Semi-automatic Web Service Classification Using Machine Learning

Efficient and fast service discovery becomes an extremely challenging task due to the proliferation and availability of functionally-similar web services. Service classification or service grouping is a popular and widely applied technique to classify services into several groups according to similarity, in order to ease up and expedite the discovery process. Existing research on web service classification uses several techniques, approaches and frameworks for web service classification. This study focused on a hybrid service classification approach based on a combination of ontology generation and machine learning algorithm, in order to gain more speed and accuracy during the classification process. Ontology generation is applied to capture the similarity between complicated words. Then, two machine learning classification algorithms, namely, Support Vector Machines (SVMs) and Naive Bayes (NB), were applied for classifying services according to their functionality. The experimental results showed significant improvement in terms of accuracy, precision and recall. The hybrid approach of ontology generation and NB algorithm achieved an accuracy of 94.50%, a precision of 93.00% and a recall of 95.00%. Therefore, a hybrid approach of ontology generation and NB has the potential to pave the way for efficient and accurate service classification and discovery.

Co-attentive representation learning for web services classification

In software technology, over the diversified environment, services can be rendered using an innovative mechanism of a novel paradigm called web services. In a business environment, rapid changes and requirements from various customers can be adapted using this service. For service management and discovery, the classification of Web services having the same functions is an efficient technique. However, there will be short lengthened Web services functional description documents, having less information, and sparse features. This makes difficulties in modelling short text in various topic models and leads to make an effect in the classification of Web services. A Mixed Wide and PSO-Bi-LSTM-CNN model (MW-PSO-Bi-LSTM-CNN) is proposed in this work for solving this issue. In this technique, the Web service category‟s breadth prediction is performed by combining Web services description document‟s discrete features, which exploits the wide learning model. In the next stage, the PSO-Bi-LSTM-CNN model is used for mining Web services description document word‟s context information and word order, for performing the Web service category‟s depth prediction. Here, particle swarm optimization (PSO) is integrated with the Bi-LSTM-CNN network for computing various hyper-parameters in an automatic manner. In third stage, Web service categories, results of depth, and breadth prediction are integrated using a linear regression model as final service classification result. At last, MW-PSO-Bi-LSTM-CNN, Wide&Bi-LSTM, and Wide&Deep web service classification techniques are compared and a better result with respect to web service classification accuracy is obtained using the proposed technique as shown in experimental results.

Service oriented architecture is a glue that allows web applications to work in collaboration. It has become a driving force for the service-oriented computing (SOC) paradigm. In heterogeneous environments the SOC paradigm uses web services as the basic building block to support low costs as well as easy and rapid composition of distributed applications. A web service exposes its interfaces using the Web Service Description Language (WSDL). A central repository called universal description, discovery and integration (UDDI) is used by service providers to publish and register their web services. UDDI registries are used by web service consumers to locate the web services they require and metadata associated with them. Manually analyzing WSDL documents is the best approach, but also most expensive. Work has been done on employing various approaches to automate the classification of web services. However, previous research has focused on using a single technique for classification. This research paper focuses on the classification of web services using a majority vote based classifier ensemble technique. The ensemble model overcomes the limitations of conventional techniques by employing the ensemble of three heterogeneous classifiers: Naive Bayes, decision tree (J48), and Support Vector Machines. We applied tenfold cross-validation to test the efficiency of the model on a publicly available dataset consisting of 3738 real world web services categorized into 5 fields, which yielded an average accuracy of 92 %. The high accuracy is owed to two main factors, i.e., enhanced pre-processing with focused feature selection, and majority based ensemble classification.

With the rapid growth and adoption of web-based technologies, more and more service providers offer various services on the web. This can facilitate the organizations to develop business applications using web services composition over the Internet according to their business requirements. The primary requirement for developing composite web service is to discover the most relevant web service for realizing each function of the required composite web service. Finding the most relevant web service is still a challenging task. In this paper, we propose web service discovery approach that incorporates web services clustering. Clustering technique combines latent Dirichlet allocation(LDA) and k-Medoids and reduces search space for web service discovery. Web service discovery uses a match-making algorithm that utilizes WordNet database and Hungarian method to compute the semantic similarity score for a web service based on the requested query. Experimental results show that the proposed approach performs better than LDA-based approach.

Automated service classification is the foundation for service discovery and service composition. Currently, many existing methods extracting features from functional description documents suffer the problem of data sparsity. However, beside functional description documents, the Web API ecosystem has accumulated a wealth of information that can be used to improve the accuracy of Web service (API) classification. At the moment, there is an absence of a unified way to combine functional description documents with other sources of information (e.g., attributes, interactions and external knowledge) accumulated in the Web API ecosystem for API classification. To address this issue, we present a dual-GCN framework that can effectively suppress the noise propagation of textual contents by distinguishing functional description documents and other sources of information (specifically Mashup-API co-invocation patterns by default in this paper) for API classification. This framework is extensible with the ability to include different sources of information accumulated in the Web API ecosystem. Comprehensive experiments on a real-world public dataset demonstrate that our proposed method can outperform various representative methods for API classification.

The transition of the World Wide Web from a paradigm of static Web pages to one of dynamic Web services raises a new and challenging problem of locating desired web services. With the expected growth of the number of Web services available on the web, the need for mechanisms that enable the automatic categorization to organize this vast amount of data, becomes important. In this paper we propose Tensor space model for data representation and Rough Set based approach for the classification of Web services. The proposed tensor space model captures the information from internal structure of WSDL documents along with the corresponding text content. Rough sets are used here to combine information of the individual tensor components for providing classification results. Two step improvement on the existing classification results of web services has been shown here. In the first step we achieve better classification results over existing, by using proposed tensor space model. In the second step further improvement of the results has been obtained by using Rough set based ensemble classifier.

A web service is a software system that supports interoperable machine to machine interaction over the network. It can be any kind of service provided on the web that can exchange data between applications. In simple terms, it is any program that is callable by another program across the web using standard protocols. Web services are special because they are independent of the platform, programming language, and model design. In current web applications, more businesses are gradually publishing their business as services over the web. This growing number of web services available within an organization and on the Web raises a new and challenging search problem: locating desired web services. Searching for web services with conventional web search engines is insufficient in this context. In this paper, a latent semantic web service discovery and composition framework, addressing both discovery and composition of web services, is introduced. The framework provides a graphical visualization interface that can facilitate users in finding the desired web service beyond the general keyword search and provide composition of complex services if the service does not exist. A novel approach based on an information retrieval technique known as latent semantic analysis is applied to a large web service descriptions data set collection. By using co-occurrence patterns across the entire data collection, and then using those patterns to infer semantic relationships between documents, relevant results in a ranked order can be retrieved. Moreover, a back to front algorithm is applied to the composition search interface in order to compose complex web services in full or partial terms from pre-existing web services when the requested service does not exist. An experimental study conducted on a collection of 2525 publicity available web service shows improved performance to the technique applied in obtaining much better search results.

Shared-Repository based Approach for Storing and Discovering web Services

Web services (WS) is called composite or compound when its execution involves interactions with other WS to utilize their features. The service providers published the web services through the internet as independent software components that are fulfilling the requirements of customer request. Clustering is more necessary for efficient web service discovery and web service composition processes. Clustering process groups the similar type of web services. In this paper, efficient clustering methods such as k-means clustering, Hierarchical agglomerative clustering and Balanced Iterative Reducing and Clustering using Hierarchies (BIRCH) clustering are introduced for web service clustering. The k-means clustering is a kind of partitional clustering where the web pages are divided into subsets with no hierarchy defined over them and the hierarchical agglomerative clustering is a type of hierarchical clustering where the web pages are arranged in tree structure in which leaves represents the data points and nodes denotes the clusters. BIRCH is an integrated hierarchical clustering algorithm uses the clustering features and cluster feature tree for general cluster description. Based on these clustering methods, web pages are clustered which are used for web service discovery and web service composition. The experiments are conducted on number of web services and the efficiency is evaluated in terms of accuracy, precision, recall and run time.

Web services are the new paradigm for distributed computing. They offer interoperability of application and integration of large scale distributed systems. Amongst the web service standards, service registry is crucial to the overall utility of web service technology. However, the web services discovery mechanism pertaining with the UDDI [1] registry is inefficient because it only supports keyword-based and category-based matching, leading to irrelevant matches. Therefore, the provision of the scalable, flexible and robust web services publication and discovery system is crucial for the overall utility of web services. In this paper, we develop a scalable, robust and efficient infrastructure for web services publication and discovery amongst domain-based registries. More importantly, a novel algorithm is provided to classify web services into domain-based registries based on their semantic description. Experimental results show that our approach outperforms a centralized approach in terms of service discovery time.

Web service (WS) discovery is a prerequisite for achieving WS composition and orchestration. Although a lot of research has been conducted on the functional discovery of WSs, the proposed techniques fall short when faced with the foreseen increase in the number of (potentially functionally-equivalent) WSs. The above situation can be resolved with the addition of non-functional (quality of service (QoS)) discovery mechanisms to WS discovery engines. QoS-based WS matchmaking algorithms have been devised for this reason. However, they are either slow - as they are based on ontology reasoners - or produce inaccurate results. Inaccuracy is caused both by the syntactic matching of QoS concepts and by wrong matchmaking metrics. In this paper, we present two constraint programming (CP) QoS-based WS discovery algorithms for unary constrained WS specifications that produce accurate results with good performance. We also evaluate these algorithms on matchmaking time, precision and recall in different settings in order to demonstrate their efficiency and accuracy.

Technologies for Web services facilitate the creation of business process solutions in an efficient, standard way. However, the automation of process integration with Web service technologies requires the automation of discovery and composition of Web services. In this paper, we focus on two problems of the Web service-based business process integration: (1) the discovery of Web services based on the capabilities and properties of published services, and (2) the composition of business processes based on the business requirements of submitted requests. We propose a solution to these problems, which comprises multiple matching algorithms, a micro-level matching algorithm, which matches the capabilities of services with activities in a process request, and macro-level matching algorithms, which are used to compose a business process by identifying services that satisfy the business requirements and constraints of the request. The solution from the macro-level matching algorithms is optimal in terms of meeting a certain business objective, e.g., minimizing the cost or execution time, or maximizing the total utility value of business properties of interest. Numerical examples are illustrated to show how to select the best Web service candidate for a chosen business process through the use of proposed macro-level matching algorithms. Furthermore, we show how existing Web service standards, UDDI and BPEL4WS, can be used and extended to specify the capabilities of services and the business requirements of requests, respectively.

Nowadays, with the advancement in mobile technologies, the use of web services has seen an explosion in interest for mobile computing environment. However, the heavyweight nature of conventional web services to be deployed on mobile devices brings new challenges in mobile computing in the coming future. Web services technology and mobile computing domains are converging at their intersection which leads to Mobile Web Services (MWS) which enables the service access in users’ mobile device. As the numbers of web services are increasing dramatically, the web service discovery becomes more important to discover the usable web services in effective and efficient manner. This paper discusses an overview of mobile challenges in MWS recent research, and a summary of the issues. Moreover, current approaches of Mobile Web Services Discovery are classified and these approaches are compared to some criteria. The results of this study will help researchers to deliver more applicable solutions with the most appropriate approaches for Mobile Web Services Discovery.KeywordsWeb ServicesMobile Web Serviceweb service discoverymobile computingmobile web service discovery

Semantic Web Services are the key technology providing services for the users' convenience in the semantic web environment. Many companies in various fields are researching and developing languages for constructing Semantic Web Services such as DAML-S, WSDL, X-LANG and BPEL4WS. DAML-S is a method that accesses the existing Web Service method from a semantic web environment. However, the current Semantic Web Services Discovery System does not provide sufficient processing of quality factors (performance, accessibility, availability, reliability and transaction) nor reliable and accurate service results desired by the user through a UDDi search method. The present study analyzes the disadvantages of the current web services and suggests a Semantic Web Services Discovery System based on QoS applying semantic web technologies as a solution to the problems. The suggested system provides a UDDi and DAML-S based discovery engine to allow efficient web service discovery and composition.