approach. 1. Introduction Natural language is a desirable access mechanism for data base systems because it frees the user from the task of understanding the details of the data base structure. A number of systems have provided
7、natural language query capabilities (for example, Hendrix et al. 1978); however, few of these allow the user to perform updates (changes) to the data base using nat- ural language. (For an example of one that does allow simple updates, see Henisz-Dostert and Thompson 1974.) 1 This work is part of th
8、e Knowledge Base Management project (DARPA contract #N00039-82-C-0250), which is investi- gating the application of artificial intelligence techniques to issues in the field of data base management. The views and conclusions contained in this document are those of the authors and should not be inter
9、preted as representative of the official policies, either ex- pressed or implied, of DARPA or the U.S. Government. 2 Authors current address: Teknowledge Inc., 525 University Avenue, Palo Alto, CA 94301 The provision of update capabilities introduces prob- lems not seen in handling queries. These pr
10、oblems arise because the user is phrasing his requests with respect to his view of the data base, which may be a simplification or transformation of the actual data base structure. While a well-formed query expressed in terms of the users view of the data base will always result in the same answer,
11、regardless of how the query may be mapped into the actual data base structure for execution, this is not the case for an update expressed on a view. Since updates request modification of the content of the data base, different mappings of the update re- quest into the actual data base structure may
12、result in different effects. Some of these effects may be unde- sirable or unanticipated. Specifically, the user may make requests that are impossible (cannot be per- formed in any way, due to hidden constraints on the data base), ambiguous (can be performed in several ways), or pathological (can be
13、 performed only in ways that cause unanticipated side effects). While human Copyright 1983 by the Association for Computational Linguistics. Permission to copy without fee all or part of this material is granted provided that the copies are not made for direct commercial advantage and the Journal re
14、ference and this copyright notice are included on the first page. To copy otherwise, or to republish, requires a fee and/or specific permission. 0362-613X/83/020057-12503.00 American Journal of Computational Linguistics, Volume 9, Number 2, April-June 1983 57 James Davidson and S. Jerrold Kaplan Nat
15、ural Language Access to Data Bases: Interpreting Update Requests speakers would intuitively reject these unusual read- ings, a computer program may be unable to distinguish them from more appropriate ones. For example, a simple request to Change the teach- er of CS345 from Smith to Jones might be ca
16、rried out by altering the number of a course that Jones al- ready teaches to be CS345, by changing Smiths name to be Jones, or by modifying a teaches link in the data base. While all of these may literally carry out the update, they may implicitly cause unanticipated changes such as altering Joness salary to be Smiths. Our approach to this problem is to treat updates as requesting that the data base be put into a self- consistent state in which the request is satisfied; th
