Patent Application
20210350078
The present invention relates generally to the field of computing, and more particularly to online recruiting systems.
In a cognitive recruiting scenario, matching candidates to available positions may involve a multi-faceted matching process based on the candidates' skills, history, and competencies (e.g., comprising at least one level of granularity such as basic, experienced, and/or expert). The process of matching candidates to available positions may additionally occur in an ever-changing market, in which the requirements of an available position may be based on an industry context, and wherein available positions with the same name existing in different industries may have different skills requirements.
Embodiments of the present invention disclose a method, computer system, and a computer program product for semantic matching. The present invention may include pre-processing and normalizing a job title. The present invention may include deconstructing the job title based on at least one semantic element. The present invention may include training a machine learning model. The present invention may include creating a contextual word representation of the job title using the at least one semantic element of the job title. The present invention may include computing a similarity score for each of the at least one semantic element of the job title. The present invention may lastly include applying a weight to the computed similarity score before making a final match assessment.
These and other objects, features and advantages of the present invention will become apparent from the following detailed description of illustrative embodiments thereof, which is to be read in connection with the accompanying drawings. The various features of the drawings are not to scale as the illustrations are for clarity in facilitating one skilled in the art in understanding the invention in conjunction with the detailed description. In the drawings:
Detailed embodiments of the claimed structures and methods are disclosed herein; however, it can be understood that the disclosed embodiments are merely illustrative of the claimed structures and methods that may be embodied in various forms. This invention 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 invention 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.
The present invention may be a system, a method, and/or a computer program product at any possible technical detail level of integration. The computer program product may include a computer readable storage medium (or media) having computer readable program instructions thereon for causing a processor to carry out aspects of the present invention.
The computer readable storage medium can be a tangible device that can retain and store instructions for use by an instruction execution device. The computer readable storage medium may be, for example, but is not limited to, an electronic storage device, a magnetic storage device, an optical storage device, an electromagnetic storage device, a semiconductor storage device, or any suitable combination of the foregoing. A non-exhaustive list of more specific examples of the computer readable storage medium includes the following: 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), a static random access memory (SRAM), a portable compact disc read-only memory (CD-ROM), a digital versatile disk (DVD), a memory stick, a floppy disk, a mechanically encoded device such as punch-cards or raised structures in a groove having instructions recorded thereon, and any suitable combination of the foregoing. A computer readable storage medium, as used herein, is not to be construed as being transitory signals per se, such as radio waves or other freely propagating electromagnetic waves, electromagnetic waves propagating through a waveguide or other transmission media (e.g., light pulses passing through a fiber-optic cable), or electrical signals transmitted through a wire.
Computer readable program instructions described herein can be downloaded to respective computing/processing devices from a computer readable storage medium or to an external computer or external storage device via a network, for example, the Internet, a local area network, a wide area network and/or a wireless network. The network may comprise copper transmission cables, optical transmission fibers, wireless transmission, routers, firewalls, switches, gateway computers and/or edge servers. A network adapter card or network interface in each computing/processing device receives computer readable program instructions from the network and forwards the computer readable program instructions for storage in a computer readable storage medium within the respective computing/processing device.
Computer readable program instructions for carrying out operations of the present invention may be assembler instructions, instruction-set-architecture (ISA) instructions, machine instructions, machine dependent instructions, microcode, firmware instructions, state-setting data, configuration data for integrated circuitry, or either source code or object code written in any combination of one or more programming languages, including an object oriented programming language such as Smalltalk, C++, or the like, and procedural programming languages, such as the “C” programming language or similar programming languages. The computer readable program instructions 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). In some embodiments, electronic circuitry including, for example, programmable logic circuitry, field-programmable gate arrays (FPGA), or programmable logic arrays (PLA) may execute the computer readable program instructions by utilizing state information of the computer readable program instructions to personalize the electronic circuitry, in order to perform aspects of the present invention.
Aspects of the present invention are described herein 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 readable program instructions.
These computer readable 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 readable program instructions may also be stored in a computer readable storage medium that can direct a computer, a programmable data processing apparatus, and/or other devices to function in a particular manner, such that the computer readable storage medium having instructions stored therein comprises an article of manufacture including instructions which implement aspects of the function/act specified in the flowchart and/or block diagram block or blocks.
The computer readable program instructions may also be loaded onto a computer, other programmable data processing apparatus, or other device to cause a series of operational steps to be performed on the computer, other programmable apparatus or other device to produce a computer implemented process, such that the instructions which execute on the computer, other programmable apparatus, or other device implement the functions/acts specified in the flowchart and/or block diagram block or blocks.
The flowchart and block diagrams in the Figures illustrate the architecture, functionality, and operation of possible implementations of systems, methods, and computer program products according to various embodiments of the present invention. In this regard, each block in the flowchart or block diagrams may represent a module, segment, or portion of instructions, which comprises one or more executable instructions for implementing the specified logical function(s). In some alternative implementations, the functions noted in the blocks may occur out of the order noted in the Figures. For example, two blocks shown in succession may, in fact, be executed substantially concurrently, or the blocks may sometimes be executed in the reverse order, depending upon the functionality involved. It will also be noted that each block of the block diagrams and/or flowchart illustration, and combinations of blocks in the block diagrams and/or flowchart illustration, can be implemented by special purpose hardware-based systems that perform the specified functions or acts or carry out combinations of special purpose hardware and computer instructions.
The following described exemplary embodiments provide a system, method and program product for semantic matching. As such, the present embodiment has the capacity to improve the technical field of cognitive recruitment by accounting for an inherent semantic structure of a job title, augmenting the job title with contextual knowledge of the semantic elements of the job title, and by computing a similarity score based on both general and domain specific knowledge bases, which score may indicate the similarity of components of the job title to components of a second job title. More specifically, the present invention may include pre-processing and normalizing a job title. The present invention may include deconstructing the job title based on at least one semantic element. The present invention may include training a machine learning model. The present invention may include creating a contextual word representation of the job title using the at least one semantic element of the job title. The present invention may include computing a similarity score for each of the at least one semantic element of the job title. The present invention may lastly include applying a weight to the computed similarity score before making a final match assessment.
As described previously, in a cognitive recruiting scenario, matching candidates to available positions may involve a multi-faceted matching process based on the candidates' skills, history, and competencies (e.g., comprising at least one level of granularity such as basic, experienced, and/or expert). The process of matching candidates to available positions may additionally occur in an ever-changing market, in which the requirements of an available position may be based on an industry context, and wherein available positions with the same name existing in different industries may have different skills requirements.
Therefore, it may be advantageous to, among other things, provide a means by which cognitive recruiting systems can provide a large-scale (e.g., macro, overall) overview of an available position based on a title of the available position (i.e., a job title), wherein the large-scale overview goes beyond a surface level matching of strings and words.
Instead, cognitive recruiting systems may utilize term-based vector matching, including term frequency-inverse document frequency (tf-idf) (e.g., a numerical statistic to determine how important a word is in a given document), to account for the inherent semantic structure of the title of the available position (i.e., the job title), which may be augmented with contextual knowledge of elements of the job title, in order to determine a similarity score of two job titles based on information included in one or more connected knowledge bases.
According to at least one embodiment, in a job recruitment scenario, each component of a candidate's curriculum vitae (CV), which includes, but is not limited to including, skills, education, work history and/or job title, may be given a score based on the component's similarity to a corresponding component of a job requisition and/or job description. In this case, the candidate's skills may be matched to any required skills of a job while the candidate's experience and/or job history may be factored in determining an appropriate job level.
According to at least one embodiment, a semantic similarity of two job titles may be determined (e.g., in instances where no other metadata is available and/or metadata is limited) in order to contribute to an overall job matching score.
According to at least one embodiment, a similarity score may be derived by first pre-processing and normalizing inputted job titles, and by then decomposing and reconstructing semantic elements within the inputted job titles. A job title may be a combination of semantic elements, including but not limited to a job role, a job level, and/or a job domain or functional area.
According to at least one embodiment, a similarity between job title elements may be computed using general purpose and domain-specific word embedding models and a similarity score may be assigned for each semantic element of the job titles.
Referring to
The client computer 102 may communicate with the server computer 112 via the communications network 116. The communications network 116 may include connections, such as wire, wireless communication links, or fiber optic cables. As will be discussed with reference to
According to the present embodiment, a user using a client computer 102 or a server computer 112 may use the semantic matching program 110a, 110b (respectively) to utilize term-based vector matching to account for the inherent semantic structure of a title (i.e., a job title), which may be augmented with contextual knowledge of elements of the title, in order to determine a similarity score of two titles based on information included in one or more connected knowledge bases. The semantic matching method is explained in more detail below with respect to
Referring now to
At 202, job title (i.e., title) pre-processing and normalization occurs. Once job titles are received for analysis (e.g., from a human user), the semantic matching program 110a, 110b may subject job titles to the following cleaning steps, which may be done in the following order: word tokenization (e.g., where a job title is segmented into single-word tokens); standard and domain-specific stop word removal (e.g., where informal words such as “on” and “in” are removed); lemmatization (e.g., where inflection endings are removed and words are converted to their dictionary form); punctuation removal (e.g., punctuation including, but not limited to, periods and question marks are removed from the title); geographic data removal (e.g., removal of cities, states, and/or countries, among other geographic location data); and removal of numeric-only terms.
Word tokenization may be a process by which a text string (i.e., the job title) is split into the individual words (e.g., tokens) that comprise the text string so that natural language processing may done on each token.
Standard and domain-specific stop word removal may be done by comparing each token (e.g., each individual word comprising the text string) to a standard and a domain-specific stop word list which may be preprogrammed within the semantic matching program 110a, 110b. The standard and domain-specific stop word lists may also be dynamic and modifiable by users of the semantic matching program 110a, 110b to reflect ever-changing industries to which the lists pertain.
Lemmatization may be done by converting words to their dictionary form, based on one or more dictionaries utilized by the semantic matching program 110a, 110b. The use of one or more dictionaries by the semantic matching program 110a, 110b may be at the discretion of a user of the semantic matching program 110a, 110b, and may include both standard and/or domain-specific dictionaries.
Punctuation, geographic data and numeric terms may be determined based on connected databases defining punctuation, geographic data, and numeric terms, respectively, which may each be dynamic and modifiable by a user of the semantic matching program 110a, 110b.
After pre-processing of the job title, the job title may be augmented by adding contextual data which may include word synonyms derived from existing pre-curated dictionaries (e.g., wordnet and/or abbreviation dictionaries) and by expanding various abbreviated terms to their unabbreviated forms, with all abbreviations in the job title being resolved based on a pre-curated dictionary (e.g., CTO may become Chief Technology Officer and QA may become Quality Assurance). The pre-curated dictionary may be dynamic and modifiable by a user of the semantic matching program 110a, 110b.
For example, two job titles, “Principal Data Engineer” and “Senior Machine Learning Engineer” may be compared by the semantic matching program 110a, 110b to determine a degree of semantic matching.
At 204, semantic decomposition (i.e., deconstruction) of the job title is done. A job title may be a combination of semantic elements, including but not limited to a job role (e.g., a manager, an engineer, or an analyst, etc.), a job level (e.g., junior, senior, lead, etc.), and/or a job domain or functional area (e.g., software, energy, etc.). Based on this categorization, a job title may have a minimum of 1 and a maximum of 3 semantic elements. A job title that includes a job role without a job level or a job domain may be a valid job title (e.g., there may be instances where a job title is comprised solely of a job role), however, a job level or job domain without a job role may not be a valid job title. For example, the job title “software engineering” does not contain a job role or a job level and may not be considered a valid job title because no details on a desired applicant may be surmised from the job domain alone. However, “software engineer” or “software engineering manager” both contain a job role (e.g., engineer and manager, respectively) and may be considered valid job titles since a desired applicant may be determined from the job title.
The semantic matching program 110a, 110b may determine whether two or more job titles are semantically similar based on a comparison of the deconstructed elements of the job title (e.g., by a comparison of the individual tokens, as described previously with respect to step 202 above). If, for example, two jobs are semantically similar, then the two jobs may be similar in both job role and/or job domain or functional area.
A matching of all three semantic elements of the job title may result in a match score which is close to 1, which may imply that the semantic relatedness of the job titles is higher (e.g., where a match score is a range from 0-1 with 1 being a perfect or equal match).
Once a job title is pre-processed and normalized, as described previously with respect to step 202 above, then each token (i.e., word) of the job title may be compared and may be classified into a level, a role, and/or a domain. Terms corresponding to the level, the role, and the domain may be defined within dynamic, pre-curated dictionaries stored within the semantic matching program 110a, 110b.
For example, a semantic decomposition may be achieved using a pre-curated job role dictionary, job domain dictionary, and/or a job level dictionary. The job role dictionary and the job level dictionary may be based on a job taxonomy (e.g., including information corresponding to job content and industries and/or sectors) which may be curated using statistical and textual mining techniques on a large domain-specific dataset such as the CareerBuilder™ dataset (CareerBuilder is a trademark or registered trademark of CareerBuilder, LLC in the United States, and/or other countries).
Each of the connected dictionaries may be dynamic in nature, permitting users of the semantic matching program 110a, 110b to modify and/or extend the dictionaries as jobs and industries change and/or as new jobs emerge.
Continuing with the above example, semantic decomposition of the job titles may be done. The semantic elements of “Principal Data Engineer” may be as follows: job role: engineer; job domain: data; job level: principal. The semantic elements of “Senior Machine Learning Engineer” may be as follows: job role: engineer; job domain: machine learning; job level: senior.
At 206, Word2vec training is done. The Word2vec models, which may be shallow neural networks (e.g., a type of machine learning model often used in natural language processing applications) used by the semantic matching program 110a, 110b, may be trained following the pre-processing steps and hyper parameter tuning on short text and job related data. Model training may be done using either generic data (e.g., data covering all domains) such as that found on Wikipedia® (Wikipedia is a registered trademark of the Wikimedia Foundation, Inc. in the United States and/or other countries) or domain-specific data such as that found on CareerBuilder™, or a combination of the two.
Word2vec models may be trained using either a continuous bag of words (CBOW) approach or a skip-gram approach. For short text matching, a CBOW approach may be more accurate than a skip gram approach. Parameters such as vector size and window size may be configurable and may be adjusted depending on the application context. In this case, a vector size of 300 may be a recommended default, however, the parameter may be configurable. A window size of 3-5 may also be a recommended default based on the nature of a job title, which may not be considered a long sentence. Thus, it may be expected that a window size of 3-5 may capture the semantics of the present case sufficiently.
This may result in a matrix of size m*n, where m may be a size of the vocabulary (e.g., a number of unique words existing in the training data) and n may be a dimension parameter for the model with each word being a row with 300 additional values as its context. The context window may be a parameter used to train the model, with m being a number of words in the corpus used to build the model and n being a dimension vector for each word.
At 208, contextual word representation of the job title is done. These representations may be derived from Word2vec models which may have been previously trained. Similarity between semantic elements of a job title may be determined using general purpose and/or domain-specific word to vector language models (e.g., word embeddings) including Word2vec (Mikolov et al. 2013). These language models may be shown to capture semantic relationships between words by means of a cooccurrence probability which may be relevant to a lexical interpretation. Word to vector models may be shallow neural network-based word representation techniques which may capture word semantics within a word context.
However, most of the general purpose word embedding models (e.g., GloVe, Jeffrey Pennington et al. 2014; and/or Wikipedia®-based (Wikipedia is a registered trademark of the Wikimedia Foundation, Inc. in the United States and/or other countries) Word2vec models, etc.) may be less effective for technical and domain-specific terms and may fail to capture the intended semantics due to a lack of statistical cues in the underlying training corpus. Therefore, a domain-specific model may be used in combination with the general purpose model.
A domain-specific word embedding model may use job related content (e.g., historical job postings) to capture semantics of words in the relevant domain-specific context. The historical job postings may be mined from connected datasets, including CareerBuilder™, as described previously with respect to step 204 above.
This may be further refined based on an industry of the job, which may be determined based on the semantic decomposition of the job title, as described previously with respect to step 204 above, and a target model may be adjusted before computing a similarity score (the similarity score will be discussed in more detail with respect to step 210 below). For example, in a domain-specific word embedding model scenario, a finance job may rely on a model trained on finance, banking, and/or insurance job postings instead of models trained on information technology jobs.
Specifically, the target model may be based on a domain of a job posting. For example, if a job posting relates to banking and/or finance, historical job postings used to build the target model may also relate to banking and/or finance. Word embeddings may be extracted from these adjusted models such that the domain semantic may be captured more reliably than a model trained using job data which spans multiple industries.
Domain-specific word embedding models may include, but are not limited to including, global reference models (e.g., using Wikipedia® as a corpus (Wikipedia is a registered trademark of the Wikimedia Foundation, Inc. in the United States and/or other countries)), job models (e.g., based on job postings), and/or job domain models (e.g., using job domains or functional areas).
At 210, a similarity score is computed. The semantic matching program 110a, 110b may compute three different similarity scores; a job role similarity score, a job domain similarity score, and a job level similarity score. The job role similarity score and the job domain similarity score may be computed using a cosine distance between the averaged word vectors, which may be an accepted distance measure to accurately capture the word semantics. To compute the job level similarity score, the job level terms may be extracted from the job title and classified into four groups (e.g., “junior/entry,” “mid/intermediate,” “senior/experienced,” and “executive/management”) and a similarity score may thereafter be assigned. This categorization technique may be done to account for an inherent difficulty in capturing the true meaning of a job level and thus matching two or more job levels. A job level may be based on organizational hierarchies and may not be standardized based on industry. Therefore, by extracting terms indicative of a job level and by categorizing the job levels, a match score may be computed.
Once similarity scores are computed for the job role and the job domain, the semantic matching program 110a, 110b may use the computed scores to create a combined match score. The combined match score may not be a simple average but may be a weighted average, as will be described in more detail with respect to step 212 below, with the weight being determined based on the informativeness (i.e., specificity) of the job role.
To give weight to each job role, a global role relevance score may be computed by assessing the job role's usage statistics in the larger text corpus. For example, the term “manager” may have a wider usage in any text corpus than the term “recruiter” or “dentist.” Hence, when the job role is manager, it may be considered less informative as compared to an instance where the job role is dentist. The weight of a job role may be determined in multiple ways; however, the present technique may be a simple yet effective means by which to calculate the job role weight.
Based on the above, if the job role is generic, then the weight of the job role similarity score reduces as compared to the weight of the job domain similarity score.
Continuing with the above example, a similarity score of the job titles may be computed. The role similarity score (e.g., comparing “engineer” to “engineer”) may be determined to be 1.0 based on a cosine similarity of the two vectors (e.g., the vectors here are the same as the words are identical). The domain similarity score (e.g., comparing “data” and “machine learning”) may be determined to be 0.63 based on a cosine similarity of the two vectors. The level similarity score may be determined to be 1.0 based on contextual data (e.g., an analysis of synonyms of the determined job level reveals that “principal” and “senior” fall under the same category).
At 212, differential weighted scores are applied to job elements and a final match assessment is made. Different score weights may be used depending on a context of the job posting. If the job roles are generic (e.g., manager, engineer, etc.) then the job domain score may achieve a higher weight than the job role score. Alternatively, if the roles are specific (e.g., dentist, recruiter, etc.) then the score weights may be reversed. The differential weighed scores may be based on a word entropy (i.e., informativeness) or on term frequency-inverse document frequency (tf-idf) score. As described previously, the combined match score may not be a simple average but may be a weighted average, with the weight being determined based on the informativeness (i.e., specificity) of the job role. For example, if the job role is generic, then the weight of the job role similarity score reduces as compared to the weight of the job domain similarity score.
A final match assessment may be made based on the generated match score, as described above. For example, based on the match score, a match category of high, medium, or low may be presented to a human user of the semantic matching program 110a, 110b indicating a degree of similarity between two job titles. The match categories may be configurable by a human user of the semantic matching program 110a, 110b and may be based on intuitive decision making. For example, a match category of high may have an associated match score of above 0.7, a match category of medium may have an associated match score of 0.4-0.7, and a match category of low may have an associated match score of less than 0.4. These configured categories may be configured and influenced by the relevant domain of the application, and may not be predetermined solely by the semantic matching program 110a, 110b.
The final match assessment may also detail how the match category was determined. For example, the final match assessment might indicate that despite two job titles being dissimilar (e.g., “Executive Vice President, Cloud” and “Senior Cloud Engineer”), the semantic matching program 110a, 110b has realized (e.g., based on the functionality described above) that the functions of the two jobs (i.e., the job roles) are the same. Based on this, the semantic matching program 110a, 110b generates a match score which falls within a match category of medium. Accordingly, the semantic matching program 110a, 110b returns to the user a final match assessment detailing the generated match scores and supporting comparison reasoning (e.g., the semantic matching program 110a, 110b outputs to the user, along with the final match category, the portions of text that are determined to be semantically similar). A format of the displayed text (e.g., a level of detail to include in the user's output) may be modified by a user of the semantic matching program 110a, 110b.
Referring now to
Semantic decomposition of the job title may then be done, as described previously with respect to step 204 above, and the job role 306, job domain 308, and job level 310 may be extracted from the Job Titles A and B. A job role similarity score 312, a job domain similarity score 314 and a job level similarity score 316 may thereafter be computed as described previously with respect to step 210 above.
A match score 318 may be computed based on the job role similarity score 312, job domain similarity score 314 and job level similarity score 316, described above.
It may be appreciated that
Data processing system 902, 904 is representative of any electronic device capable of executing machine-readable program instructions. Data processing system 902, 904 may be representative of a smart phone, a computer system, PDA, or other electronic devices. Examples of computing systems, environments, and/or configurations that may represented by data processing system 902, 904 include, but are not limited to, personal computer systems, server computer systems, thin clients, thick clients, hand-held or laptop devices, multiprocessor systems, microprocessor-based systems, network PCs, minicomputer systems, and distributed cloud computing environments that include any of the above systems or devices.
User client computer 102 and network server 112 may include respective sets of internal components 902a, b and external components 904a, b illustrated in
Each set of internal components 902a, b also includes a R/W drive or interface 918 to read from and write to one or more portable computer-readable tangible storage devices 920 such as a CD-ROM, DVD, memory stick, magnetic tape, magnetic disk, optical disk or semiconductor storage device. A software program, such as the software program 108 and the semantic matching program 110a and 110b can be stored on one or more of the respective portable computer-readable tangible storage devices 920, read via the respective R/W drive or interface 918 and loaded into the respective hard drive 916.
Each set of internal components 902a, b may also include network adapters (or switch port cards) or interfaces 922 such as a TCP/IP adapter cards, wireless wi-fi interface cards, or 3G or 4G wireless interface cards or other wired or wireless communication links. The software program 108 and the semantic matching program 110a in client computer 102 and the semantic matching program 110b in network server computer 112 can be downloaded from an external computer (e.g., server) via a network (for example, the Internet, a local area network or other, wide area network) and respective network adapters or interfaces 922. From the network adapters (or switch port adaptors) or interfaces 922, the software program 108 and the semantic matching program 110a in client computer 102 and the semantic matching program 110b in network server computer 112 are loaded into the respective hard drive 916. The network may comprise copper wires, optical fibers, wireless transmission, routers, firewalls, switches, gateway computers and/or edge servers.
Each of the sets of external components 904a, b can include a computer display monitor 924, a keyboard 926, and a computer mouse 928. External components 904a, b can also include touch screens, virtual keyboards, touch pads, pointing devices, and other human interface devices. Each of the sets of internal components 902a, b also includes device drivers 930 to interface to computer display monitor 924, keyboard 926 and computer mouse 928. The device drivers 930, R/W drive or interface 918 and network adapter or interface 922 comprise hardware and software (stored in storage device 916 and/or ROM 910).
It is understood in advance that although this disclosure includes a detailed description on cloud computing, implementation of the teachings recited herein are not limited to a cloud computing environment. Rather, embodiments of the present invention are capable of being implemented in conjunction with any other type of computing environment now known or later developed.
Cloud computing is a model of service delivery for enabling convenient, on-demand network access to a shared pool of configurable computing resources (e.g. networks, network bandwidth, servers, processing, memory, storage, applications, virtual machines, and services) that can be rapidly provisioned and released with minimal management effort or interaction with a provider of the service. This cloud model may include at least five characteristics, at least three service models, and at least four deployment models.
Characteristics are as follows:
On-demand self-service: a cloud consumer can unilaterally provision computing capabilities, such as server time and network storage, as needed automatically without requiring human interaction with the service's provider.
Broad network access: capabilities are available over a network and accessed through standard mechanisms that promote use by heterogeneous thin or thick client platforms (e.g., mobile phones, laptops, and PDAs).
Resource pooling: the provider's computing resources are pooled to serve multiple consumers using a multi-tenant model, with different physical and virtual resources dynamically assigned and reassigned according to demand. There is a sense of location independence in that the consumer generally has no control or knowledge over the exact location of the provided resources but may be able to specify location at a higher level of abstraction (e.g., country, state, or datacenter).
Rapid elasticity: capabilities can be rapidly and elastically provisioned, in some cases automatically, to quickly scale out and rapidly released to quickly scale in. To the consumer, the capabilities available for provisioning often appear to be unlimited and can be purchased in any quantity at any time.
Measured service: cloud systems automatically control and optimize resource use by leveraging a metering capability at some level of abstraction appropriate to the type of service (e.g., storage, processing, bandwidth, and active user accounts). Resource usage can be monitored, controlled, and reported providing transparency for both the provider and consumer of the utilized service.
Service Models are as follows:
Software as a Service (SaaS): the capability provided to the consumer is to use the provider's applications running on a cloud infrastructure. The applications are accessible from various client devices through a thin client interface such as a web browser (e.g., web-based e-mail). The consumer does not manage or control the underlying cloud infrastructure including network, servers, operating systems, storage, or even individual application capabilities, with the possible exception of limited user-specific application configuration settings.
Platform as a Service (PaaS): the capability provided to the consumer is to deploy onto the cloud infrastructure consumer-created or acquired applications created using programming languages and tools supported by the provider. The consumer does not manage or control the underlying cloud infrastructure including networks, servers, operating systems, or storage, but has control over the deployed applications and possibly application hosting environment configurations.
Infrastructure as a Service (IaaS): the capability provided to the consumer is to provision processing, storage, networks, and other fundamental computing resources where the consumer is able to deploy and run arbitrary software, which can include operating systems and applications. The consumer does not manage or control the underlying cloud infrastructure but has control over operating systems, storage, deployed applications, and possibly limited control of select networking components (e.g., host firewalls).
Deployment Models are as follows:
Private cloud: the cloud infrastructure is operated solely for an organization. It may be managed by the organization or a third party and may exist on-premises or off-premises.
Community cloud: the cloud infrastructure is shared by several organizations and supports a specific community that has shared concerns (e.g., mission, security requirements, policy, and compliance considerations). It may be managed by the organizations or a third party and may exist on-premises or off-premises.
Public cloud: the cloud infrastructure is made available to the general public or a large industry group and is owned by an organization selling cloud services.
Hybrid cloud: the cloud infrastructure is a composition of two or more clouds (private, community, or public) that remain unique entities but are bound together by standardized or proprietary technology that enables data and application portability (e.g., cloud bursting for load-balancing between clouds).
A cloud computing environment is service oriented with a focus on statelessness, low coupling, modularity, and semantic interoperability. At the heart of cloud computing is an infrastructure comprising a network of interconnected nodes.
Referring now to
Referring now to
Hardware and software layer 1102 includes hardware and software components. Examples of hardware components include: mainframes 1104; RISC (Reduced Instruction Set Computer) architecture based servers 1106; servers 1108; blade servers 1110; storage devices 1112; and networks and networking components 1114. In some embodiments, software components include network application server software 1116 and database software 1118.
Virtualization layer 1120 provides an abstraction layer from which the following examples of virtual entities may be provided: virtual servers 1122; virtual storage 1124; virtual networks 1126, including virtual private networks; virtual applications and operating systems 1128; and virtual clients 1130.
In one example, management layer 1132 may provide the functions described below. Resource provisioning 1134 provides dynamic procurement of computing resources and other resources that are utilized to perform tasks within the cloud computing environment. Metering and Pricing 1136 provide cost tracking as resources are utilized within the cloud computing environment, and billing or invoicing for consumption of these resources. In one example, these resources may comprise application software licenses. Security provides identity verification for cloud consumers and tasks, as well as protection for data and other resources. User portal 1138 provides access to the cloud computing environment for consumers and system administrators. Service level management 1140 provides cloud computing resource allocation and management such that required service levels are met. Service Level Agreement (SLA) planning and fulfillment 1142 provide pre-arrangement for, and procurement of, cloud computing resources for which a future requirement is anticipated in accordance with an SLA.
Workloads layer 1144 provides examples of functionality for which the cloud computing environment may be utilized. Examples of workloads and functions which may be provided from this layer include: mapping and navigation 1146; software development and lifecycle management 1148; virtual classroom education delivery 1150; data analytics processing 1152; transaction processing 1154; and semantic matching 1156. A semantic matching program 110a, 110b provides a way to utilize term-based vector matching to account for the inherent semantic structure of a title (i.e., a job title), which may be augmented with contextual knowledge of elements of the title, in order to determine a similarity score of two titles based on information included in one or more connected knowledge bases.
The descriptions of the various embodiments of the present invention have been presented for purposes of illustration, but are not intended to be exhaustive or limited to the embodiments disclosed. Many modifications and variations will be apparent to those of ordinary skill in the art without departing from the scope of the described embodiments. The terminology used herein was chosen to best explain the principles of the embodiments, the practical application or technical improvement over technologies found in the marketplace, or to enable others of ordinary skill in the art to understand the embodiments disclosed herein.
