# US 2018/0024993 A1 · Computer-readable recording medium, determination device and determination method > Agents: this data is available as JSON from the patentgrep API and MCP server — see https://patentgrep.com/llms.txt - Number: US 2018/0024993 A1 - Document ID: US20180024993A1 - Status: Published application - Kind: A1 - Application number: 15/624,361 - Filed: 2017-06-15 - Published: 2018-01-25 - Earliest claimed: 2016-07-25 - Assignees: Fujitsu Limited - Applicants: Fujitsu Limited - Inventors: Naohiro Itou; Nobuko Takase; Kazuo Mineno - Web page: https://patentgrep.com/patent/US20180024993A1 ## CPC - G06F 40/247: Thesauruses; Synonyms - G06F 40/242: Dictionaries - G06F 40/253: Grammatical analysis; Style critique - G06F 40/268: Morphological analysis - G06F 40/30: Semantic analysis - G06F 40/35: Discourse or dialogue representation - G06F 40/53: Processing of non-Latin text (kana-to-kanji conversion G06F40/129; vowelisation G06F40/232) - G06F: ELECTRIC DIGITAL DATA PROCESSING (computer systems based on specific computational models G06N) ## Abstract An information processing device receives sentence data. The information processing device generates sets of information, each of the sets indicating a relationship between each of words included in the received sentence data and another word among the words by executing a semantic analysis process on the words. The information processing device determines a similarity between the words based on a similarity between the generated the sets of information. The information processing device outputs a result of the determining. ## Claims ### Claim 1 A non-transitory computer-readable recording medium having stored therein a program that causes a computer to execute a process comprising: receiving sentence data; generating sets of information, each of the sets indicating a relationship between each of words included in the received sentence data and another word among the words by executing a semantic analysis process on the words; determining a similarity between the words based on a similarity between the generated the sets of information; and outputting a result of the determining. ### Claim 2 (depends on 1) The non-transitory computer-readable recording medium according to claim 1, wherein the generating includes, with regard to each word included in the words, generating idea information that indicates an idea symbol that represents an idea possessed by the word, a relationship symbol that represents a relationship between the word and a different word included in the words, and a direction of a corresponding relationship. ### Claim 3 (depends on 2) The non-transitory computer-readable recording medium according to claim 2, wherein the determining includes determining a similarity between the words by making a determination as to whether a word with an idea symbol included in the idea information is connected to each of a plurality of different words included in the words with an identical relationship symbol and in an identical direction of a relationship, and the outputting includes, as a result of the determination, outputting a list of different words that are connected with an identical relationship symbol to a first list information. ### Claim 4 (depends on 3) The non-transitory computer-readable recording medium according to claim 3, wherein the determining includes applying a specific hash function to an idea symbol that is possessed by the word, a relationship symbol that is possessed by the word, and a direction of a relationship to calculate a hash value; adding, to the first list information, the different word that is connected to the word in relation to the hash value calculated; and deleting an entry of a hash value, of which a number of the added different words is one, from the first list information. ### Claim 5 (depends on 4) The non-transitory computer-readable recording medium according to claim 4, the process further including: generating, from the first list information, second list information that traces in reverse from the different word to an idea symbol of an original word; extracting, as a pair, entries that include the different word that is identical by using the first list information and the second list information; and when extracted two pairs of entries have a common relationship, determining that different words in the two pairs of entries are synonym candidates. ### Claim 6 A determination device comprising: a processor; a memory, wherein the processor executes a process comprising: receiving sentence data; generating sets of information, each of the sets indicating a relationship between each of words included in the received sentence data and another word among the words by executing a semantic analysis process on the words; determining a similarity between the words based on a similarity between the generated the sets of information; and outputting a result of determining. ### Claim 7 A determination method to be performed by a computer, the determination method comprising: generating sets of information, each of the sets indicating a relationship between each of words included in the received sentence data and another word among the words by executing a semantic analysis process on the words using the processor; determining a similarity between the words based on a similarity between the generated the sets of information using the processor; and outputting a result of the determining using the processor.