Abstract: A document retrieval system for searching a document coinciding with a retrieval request the user inputs and further ranking the document in accordance with the degree of coincidence between the document and the retrieval request. In the document retrieval system, a word frequency calculating section finds out the number of documents where a word appears, a frequency of occurrence of the word in a document and obtains a weighting parameter for the word, and a frequency score calculating section obtains a frequency score on the basis of the output of the word frequency calculating section. In addition, a word cooccurrence relation checking section checks word cooccurrence relations of the retrieval request and the document, and a cooccurrence score calculating section calculates a cooccurrence score from the degree of coincidence therebetween. A document score calculating section calculates a document score on the basis of the frequency score and the cooccurrence score.

Abstract: A table of derivation elements and a state-transition table indicating applicable strings of derivation types are produced according to pronunciation expanding rules in a similar character string expanding apparatus. Each of the derivation elements is composed of a derived sound derived from a key sound placed at a key position of a question pronunciation character string, a sound position of the derived sound in each of character strings expanded from the question pronunciation character string and one or more derivation types indicating how the derived sound placed at the sound position is derived from the key sound placed at the key position.

Abstract: A document storing and managing system for storing plural electronic documents in each of folders according to classifications and managing the stored electronic documents in a unit of the folder has a folder managing means for managing attributes of the electronic documents included in each of the folders, a document version managing means for managing information as to version of the electronic documents included in each of the folder, and a folder version managing means for managing a correspondence relation between a version of the folder and a version of each of the electronic documents included in the folder. The document storing and managing system of this invention may set and manage a version of a folder while keeping adjustability with a version of each document.

Abstract: A new type of text search apparatus, capable of finding all occurrence positions of a search string that is an arbitrary character string, within a text which is written as a continous sequence of characters, utilizes for text position reference purposes in an index file, words which each occur (at least once within the text) as the maximum length word, referred to as an extension word, among a set of arbitrarily predefined dictionary words extending from a specific character position. Each such occurrence of a word as an extension word defines one of a set of text position elements, with that set covering all of the character positions of the text. The index file also includes a table which relates each of the extension words to the respective positions at which each of the partial character strings of the word occur within the word.

Abstract: A constructing method of a finite state machine with failure transitions FFM is disclosed. The machine FFM is constructed from a nondeterministic finite-state machine and a string of external inputs. States <p,q> in the machine FFM is formed of a state set q included in the nondeterministic finite-state machine and a set p defined as a subset of the state set q, and the number of states <p,q> is finite. Also, an external input c takes the machine FFM from a current state s to a next state g(s,c) and an output .mu.(s) is output from the next state g(s,c) in cases where a value g(s,c) of a success function g is defined, and an external input c takes the machine FFM from the current state s to a state g(f(f...f(s)...)) determined by repeatedly calculating a value f(s) of a failure function f until a value g(f(f...f(s)...)) defined is found out in cases where the value g(s,c) of the success function g is not defined.

Abstract: A co-occurrence dictionary is built through a process for calculating three kinds of co-occurrence information and a real number vector corresponding to each category. The co-occurrence dictionary is updated through a process for selecting the opposite phrase of the co-occurrence for the additional co-occurrence information and a process for calculating a real number vector corresponding to an additional word on the basis of the additional co-occurrence information. A co-occurrence analysis is effected through a process for calculating in real number the degree of the co-occurrence on the basis of the real number vectors corresponding to two categories to be checked in the co-occurrence relation, and a semantic analysis is effected through a process for indicating, by a numerical value, the propriety of the interpretation on the basis of the degree of each co-occurrence.

Abstract: A concept dictionary management device includes a fundamental concept dictionary data holding portion for holding fundamental concept network connection information which represents fundamental concept network connections among words stored in a concept dictionary, a first supplemental concept dictionary data holding portion for holding first supplemental concept network connection information to be used for adding words to and deleting words from the concept network connections represented by the fundamental concept network connection information, a second supplemental concept dictionary data holding portion for holding second supplemental concept network connection information to be used only for personal use, to add one or more words to and deleting one or more words from the concept network connections obtained as a result of an addition of a deletion of the words connection by using the first supplemental concept network connection information, a concept dictionary retrieval portion for retrieving a conc

Abstract: A finite state automaton (FSA) text search apparatus converts each of successively received binary code numbers, representing successive characters of a text to be searched, into two portions. Two portions are extracted from a state number which is currently being read out from a first table memory, with one portion being combined with one of the input character portions and the combination used to select one of a set of 1st displacement numbers from a second table memory, while a second portion of the current state number and the other portion of the input character number similarly select one of a set of 2nd displacement numbers from a third table memory, and the entire current state number selects one of a set of base numbers from a fourth table memory. The selected 1st and 2nd displacement numbers are combined with the selected base number, and the result used to select the next state number from the first table memory.