Catálogo de publicaciones - libros

With the rapid growth of online information, a simple search query may return thousands or even millions of results. There is a need to help user to access and identify relevant information in a flexible way. This paper describes a methodology that automatically map web search results into user defined categories. This allows the user to focus on categories of their interest, thus helping them to find for relevant information in less time. Text classification algorithm is used to map search results into categories. This paper focuses on feature selection method and term weighting measure in order to train an optimum and simple category model from a relatively small number of training texts. Experimental evaluations on real world data collected from the web shows that our classification algorithm gives promising results and can potentially be used to classify search results returned by search engines.

One problem domain of meta search is to combine and improve the precision of ranking results from various search systems. This paper describes a rank aggregation model that incorporates text analysis measure with existing rank-based method, e.g. Best Rank and Borda Rank, to aggregate search results from various search systems. This approach provides means to normalize the differences of rank methodology used by different search systems, justifying the potential of using contents analysis to improve the results relevancy in meta search. In this paper, we fully describe our approach on text normalization for meta search and present our rationality of using two rank-based methods in our model. We then evaluate and benchmark the performance of our model based on user judgment on results relevancy. Our experiment results show that when text analysis factor is taken into account, the results outperform the rank-based methods alone. This shows the potential of our model to complement current rank aggregation methods used in meta search.

We argue that the emphasis normally placed on query-similarity in Web search limits search precision. We draw on related work in case-based reasoning (CBR) and recommender systems research, which shows how enhancing diversity can improve the quality of retrieved cases and recommendations. We investigate the use of related diversity-enhancing retrieval techniques in Web search, showing that similar benefits are available, i.e. that result diversity can be significantly enhanced without compromising query similarity or result precision and recall.

Constraint satisfaction problem (CSP) methodologies are intended to solve (mostly combinatorial) problems especially in areas of planning and scheduling. Our paper focuses on a model enforcing arc consistency without changing the structure of the constraint network (CN), i.e., only by efficiently removing values from variable domains. The propagation of a variable domain to the constraints related to this variable allows the model to keep under control the size of the search space, by enumerating only the variable values which are part of a solution.

Based on ISAR imaging technology and Computer Vision theory, a novel target recognition method extracting the 3-Dimensional characteristic of targets from ISAR images sequence is presented in this paper, which can get higher recognition rate with fewer samples.

In this paper we discuss uncertain reasoning and decision making. Our proposal is based on the knowledge we have and entirely formalized within the so called classical two valued logic, so it has a solid foundation. Basic notions of various items are defined formally; formulas of supporting degree for uncertain reasoning and supporting degree with safety for decision making are introduced. Evaluation of “weighted facts”, which represents the different importance of facts, is clearly presented within our proposal without anything else. The relation between uncertain reasoning and decision making is discussed in detail. Examples in the paper are comprehensively exhibited, which shows that our proposal is reasonable and computer-operative,

Model-based software debugging helps users to find program errors and thus to reduce the overall costs for software development. In this paper, we extend our previous work to diagnose common data structure errors. The proposed logical program model derives from a collection of indexed object relations, which capture the underlying data structures at the abstraction level of objects. A case study suggests that the counterexample with the diagnoses can help the user to understand the nature of program errors and thus speed up error correction.

Well logging analysis plays an essential role in petroleum exploration and exploitation. It is used to identify the pay zones of gas or oil in the reservoir formations. This paper applies intelligent technology for well logging analysis, particular combining data mining and expert system together, and proposes an intelligent system for well log analysis called IntWeL Analyzer in terms of data mining platform MSMiner and expert system tool OKPS. The architecture of IntWeL Analyzer and data mining algorithms, including Ripper algorithm and MOUCLAS algorithm are also presented. MOUCLAS is based on the concept of the fuzzy set membership function that gives the new approach a solid mathematical foundation and compact mathematical description of classifiers. The aim of the study is the use of intelligent technology to interpret the pay zones from well logging data for the purpose of reservoir characterization. This approach is better than conventional techniques for well logging interpretation that cannot discover the correct relation between the well logging data and the underlying property of interest.

In this paper we propose an extension, , of so that sets can be naturally constructed in logic programming. In , sets can be defined by statements so it has a strong capability in creating sets. Three deductive rules are also introduced in this paper, which make strong in deductions and programming even when sets are involved in deductions. The syntactical description and the semantical interpretation of are comprehensively discussed in detail. The soundness and completeness theorem of is proved, which provides a solid foundation of .

In this paper, we calculate the time intervals of page views, and analyze the time intervals to obtain a certain threshold, which is then used to break the web logs into sessions. Based on the time intervals, frequencies for each interval are counted and frequency vectors are obtained for each IP. Some IPs with special features of frequency distributions can be deemed as single users. For these IPs, we can define threshold for each individual IP, and separate sessions at the points of long access time intervals.