Sui Yuefei

Progress in the development of national knowledge infrastructure

This paper presents the recent process in a long-term research project, called National Knowledge Infrastructure (or NKI). Initiated in the early 2000, the project aims to develop a multi-domain shareable knowledge base for knowledge-intensive applications. To develop NKI, we have used domain-specific ontologies as a solid basis, and have built more than 600 ontologies. Using these ontologies and our knowledge acquisition methods, we have extracted about 1.1 millions of domain assertions. For users to access our NKI knowledge, we have developed a uniform multi-modal human-knowledge interface. We have also implemented a knowledge application programming interface for various applications to share the NKI knowledge.This paper presents the recent process in a long-term research project, called National Knowledge Infrastructure (or NKI). Initiated in the early 2000, the project aims to develop a multi-domain shareable knowledge base for knowledge-intensive applications. To develop NKI, we have used domain-specific ontologies as a solid basis, and have built more than 600 ontologies. Using these ontologies and our knowledge acquisition methods, we have extracted about 1.1 millions of domain assertions. For users to access our NKI knowledge, we have developed a uniform multi-modal human-knowledge interface. We have also implemented a knowledge application programming interface for various applications to share the NKI knowledge.

The Description Logic for Relational Databases

Description logics are widely used to express structured data and provide reasoning facility to query and integrate data from different databases. This paper presents a many-sorted description logic \({\cal MDL}\) to represent relational databases. We give a translation from relational databases to the description logic \({\cal MDL}\), and show this translation completely and faithfully captures the information in the relational database. Moreover, we show that some relational algebra operations could be expressed in \({\cal MDL}\).Description logics are widely used to express structured data and provide reasoning facility to query and integrate data from different databases. This paper presents a many-sorted description logic MDL to represent relational databases. We give a translation from relational databases to the description logic MDL, and show this translation completely and faithfully captures the information in the relational database. Moreover, we show that some relational algebra operations could be expressed in MDL.

Two Online Algorithms for the Ambulance Systems

An ambulance system consists of a collectionS={s1,...sm} of emergency centers in a metric spaceM. Each emergency centersi has a positive integral capacityci to denote, for example, the number of ambulances at the center. There aren = Σi = 1mci patients requiring ambulances at different timestj and every patient is associated with a numberbj, the longest time during which the patient can wait for ambulance. An online algorithmA will decide which emergency center sends an ambulance to serve a request for ambulance from a patient at some time. If algorithmA sends an ambulance insi to serve a patientrj, then it must be observed thatdi,j/v<bj, wheredi,j is the distance between emergency centersi and patientrj, andv is the velocity of ambulance. A fault of algorithmA is such that to a request for ambulance at some timetj withj≤n, for everyi withdi,j/v<bj, there is no ambulance available insi. The cost of an algorithmA is the number of faultsA makes. This paper gives two algorithmsB andC, whereB is the local greedy algorithm andC is a variant of balancing costs, and proves that bothB andC have no bounded competitive ratios. Moreover, given any sequence σ of requests for ambulances without optimal faults, the cost ofC on σ is less than or equal to [n/3] and that ofB is less than or equal to [n/2].

The Variant Gentzen System for the Propositional Logic

The Gentzen system for the propositional logic is a deduction system which is logically equivalent to the corresponding axiomatic system. In the Gentzen system, the validity and provability of a sequence Γ ⇒ Δ are considered. The validity(⊨) corresponds to the provability(⊢). In this paper, we propose the variant Gentzen system which is a dual-system of the Gentzen system. A co-sequence Γ|Δ is introduced, and the variant Gentzen system is defined such that the satisfiability(⊨^co) of the co-sequence corresponds to the provability(⊢^co) of it. And another dual-Gentzen system is given such that the unsatisfiable of the co-sequence corresponds to the provability. The soundness and the completeness theorems of the both system are proved.

The B4-valued propositional Belnap-Dunn logic1

A B4-valued propositional Belnap-Dunn logic will be proposed in this paper which there are two unary connectives [ ], [⊥] to represent the two middle values , ⊥, respectively, and a Gentzen-typed deduction system will be given so that the system is sound and complete with the four-valued semantics of the propositional Belnap-Dunn logic.

A General Approach to Extracting Full Names and Abbreviations for Chinese Entities from the Web

Identifying Full names/abbreviations for entities is a challenging problem in many applications, e.g. question answering and information retrieval. In this paper, we propose a general extraction method of extracting full names/abbreviations from Chinese Web corpora. For a given entity, we construct forward and backward query items and commit them to a search engine (e.g. Google), and utilize search results to extract full names and abbreviations for the entity. To verify the results, filtering and marking methods are used to sort all the results. Experiments show that our method achieves precision of 84.7% for abbreviations, and 77.0% for full names.

Well limit behaviors of term rewriting systems

The limit behaviors of computations have not been fully explored. It is necessary to consider such limit behaviors when we consider the properties of infinite objects in computer science, such as infinite logic programs, the symbolic solutions of infinite polynomial equations. Usually, we can use finite objects to approximate infinite objects, and we should know what kinds of infinite objects are approximable and how to approximate them effectively. A sequence {Rk: k ε ω} of term rewriting systems has the well limit behavior if under the condition that the sequence has the set-theoretic limit or the distance-based limit, the sequence {Th(Rk): k ε ω} of corresponding theoretic closures of Rk has the set-theoretic or distance-based limit, and limk→∞Th(Rk) is equal to the theoretic closure of the limit of {Rk: k ε ω}. Two kinds of limits of term rewriting systems are considered: one is based on the set-theoretic limit, the other is on the distance-based limit. It is proved that given a sequence {Rk: κ ε ω} of term rewriting systems Rk, if there is a well-founded ordering ≺ on terms such that every Rk is ≺-well-founded, and the set-theoretic limit of {Rk: κ ε ω} exists, then {Rk: κ ε ω} has the well limit behavior; and if (1) there is a well-founded ordering ≺ on terms such that every Rk is ≺-well-founded, (2) there is a distance d on terms which is closed under substitutions and contexts and (3) {Rk: k ε ω} is Cauchy under d then {Rκ: κ ε ω} has the well limit behavior. The results are used to approximate the least Herbrand models of infinite Horn logic programs and real Horn logic programs, and the solutions and Gröbner bases of (infinite) sets of real polynomials by sequences of (finite) sets of rational polynomials.

The contiguity in R/M

An r.e. degree c is contiguous if degwtt(A) = degwtt(B) for any r.e. sets A, B ∈ c. In this paper, we generalize the notation of contiguity to the structure R/M, the upper semilattice of the r.e. degree set R modulo the cappable r.e. degree set M. An element [c] ∈ R/M is contiguous if [degwtt(A)] = [degwtt(B)] for any r.e. sets A, B such that degT(A), degT(B) ∈ [c]. It is proved in this paper that every nonzero element in R/M is not contiguous, i.e., for every element [c] ∈ R/M, if [c] ≠ [o] then there exist at least two r.e. sets A, B such that degT(A),degT(B) ∈ [c] and [degwtt(A)] ≠ [degwtt(B)].Anr.e. degree c is contiguous if degwtt(A)=degwtt(B) for anyr.e. setsA, B∈c. In this paper, we generalize the notation of contiguity to the structure R/M, the upper semilattice of ther.e. degree set R modulo the cappabler.e. degree set M. An element [c]∈R/M is contiguous if [degwtt(A)]=[degwtt(B)] for anyr.e. setsA, B such that degT(A) degT(B)∈[c]. It is proved in this paper that every nonzero element in R/M is not contiguons, i.e., for every element [c]∈R/M, if [c]≠[o] then there exist at least twor.e. setsA, B such that degT(A), degT(B)∈[c] and [degwtt(A)]≠[degwtt(B)].

Based-CR knowledge interconnection in National Knowledge Infrastructure

Knowledge processing is an important research field in AI. A lot of research practice has proved that it is necessary for a computer to really realize intelligence to have enough knowledge (i.e. knowledge base) and interconnections among knowledge. Otherwise, Johnson has advanced a famous three stage knowledge acquisition model, where knowledge is not isolated but mutually relevant. These sufficiently show that it is necessary and feasible to research knowledge interconnection in knowledge processing. We firstly introduce the CR (concept-relation) model and hierarchical concept graph (HCG), and then research knowledge interconnection in the National Knowledge Infrastructure (NKI).

Local noncuppability in R/M

Given any [c], [a], [d] xxxxxxxxR/M such that [d] ≤ [a] ≤ [c], [a] is locally noncuppable between [c] and [d] if [d] < [a] ≤ [c] and [a] ∨ [b] < [c] for any [b] xxxxxxxxR/M such that [d] ≤ [b] < [c]. It will be shown that given any nonzero [c] xxxxxxxxR/M, there are [a], [d] xxxxxxxxR/M such that [d] < [a] ≤ [c] and [a] is locally noncuppable between [c] and [d].