# US 2017/0286399 A1 · System, method, and recording medium for corpus pattern paraphrasing > Agents: this data is available as JSON from the patentgrep API and MCP server — see https://patentgrep.com/llms.txt - Number: US 2017/0286399 A1 - Document ID: US20170286399A1 - Status: Published application - Kind: A1 - Application number: 15/087,023 - Filed: 2016-03-31 - Published: 2017-10-05 - Earliest claimed: 2016-03-31 - Assignees: International Business Machines Corporation - Applicants: International Business Machines Corporation - Inventors: Octavian Popescu; Vadim Sheinin - Web page: https://patentgrep.com/patent/US20170286399A1 ## CPC - G06F 40/30: Semantic analysis - G06F 40/211: Syntactic parsing, e.g. based on context-free grammar [CFG] or unification grammars - G06F 40/247: Thesauruses; Synonyms - G06F 40/284: Lexical analysis, e.g. tokenisation or collocates - G06F 40/205: Parsing - G06F 40/237: Lexical tools - G06F 40/289: Phrasal analysis, e.g. finite state techniques or chunking - G06F 40/40: Processing or translation of natural language (natural language analysis G06F40/20; semantic analysis G06F40/30) - G06F 40/44: Statistical methods, e.g. probability models - G06F 40/45: Example-based machine translation; Alignment - G10L 15/1822: {Parsing for meaning understanding} - G06F: ELECTRIC DIGITAL DATA PROCESSING (computer systems based on specific computational models G06N) - G10L: SPEECH ANALYSIS TECHNIQUES OR SPEECH SYNTHESIS; SPEECH RECOGNITION; SPEECH OR VOICE PROCESSING TECHNIQUES; SPEECH OR AUDIO CODING OR DECODING ## Abstract A corpus pattern paraphrasing method, system, and non-transitory computer readable medium, include an analyzing circuit configured to analyze a corpus of sentences stored in a database to determine regular structures including a plurality of substitute words for verbs expressed as patterns and apply deep learning of the regular structures over the patterns, a representative word determining circuit configured to determine a plurality of representative words that represents each class of word of the regular structures, and an aligning circuit configured to align word slots of a paraphrase pattern of the classes of words replaced with substitute words and representative words in the paraphrase pattern to give a same semantic meaning to the paraphrase pattern as a sentence of the corpus of sentences. ## Claims ### Claim 1 A corpus pattern paraphrasing system comprising: an analyzing circuit configured to: analyze a corpus of sentences stored in a database to determine regular structures including a plurality of substitute words for verbs expressed as patterns and apply deep learning of the regular structures over the patterns; and create a plurality of regular rules based on the corpus of sentences analyzed, the plurality of regular rules being created independently to the database to predict a meaning of the corpus of sentences without referring to the database; a representative word determining circuit configured to determine a plurality of representative words that represents each class of word of the regular structures; and an aligning circuit configured to align word slots of a paraphrase pattern of the classes of words replaced with substitute words and representative words in the paraphrase pattern to give a same semantic meaning to the paraphrase pattern as a sentence of the corpus of sentences. ### Claim 2 (depends on 1) The system of claim 1, wherein the aligning circuit determines the same semantic meaning according to a probability calculated based on ordering the substitute words in the paraphrase pattern in each possible order. ### Claim 3 (depends on 1) The system of claim 1, wherein the analyzing circuit determines a syntactical feature of each word in the corpus of sentences based on lexical features of the sentences in the corpus of sentences and assigns a generic placeholder for the syntactical feature such that the representative word determining circuit replaces the generic placeholder with a different word having the same syntactical feature. ### Claim 4 (depends on 1) The system of claim 1, wherein the aligning circuit computes a probability of the paraphrase pattern having the same semantic meaning for each order of the classes of words replaced with the substitute words. ### Claim 5 (depends on 1) The system of claim 1, wherein the substitute words have a same syntactical meaning as the classes of words. ### Claim 6 (depends on 1) The system of claim 1, wherein the plurality of substitute words comprise a plurality of most frequently used words in the corpus of sentences that have a same lexical feature. ### Claim 7 (depends on 1) The system of claim 1, wherein the aligning circuit determines the same semantic meaning according to a probability that the word slots of a paraphrase pattern have a same lexical feature in the respective word slots. ### Claim 8 (depends on 1) The system of claim 1, wherein the aligning circuit aligns the word slots such that verbs are positioned in each possible order with the classes of words. ### Claim 9 (depends on 8) The system of claim 8, wherein the aligning circuit calculates a probability of having the same semantic meaning for each possible order of the verbs and the representative words, and wherein the paraphrase pattern comprises the order of the verbs and the representative words having a highest probability of the same semantic meaning. ### Claim 10 (depends on 1) The system of claim 1, wherein the analyzing circuit determines verbal phrases that are synonymic with a sentence of the corpus of sentences. ### Claim 11 A corpus pattern paraphrasing method, comprising: analyzing a corpus of sentences stored in a database to determine regular structures including a plurality of substitute words for verbs expressed as patterns and apply deep learning of the regular structures over the patterns; creating a plurality of regular rules based on the corpus of sentences analyzed, the plurality of regular rules being created independently to the database to predict a meaning of the corpus of sentences without referring to the database; determining a plurality of representative words that represents each class of word of the regular structures; and aligning word slots of a paraphrase pattern of the classes of words replaced with substitute words and representative words in the paraphrase pattern to give a same semantic meaning to the paraphrase pattern as a sentence of the corpus of sentences. ### Claim 12 (depends on 11) The method of claim 11, wherein the aligning determines the same semantic meaning according to a probability calculated based on ordering the substitute words in the paraphrase pattern in each possible order. ### Claim 13 (depends on 11) The method of claim 11, wherein the analyzing determines a syntactical feature of each word in the corpus of sentences based on lexical features of the sentences in the corpus of sentences and assigns a generic placeholder for the syntactical feature such that the representative word determining circuit replaces the generic placeholder with a different word having the same syntactical feature. ### Claim 14 (depends on 11) The method of claim 11, wherein the aligning circuit computes a probability of the paraphrase pattern having the same semantic meaning for each order of the classes of words replaced with the substitute words. ### Claim 15 (depends on 11) The method of claim 11, wherein the substitute words have a same syntactical meaning as the classes of words. ### Claim 16 A non-transitory computer-readable recording medium recording a corpus pattern paraphrasing program, the program causing a computer to perform: analyzing a corpus of sentences stored in a database to determine regular structures including a plurality of substitute words for verbs expressed as patterns and apply deep learning of the regular structures over the patterns; creating a plurality of regular rules based on the corpus of sentences analyzed, the plurality of regular rules being created independently to the database to predict a meaning of the corpus of sentences without referring to the database; determining a plurality of representative words that represents each class of word of the regular structures; and aligning word slots of a paraphrase pattern of the classes of words replaced with substitute words and representative words in the paraphrase pattern to give a same semantic meaning to the paraphrase pattern as a sentence of the corpus of sentences. ### Claim 17 (depends on 16) The non-transitory computer-readable recording medium of claim 16, wherein the aligning determines the same semantic meaning according to a probability calculated based on ordering the substitute words in the paraphrase pattern in each possible order. ### Claim 18 (depends on 16) The non-transitory computer-readable recording medium of claim 16, wherein the analyzing determines a syntactical feature of each word in the corpus of sentences based on lexical features of the sentences in the corpus of sentences and assigns a generic placeholder for the syntactical feature such that the representative word determining circuit replaces the generic placeholder with a different word having the same syntactical feature. ### Claim 19 (depends on 16) The non-transitory computer-readable recording medium of claim 16, wherein the aligning circuit computes a probability of the paraphrase pattern having the same semantic meaning for each order of the classes of words replaced with the substitute words. ### Claim 20 (depends on 16) The non-transitory computer-readable recording medium of claim 16, wherein the substitute words have a same syntactical meaning as the classes of words.