Caroline M. Eastman

Hyper-semantic data modeling

Abstract This paper describes a new area of data modeling, a model in this new area, and the schema specification language for the model. The Knowledge/Data Model captures both knowledge semantics, as specified in Knowledge Based Systems, and data semantics, as represented by Semantic Data Models. The Knowledge/Data Model is an instance of a new class of models, called hyper-semantic data models, which facilitate the incorporation of knowledge in the form of heuristics, uncertainty, constraints and other Artificial Intelligence concepts, together with object-oriented concepts found in Semantic Data Models. The unified knowledge/data modeling features are provided via the constructs of the Knowledge/Data Language.

DESIGNING USER INTERFACES FOR OLDER ADULTS

As the computer industry puts more emphasis on software interface design, the average user has found todays applications more powerful and easier to use. However, many older adults still find the technology awkward and confusing. The physical and behavioral characteristics of seniors that prevent them from using popular applications are examined to identify possible software solutions. The results of a survey of older adults also are presented. This survey was conducted to help determine the effect these characteristics have on older adults’ computer skills. Finally, several software solutions are proposed, and a prototype implementation of these solutions is given.

30,000 hits may be better than 300: precision anomalies in internet searches

Results of searches using a variety of query formulations with several Internet search engines show that strategies intended to give narrower and more precise results may not give improvements in precision even though they result in fewer hits. Searches were performed by students in graduate information retrieval courses using different formulations for the same topic.

Tree structures for high dimensionality nearest neighbor searching

Abstract A nearest neighbor searching algorithm which is an extension of the multidimensional binary tree ( k-d tree) for high dimensional spaces is discussed. A model of its behavior, which is applicable under restricted conditions, shows that the search time required is bounded by 0( log 2 N ) α , where N is the number of records and α is a system-dependent parameter. Experiments with a document collection show that the model provides a reasonable guide to performance, and that some savings over a sequential search can be achieved in this type of application. A probabilistic version of the algorithm is presented which provides significantly faster searching with little degradation in retrieval quality.

The inference problem: Maintaining maximal availability in the presence of database updates

In this paper, we present the Dynamic Disclosure Monitor (D^2Mon) architecture to prevent illegal inferences via database constraints. D^2Mon extends the functionality of Disclosure Monitor (DiMon) to address database updates while preserving the soundness and completeness properties of the inference algorithms. We study updates from the perspective of increasing data availability. That is, updates on tuples that were previously released may affect the correctness of the user inferences over these tuples. We develop a mechanism, called Update Consolidator (UpCon), that propagates updates to a history file to ensure that no query is rejected based on inferences derived from outdated data. The history file is used by the Disclosure Inference Engine (DiIE) to compute inferences. We show that UpCon and DiIE working together guarantee confidentiality (completeness property of the data-dependent disclosure inference algorithm) and maximal availability (soundness property of the data-dependent disclosure inference algorithm) even in the presence of updates. We also present our implementation of D^2Mon and our empirical results.

Dynamic disclosure monitor ( D 2 Mon ): an improved query processing solution

The Dynamic Disclosure Monitor (D2Mon) is a security mechanism that executes during query processing time to prevent sensitive data from being inferred. A limitation of D2Mon is that it unnecessarily examines the entire history database in computing inferences. In this paper, we present a process that can be used to reduce the number of tuples that must be examined in computing inferences during query processing time. In particular, we show how a priori knowledge of a database dependency can be used to reduce the search space of a relation when applying database dependencies. Using the database dependencies, we develop a process that forms an index table into the database that identifies those tuples that can be used in satisfying database dependencies. We show how this process can be used to extend D2Mon to reduce the number of tuples that must be examined in the history database when computing inferences. We further show that inferences that are computed by D2Mon using our extension are sound and complete.

The effects of search engines and query operators on top ranked results

We examine whether the use of query operators changes the documents retrieved by three popular Web search engines. One hundred queries containing query operators were selected from the transaction log of a major Web search service. The query operators were then removed from these one hundred advanced queries. Both the original and modified queries were submitted to three major Web search engines. A total of 600 queries were submitted, and 5748 of the documents retrieved were examined. Changes in the ranking of the top documents retrieved were examined. The significant results of our research are that the effectiveness of query operators is dependent on the specific search engine utilized, and that generally there is approximately 60% similarity between retrieved results across all search engines. Implications on the effectiveness of current searching techniques, for future search engine design, and of future research are discussed.

A tree algorithm for nearest neighbor searching in document retrieval systems

The problem of finding nearest neighbors to a query in a document collection is a special case of associative retrieval, in which searches are performed using more than one key. A nearest neighbors associative retrieval algorithm, suitable for document retrieval using similarity matching, is described. The basic structure used is a binary tree, at each node a set of keys (concepts) is tested to select the most promising branch. Backtracking to initially rejected branches is allowed and often necessary. Under certain conditions, the search time required by this algorithm is 0(log 2 N) k . N is the number of documents, and k is a system-dependent parameter. A series of experiments with a small collection confirm the predictions made using the analytic model; k is approximately 4 in this situation. This algorithm is compared with two other searching algorithms; sequential search and clustered search. For large collections, the average search time for this algorithm is less than that for a sequential search and greater than that for a clustered search. However, the clustered search, unlike the sequential search and this algorithm, does not guarantee that the near neighbors found are actually the nearest neighbors.

The ambiguity of negation in natural language queries to information retrieval systems

A prototype system to handle negation in natural language queries to information retrieval systems is presented. Whenever a query that has negation is entered, the system will determine whether or not it is necessary for the user to clarify exactly what constituents in the query are being negated. If clarification is needed, the user is presented with a list of choices and asked to select the appropriate one. The algorithm used is based on the results of a survey administered to 64 subjects. The subjects were given a number of queries using negation. For each query, several possible choices for the negated constituent(s) were given. Whenever a lexical unit composed of nouns connected by the conjunction and was negated, there was general agreement on the response. But whenever there were multiple lexical units involved, such as complex lexical units connected by and or prepositional phrases, the subjects were divided on the choices. The results of this survey indicate that it is not possible for a system to automatically disambiguate all uses of negation. However, it is possible for the user interface to handle disambiguation through a clarification dialog during which a user is asked to select from a list of possible interpretations.

Automatic evaluation of credibility on the Web

A system to automatically assess credibility of web sites is described. The system uses information about credentials, advertisements, and web design to produce a credibility measure. The system was tested on several web sites in the medical domain. It selected the same three domains as most credible as did two manual rankings from the Wall Street Journal and a Stanford University research study.