Back to results
US 8,285,714 B2Granted
Method and apparatus for providing related words for queries using word co-occurrence frequency
AssigneeSungkyunkwan University Foundation for Corporate Collaboration
InventorsHyungsuk Ji, Hyunseung Choo
Life of this patent
Estimating the term…
Abstract
A system and a method is provided for extracting, classifying, and displaying related words for a query, if an internet or local computer or a mobile appliance user inputs the query. In a system for providing related words composed of a client and a server connected to the client through a network, the method of providing related words for a query includes the steps of the server searching for the co-occurrence frequency by words from source data, the server storing the result of the search in a database, the server receiving an input of a query from the client, and the server extracting related words related to the input query from the database.
Claims22
An apparatus for providing related words for a query, comprising:
- a communication module connected to an external client through a network to receive an input query from the client;
- a relational database storing data including at least one first word, at least one second word co-occurring with the at least one first word, and a co-occurrence frequency, wherein the co-occurrence frequency indicates how many times the at least one second word occurs in relation to the at least one first word;
- and
- a related word control module extracting first words included in source data by searching the source data, extracting and storing second words co-occurring by the extracted first words and the co-occurrence frequency in the relational database, wherein the related word control module is configured to search for third words co-occurring by at least one of the second words, and to determine that the at least one second word is related to at least one of the first words if the third words include the at least one of the first words, and wherein the related word control module is configured to extract the query input through the communication module and related words from the relational database, wherein at least one of the communication module, the relational database, and the related word control module comprises a processor, wherein the related word control module comprises an advanced extraction module reading the co-occurrence words by words from the relational database, extracting related words related to the word among the co-occurrence words, and storing the extracted related words in the relational database, wherein the advanced extraction module reads the co-occurrence words by specified words from the relational database, extracts N upper co-occurrence words by the specified words in the order of the co-occurrence frequency, extracts second co-occurrence words with respect to the N co-occurrence words when the co-occurrence words occur, and judges the co-occurrence word for the specified word as being a related word if the specified word among the second co-occurrence words is within the order predetermined by the co-occurrence frequency order.
1 more element
A system for providing related words for a query, comprising:
- a client receiving a query from a user and transmitting the query through a network;
- and
- a server receiving the query from the client, extracting and transmitting related words related to the query to the client, the server comprising:
- a communication module connected to the external client through the network to receive an input of the query from the client;
- a relational database storing data including at least one first word, at least one second word co-occurring with the at least one first word, and a co-occurrence frequency, wherein the co-occurrence frequency indicates how many times the at least one second word occurs in relation to the at least one first word;
- and
- a related word control module extracting second words co-occurring by first words from source data, storing the second words in the relational database, wherein the related word control module is configured to search for third words co-occurring by at least one of the second words, and to determine that the at least one of the second words is related to at least one of first words if the third words include the at least one of the first words, and wherein the related word control module is configured to extract the related words related to the query input through the communication module from the database, wherein the related word control module comprises an advanced extraction module reading the co-occurrence words by words from the relational database, extracting related words related to the word among the co-occurrence words, and storing the extracted related words in the relational database, wherein the advanced extraction module reads the co-occurrence words by specified words from the relational database, extracts N upper co-occurrence words by the specified words in the order of the co-occurrence frequency, extracts second co-occurrence words with respect to the N co-occurrence words when the co-occurrence words occur, and judges the co-occurrence word for the specified word as being a related word if the specified word among the second co-occurrence words is within the order predetermined by the co-occurrence frequency order.
4 more elements
A method of providing related words for a query using a search system for providing related words composed of a client and a server connected to the client through a network, the method comprising:
- searching for a co-occurrence frequency by words from source data, wherein the co-occurrence frequency indicates how many times at least one second word occurs in relation to at least one first word;
- storing the result of the search in a database;
- receiving an input of a query from the client;
- searching for third words co-occurring by at least one of the second words;
- determining that the at least one of the second words is related to at least one of the first words if the third words include the at least one of the first words;
- and
- extracting related words related to the input query from the database, wherein the extracting the related words reads the co-occurrence words by specified words from the relational database, extracts N upper co-occurrence words by the specified words in the order of the co-occurrence frequency, extracts second co-occurrence words with respect to the N co-occurrence words when the co-occurrence words occur, and judges the co-occurrence words for the specified word as being related words if the specified word among the second co-occurrence words is within the order predetermined by the co-occurrence frequency order.
4 more elements
A method of providing related words for a query using a search system for providing related words composed of a client and a server connected to the client through a network, the method comprising:
- searching for a co-occurrence frequency by words from source data and storing the co-occurrence frequency in a database, wherein the co-occurrence frequency indicates how many times at least one second word occurs in relation to at least one first word;
- extracting related words by words on the basis of the searched co-occurrence frequency;
- storing the extracted related words in the database;
- receiving an input of a query from the client;
- searching for third words co-occurring by at least one of the second words;
- determining that at least one of the second words is related to at least one of the first words if the third words include the at least one of the first words;
- and
- extracting the related words related to the input query from the database, wherein the extracting the related words reads the co-occurrence words by specified words from the relational database, extracts N upper co-occurrence words by the specified words in the order of the co-occurrence frequency, extracts second co-occurrence words with respect to the N co-occurrence words when the co-occurrence words occur, and judges the co-occurrence words for the specified word as being related words if the specified word among the second co-occurrence words is within the order predetermined by the co-occurrence frequency order.
5 more elements
An apparatus for providing related words for a query, comprising:
- an input module receiving a query from a user;
- a relational database storing data including at least one first word, at least one second word co-occurring with the at least one first word, and a co-occurrence frequency, wherein the co-occurrence frequency indicates how many times the at least one second word occurs in relation to the at least one first word;
- and
- a related word control module extracting first words included in source data by searching the source data, extracting and storing second words co-occurring by the extracted first words and the co-occurrence frequency in the relational database, wherein the related word control module is configured to search for third words co-occurring by at least one of the second words, and to determine whether the at least one second word is related to at least one of the first words if the third words include the at least one of the first words, and wherein the related word control module is configured to extract the query inputted through the communication module and related words from the relational database, wherein at least one of the input module, the relational database, and the related word control module comprises a processor, wherein the related word control module comprises an advanced extraction module reading the co-occurrence words by words from the relational database, extracting related words related to the word among the co-occurrence words, and storing the extracted related words in the relational database, wherein the advanced extraction module reads the co-occurrence words by specified words from the relational database, extracts N upper co-occurrence words by the specified words in the order of the co-occurrence frequency, extracts second co-occurrence words with respect to the N co-occurrence words when the co-occurrence words occur, and judges the co-occurrence word for the specified word as being a related word if the specified word among the second co-occurrence words is within the order predetermined by the co-occurrence frequency order.
1 more element
A non-transitory computer-readable recording medium recorded with a program in a system for providing related words for a query composed of a client and a server connected to the client through a network, the program to perform operations comprising:
- searching for a co-occurrence frequency by words from source data, wherein the co-occurrence frequency indicates how many times at least one second word occurs in relation to at least one first word;
- storing the result of the search in a database;
- receiving an input of a query from the client;
- searching for third words co-occurring by at least one of the second words;
- determining that at least one of the second words is related to at least one of the first words if the third words include the at least one of the first words;
- and
- extracting related words related to the input query from the database, wherein the extracting the related words reads the co-occurrence words by specified words from the relational database, extracts N upper co-occurrence words by the specified words in the order of the co-occurrence frequency, extracts second co-occurrence words with respect to the N co-occurrence words when the co-occurrence words occur, and judges the co-occurrence words for the specified word as being related words if the specified word among the second co-occurrence words is within the order predetermined by the co-occurrence frequency order.
4 more elements
Dependencies are inferred from the claim text.