1. Field of the Invention
The present invention relates generally to regular expressions for natural language processing, and more specifically to optimizing generation of a regular expression, utilized for entity extraction, that can identify a word or a phrase having the word within text data (i.e., one or more strings of text) even if the word is misspelled.
2. Description of the Related Art
The tremendous growth of the Internet and computer storage capabilities has enabled people to have access to massive amounts of electronically stored data, wherein the data includes text data (i.e., one or more strings of text) stored on a computer readable tangible storage device. There are various computer software programs utilized as search tools capable of searching and identifying information within the text data. Specifically, it is known to utilize search tools having a regular expression to identify one or more specific words within the text data, in order to perform entity extraction. However, if the text data is not validated prior to being stored on the computer readable tangible storage device, then there can be a misspelling of a word within the text data. The misspelling of a word within the text data can result in a search tool not being able to identify the word that is misspelled even if the word is only slightly misspelled, which can further result in inaccurate and imprecise information entity extraction results.
Aspects of an embodiment of the present invention disclose a method, computer system, and program product for optimizing generation of a regular expression utilized for entity extraction. The computer receives a formatted query having a plain text word. The computer determines whether one or more performance optimization parameters are configured to optimize performance of generating a regular expression. In response to the computer determining the one or more performance optimization parameters are configured to optimize performance of generating the regular expression, the computer identifies syllables within the plain text word that have a high probability of at least one of an incorrect substituted and transposed character within a spelling of a word having a same number of syllables as the plain text word. The computer selects each character in the syllables identified. The computer identifies a group of characters from a confusion matrix that are commonly confused with the character selected. The computer generates a set of characters for each character selected, wherein the set of characters begin with one of the each character selected followed by and ending with the group of characters from the confusion matrix. The computer generates a regular expression by concatenating each of the set of characters.
The subject matter which is regarded as an embodiment of the present invention is particularly pointed out and distinctly claimed in the claims at the conclusion of the specification. One manner in which recited features of an embodiment of the present invention can be understood is by reference to the following detailed description of embodiments, taken in conjunction with the accompanying drawings in which:
As will be appreciated by one skilled in the art, aspects of the present invention may be embodied as a system, method or computer program product. Accordingly, aspects of the present invention may take the form of an entirely hardware embodiment, an entirely software embodiment (including firmware, resident software, micro-code, etc.) or an embodiment combining software and hardware aspects that may all generally be referred to herein as a “circuit,” “module” or “system.” Furthermore, aspects of the present invention may take the form of a computer program product embodied in one or more computer readable medium(s) having computer readable program code embodied thereon.
Any combination of one or more computer readable medium(s) may be utilized. The computer readable medium may be a computer readable signal medium or a computer readable storage medium. A computer readable storage medium may be, for example, but not limited to, an electronic, magnetic, optical, electromagnetic, infrared, or semiconductor system, apparatus, or device, or any suitable combination of the foregoing. More specific examples (a non-exhaustive list) of the computer readable storage medium would include the following: an electrical connection having one or more wires, a portable computer diskette, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or Flash memory), an optical fiber, a portable compact disc read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the foregoing. In the context of this document, a computer readable storage medium may be any tangible medium that can contain, or store a program for use by or in connection with an instruction execution system, apparatus, or device.
A computer readable signal medium may include a propagated data signal with computer readable program code embodied therein, for example, in baseband or as part of a carrier wave. Such a propagated signal may take any of a variety of forms, including, but not limited to, electro-magnetic, optical, or any suitable combination thereof. A computer readable signal medium may be any computer readable medium that is not a computer readable storage medium and that can communicate, propagate, or transport a program for use by or in connection with an instruction execution system, apparatus, or device.
Program code embodied on a computer readable medium may be transmitted using any appropriate medium, including but not limited to wireless, wireline, optical fiber cable, RF, etc., or any suitable combination of the foregoing.
Computer program code for carrying out operations for aspects of the present invention may be written in any combination of one or more programming languages, including an object oriented programming language such as JAVA, Smalltalk, C++ or the like, conventional procedural programming languages such as the “C” programming language, a hardware description language such as VERILOG, or similar programming languages. The program code may execute entirely on the user's computer, partly on the user's computer, as a stand-alone software package, partly on the user's computer and partly on a remote computer or entirely on the remote computer or server. In the latter scenario, the remote computer may be connected to the user's computer through any type of network, including a local area network (LAN) or a wide area network (WAN), or the connection may be made to an external computer (for example, through the Internet using an Internet Service Provider).
Aspects of the present invention are described below with reference to flowchart illustrations and/or block diagrams of methods, apparatus (systems) and computer program products according to embodiments of the invention. It will be understood that each block of the flowchart illustrations and/or block diagrams, and combinations of blocks in the flowchart illustrations and/or block diagrams, can be implemented by computer program instructions. These computer program instructions may be provided to a processor of a general purpose computer, special purpose computer, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, create means for implementing the functions/acts specified in the flowchart and/or block diagram block or blocks.
These computer program instructions may also be stored in a computer readable medium that can direct a computer, other programmable data processing apparatus, or other devices to function in a particular manner, such that the instructions stored in the computer readable medium produce an article of manufacture including instructions which implement the function/act specified in the flowchart and/or block diagram block or blocks.
The computer program instructions may also be loaded onto a computer, other programmable data processing apparatus, or other devices to cause a series of operational steps to be performed on the computer, other programmable apparatus or other devices to produce a computer implemented process such that the instructions which execute on the computer or other programmable apparatus provide processes for implementing the functions/acts specified in the flowchart and/or block diagram block or blocks.
Exemplary embodiments now will be described more fully herein with reference to the accompanying drawings. This disclosure may, however, be embodied in many different forms and should not be construed as limited to the exemplary embodiments set forth herein. Rather, these exemplary embodiments are provided so that this disclosure will be thorough and complete, and will fully convey the scope of this disclosure to those skilled in the art. In the description, details of well-known features and techniques may be omitted to avoid unnecessarily obscuring the presented embodiments.
Embodiments of the present invention provide a regular expression generation program having program code for optimizing generation of a regular expression, utilized for entity extraction, that can identify a word or a phrase having the word within text data (i.e., one or more strings of text) even if the word is misspelled. A regular expression is a specific pattern that provides a way to match (i.e., define and identify) strings of text such as specific characters, words, or patterns of characters within text data. As a result, the regular expression generation program can yield search results (e.g., extract entity information) that have high degree of accurate and precision.
Query program 150 is a software tool that can send a query to entity extraction program 160 in order to retrieve data from electronic repository 165. In response to receiving a query from query program 150, entity extraction program 160 can extract data from electronic repository 165. However, to accurately extract the data based on the query received, entity extraction program 160 sends a version of the query to regular expression generation program 170. Regular expression generation program 170 is program code that generates a regular expression by utilizing confusion matrix 180, the version of the query received by entity extraction program 160, and data from data analytics software 175. Entity extraction program 160 can utilize the regular expression that is generated to accurately identify a string of text within the data stored in electronic repository 165 even if the string of text is misspelled. Upon accurately identifying the string of text within the data entity extraction program 160 can extract the data, and send the data to the end-user.
If the one or more performance optimization parameters are not configured to optimize performance of generating a regular expression (decision block 205, “NO” branch), then the program code selects a character in the plain text word (block 210). Next, the program code identifies a group of characters from confusion matrix 180 that are most commonly confused with (i.e., most commonly substituted for or even transposed with) the character selected when spelling a word, thereby causing the word to be misspelled (block 225).
Otherwise, if the one or more performance optimization parameters are configured to optimize performance of generating a regular expression (decision block 205, “YES” branch), then the program code identifies syllables within the plain text word that are most likely to have incorrect substituted or transposed characters when spelling a word having a same number of syllables as the plain text word (block 215). Specifically, based on the number of syllables in the plain text word, the program code of regular expression generation program 170 can be configured to identify each of the syllables that have a high probability (e.g., a probability ranging from approximately 50% to about 99%) of spelling errors.
Data on probability of spelling errors based on the number of syllables in the plain text word may be collected from industry research and/or data analytics software 175 and utilized to configure the program code of regular expression generation program 170. Specifically, the data on the probability of spelling errors indicates syllables in which misspellings are most likely to occur within the plain text word, and can be utilized to configure the program code to select one or more of the syllables within the plain text word, in order to optimize performance of the program code generating a regular expression. For example, in the disclosed embodiment, the plain text word is “interembedded”, which has five syllables. The data on probability of spelling errors, generated by data analytics software 175, indicates that the first, middle, and last syllables of five syllable words have a high probability of spelling errors. Accordingly, since the plain text word “interembedded” is a five syllable word, the program code is configured to select first syllable “in”, middle syllable “em”, and last syllable “ed”. However, in other embodiments, the plain text word may be different and have a different number of syllables. Thus, the program code can be configured to select any combination of syllables or only a single syllable from the plain text word, by utilizing the data on probability of spelling errors generated by data analytics software 175.
Next, the program code selects each character in the syllables identified (block 220). Subsequently, for each character selected the program code identifies a group of characters from confusion matrix 180 that are most commonly confused with (i.e., most commonly substituted for or even transposed with) the character selected when spelling a word, thereby causing the word to be misspelled (block 225). In the disclosed embodiment, to identify the group of characters the program code searches confusion matrix 180 for three characters most commonly confused with the character selected by the program code. However, in other embodiments, to identify the group of characters the program code can search confusion matrix 180 for a top percentage of characters (e.g., top 80% of characters) most commonly confused with the character selected by the program code. Thus, the number of characters the program code searches for within confusion matrix 180, is configurable based on end-user requirements.
Subsequent to identifying the group of characters from confusion matrix 180, if the one or more performance optimization parameters are configured to optimize performance of generating a regular expression (decision block 230, “YES” branch), then the program code generates a set of characters, for each character selected, beginning with the character selected followed by and ending with the group of characters that are most commonly confused with the character selected (block 235). Next, the program code receives data from data analytics software (block 240).
If the program code determines from the data received that the character selected is not often omitted with a word (decision block 245, “NO” branch), then the program code generates a regular expression by concatenating each set of characters based on order that each character selected appears in the plain text word (block 250). The program code can associate a configurable numerical value to the concatenated set of characters if the Boolean edit distance parameter is set (e.g., set to a value of one). The configurable numerical value represents an amount of variation allowed between a spelling of the plain text word and the spelling of a word subsequently identified within electronic repository 165, utilizing the regular expression that is generated, in order for the word to be considered a possible match with the plain text word. A system programmer/administrator can configure the numerical value based on end-user requirements.
However, if the program code determines from the data received that the character selected is often omitted within a word, based on number of syllables in the word or some other correlating factor, thereby causing the word to be misspelled (decision block 245, “YES” branch), then prior to generating a regular expression the program code tags the set of characters (block 260). For example, the tag can take many forms such as an ampersand, an asterisk, or a question mark. The program code treats the selected character within the set of characters that is tagged as optional, which means that the regular expression can be utilized to search for spelling of the plain text word, within text data, in which the selected character is omitted. Subsequently, the program code generates a regular expression by concatenating each set of characters based on order that each character selected appears in the plain text word (block 250). Next, the program code sends the regular expression to entity extraction program 160 (block 255), and then the program code ends. Entity extraction program 160 can utilize the regular expression to identify a word or a phrase having the word within text data (i.e., one or more strings of text) even if the word is misspelled, and to extract data associated to the word and/or phrase from electronic repository 165.
Otherwise, if the one or more performance optimization parameters are not configured to optimize performance of generating a regular expression (decision block 230, “NO” branch), and if there is another character in the plain text word that has not been previously selected (decision block 265, “YES” branch), then the program code selects next character in the plain text word (block 270). Subsequently, the program code identifies a group of characters from confusion matrix 180 that are most commonly confused with (i.e., most commonly substituted for or even transposed with) the character selected when spelling a word, thereby causing the word to be misspelled (block 225). However, if there is not another character in the plain text word that has not been previously selected (decision block 265, “NO” branch), then the program code generates a regular expression by concatenating each set of characters based on order that each character selected appears in the plain text word (block 250). Next, the program code sends the regular expression to entity extraction program 160 (block 255), and then the program code ends. Entity extraction program 160 can utilize the regular expression to identify a word or a phrase having the word within text data (i.e., one or more strings of text) even if the word is misspelled, and to extract data associated to the word and/or the phrase from electronic repository 165.
Accordingly, based on the given sample size of words, data structure 300 indicates the correct character “i” was confused with incorrect character “a” one hundred eighteen times out of a total of three hundred twenty-nine times (i.e., confused approximately 35.8% of the time), confused with incorrect character “e” eighty-nine times out of a total of three hundred twenty-nine times (i.e., confused approximately 27.1% of the time), confused with incorrect character “l” thirteen times out of a total of three hundred twenty-nine times (i.e., confused approximately 4% of the time), confused with incorrect character “n” one time out of a total of three hundred twenty-nine times (i.e., confused approximately 0.3% of the time), confused with incorrect character “o” twenty-five times out of a total of three hundred twenty-nine times (i.e., confused approximately 7.6% of the time), confused with incorrect character “p” two times out of a total of three hundred twenty-nine times (i.e., confused approximately 0.6% of the time), confused with incorrect character “r” two times out of a total of three hundred twenty-nine times (i.e., confused approximately 0.6% of the time), confused with incorrect character “u” sixty-four times out of a total of three hundred twenty-nine times (i.e., confused approximately 19.5% of the time), confused with incorrect character “y” fifteen times out of a total of three hundred twenty-nine times (i.e., confused approximately 4.6% of the time).
Each set of internal components 800a and 800b includes a R/W drive or interface 832 to read from and write to one or more portable computer readable tangible storage devices 936 such as CD-ROM, DVD, memory stick, magnetic tape, magnetic disk, optical disk or semiconductor storage device. Regular expression generation program 170 on server computer 105b can be stored on one or more of the respective portable computer readable tangible storage devices 936, read via the respective R/W drive or interface 832 and loaded into the respective hard drive or computer readable tangible storage device 830.
Furthermore, each set of internal components 800a and 800b also includes a network adapter or interface 836 such as TCP/IP adapter card, wireless wi-fi interface card, or 3G or 4G wireless interface card or other wired or wireless communication link. Regular expression generation program 170 on server computer 105b can be downloaded to respective computer 105b from an external computer or external storage device via a network (for example, the Internet, a LAN, or a WAN) and respective network adapters or interfaces 836. From the network adapter or interface 836, regular expression generation program 170 on server computer 105b are loaded into respective hard drive or computer readable tangible storage device 830. The network may comprise copper wires, optical fibers, wireless transmission, routers, firewalls, switches, gateway computers and/or servers.
Each set of external components 900a and 900b can include a computer display monitor 920, a keyboard 930, and a computer mouse 934. External components 900a and 900b can also include touch screens, virtual keyboards, touch pads, pointing devices, and other human interface devices. Each set of internal components 800a and 800b also includes device drivers 840 to interface to computer display monitor 920, keyboard 930 and computer mouse 934. The device drivers 840, R/W drive or interface 832 and network adapter or interface 836 comprise hardware and software in which the software is stored in computer readable tangible storage device 830 and/or ROM 824.
It should be appreciated that
Aspects of the present invention are described above with reference to flowchart illustrations and/or block diagrams of methods, apparatus (systems) and computer program products according to embodiments of the invention. It will be understood that each block of the flowchart illustrations and/or block diagrams, and combinations of blocks in the flowchart illustrations and/or block diagrams, can be implemented by computer program instructions. These computer program instructions may be provided to a processor of a general purpose computer, special purpose computer, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, create means for implementing the functions/acts specified in the flowchart and/or block diagram block or blocks.
Computer program code for carrying out operations for aspects of an embodiment of the invention may be written in any combination of one or more programming languages, including an object oriented programming language such as JAVA, Smalltalk, C++ or the like and conventional procedural programming languages, such as the “C” programming language or similar programming languages. The program code may execute entirely on the end-user's computer, partly on the end-user's computer, as a stand-alone software package, partly on the end-user's computer and partly on a remote computer, or entirely on a remote computer or server. In the latter scenario, the remote computer may be connected to the end-user's computer through any type of network, including a LAN or WAN, or the connection may be made to an external computer (for example, through the Internet using an Internet Service Provider). Alternatively, the functions of the aforementioned programs can be implemented in whole or in part by computer circuits and other hardware (not shown).
Based on the foregoing, a method, computer system, and computer program product have been disclosed for optimizing generation of a regular expression, utilized for entity extraction, that can identify a word within text data (i.e., one or more strings of text) even if the word is misspelled. However, numerous modifications and substitutions can be made without deviating from the scope of an embodiment of the invention. Therefore, one or more embodiments of the invention have been disclosed by way of example and not limitation.