User Input Query Research Articles

Improving speculative query execution support by the use of the hypergraph representation

Speculative support for query execution in Relational Database Management Systems (RDBMSs) based on a graph-analytic approach is the subject of this paper. Optimized speculative queries are defined by analyzing graph representations of a stream of input queries arriving to a RDBMS. The analysis is permanently performed based on a Speculation Window moving on the RDBMS input query queue. The proposed optimization concerns both the definition of the speculative queries to be executed and the methods of speculative query results assignment to incoming input queries. The paper shows how to design RDBMS query execution speculative support using two kinds of the query graph representations: the query multigraphs and the query hypergraphs. In particular, the paper proposes an improvement of the use of the speculative query results defined based on the query multigraph analysis by an additional assignment to queries in an extension of the Speculation Window in the input query queue. The additional assignment is based on the analysis of the speculative and user input queries represented jointly in a common hypergraph. Two approaches to the proposed speculative results improvement are discussed which increase the speculative result life time. The properties of the speculative query execution support based on the mentioned query graph representations and algorithmic approaches have been assessed by the experiments with the simulated execution of the query testbed sets based on the popular TPC-H benchmark.

Classification on Unsupervised Deep Hashing With Pseudo Labels Using Support Vector Machine for Scalable Image Retrieval

The content-based image retrieval (CBIR) method operates on the low-level visual features of the user input query object, which makes it difficult for users to formulate the query and also does not provide adequate retrieval results. In the past, image annotation was suggested as the best possible framework for CBIR, which works on automatically signing keywords to images that support image retrieval. The recent successes of deep learning techniques, especially Convolutional Neural Networks (CNN), in solving computer vision applications have inspired me to work on this paper to solve the problem of CBIR using a dataset of annotated images

DeepSuggest: Using Neural Networks to Suggest Related Keywords for a Comprehensive Search of Clinical Notes

Abstract Objective A large amount of clinical data are stored in clinical notes that frequently contain spelling variations, typos, local practice-generated acronyms, synonyms, and informal words. Instead of relying on established but infrequently updated ontologies with keywords limited to formal language, we developed an artificial intelligence (AI) assistant (named “DeepSuggest”) that interactively offers suggestions to expand or pivot queries to help overcome these challenges. Methods We applied an unsupervised neural network (Word2Vec) to the clinical notes to build keyword contextual similarity matrix. With a user's input query, DeepSuggest generates a list of relevant keywords, including word variations (e.g., formal or informal forms, synonyms, abbreviations, and misspellings) and other relevant words (e.g., related diagnosis, medications, and procedures). Human intelligence is then used to further refine or pivot their query. Results DeepSuggest learns the semantic and linguistic relationships between the words from a large collection of local notes. Although DeepSuggest is only able to recall 0.54 of Systematized Nomenclature of Medicine (SNOMED) synonyms on average among the top 60 suggested terms, it covers the semantic relationship in our corpus for a larger number of raw concepts (6.3 million) than SNOMED ontology (24,921) and is able to retrieve terms that are not stored in existing ontologies. The precision for the top 60 suggested words averages at 0.72. Usability test resulted that DeepSuggest is able to achieve almost twice the recall on clinical notes compared with Epic (average of 5.6 notes retrieved by DeepSuggest compared with 2.6 by Epic). Conclusion DeepSuggest showed the ability to improve retrieval of relevant clinical notes when implemented on a local corpus by suggesting spelling variations, acronyms, and semantically related words. It is a promising tool in helping users to achieve a higher recall rate for clinical note searches and thus boosting productivity in clinical practice and research. DeepSuggest can supplement established ontologies for query expansion.

Abstract 187: PCaT: An integrated platform for the analysis of pediatric cancer genomic data

Abstract Pediatric cancer is the leading cause of disease related mortality in children, affecting approximately 15,000 new cases each year in the US. Compared with their adult counterparts, pediatric cancers harbor ten times less mutations. Their relatively quiet genomes dictate that most patients do not have alterations in clinically actionable targets. Transcriptomic changes reflect genetic and epigenetic perturbations, thus represent an important tool to study molecular mechanisms and therapeutic responses of pediatric cancers. However, the complexity of current multidimensional datasets poses a great challenge for researchers, especially bench scientists, to navigate these data. Here, we present a new, intuitive, and integrated tool, Pediatric Cancer Transcriptome Database (PCaT, http://www.pedtranscriptome.org), to help address this issue. PCAT focuses on transcriptome data, but integrates clinical, histopathological, genetic, and drug response data. Built under the Rshiny framework, it currently offers three function modules (single gene analysis, multi-gene analysis, preclinical testing), and is under active expansion. Each module enables pan-cancer analysis and cancer type specific analysis when appropriate, thus providing both a panoramic and focused view for user input queries. The single gene analysis module allows users to correlate gene expression with clinical, histological parameters and patient prognosis. Users can examine how copy number and mutation impact a gene's expression. Importantly, users can also identify genes and functional modules related to the query through co-expression analysis. Multiple gene analysis allows for visualization, clustering, functional enrichment analysis, pairwise correlation, and calculation of a signature score using single sample gene set enrichment analysis. A distinctive feature of PCAT is the integration of preclinical testing data for more than 60 patient derived xenograft models (PDXs) to nearly 80 FDA approved or experimental agents (preclinical module). This feature provides users with the opportunity to explore correlations between gene expression and drug responses across a diverse range of pediatric cancer types. Currently, PCAT includes multimodal datasets from Therapeutically Applicable Research to Generate Effective Treatments consortium (TARGET) and Pediatric Preclinical Testing Consortium (PPTC). In summary, PCAT is a new user-friendly online portal that offers an intuitive interface for users to explore multidimensional data, especially the transcriptome, of pediatric cancers. We are currently expanding the portal with more datasets and more functional modules, particularly gene fusions. PCAT is complementary to existing tools such as St Jude Cloud, R2 portal, and pedcbio, and offers an alternative route to navigating an increasing accumulation of pediatric cancer genomic data. Citation Format: Juechen Yang, Qilin Li, Xiaojing Wang, Peter Houghton, Siyuan Zheng. PCaT: An integrated platform for the analysis of pediatric cancer genomic data [abstract]. In: Proceedings of the Annual Meeting of the American Association for Cancer Research 2020; 2020 Apr 27-28 and Jun 22-24. Philadelphia (PA): AACR; Cancer Res 2020;80(16 Suppl):Abstract nr 187.

Statistics Based Optimization Technique for Query Processing

Semantic web consists of the data in the structure manner and query searching methods can access these structured data to provide effective search result. The query recommendation in the semantic web relevance is needed to be improved based on the user input query. Many existing methods are used to improve the query recommendation efficiency using the optimization technique such as Particle Swarm Optimization (PSO), Ant Colony Optimization (ACO). These methods involve in the use of many features which are selected from the user query. This in-turn increases the cost of a query in the semantic web. In this research, the query optimization was carried out by using the statistics method. The statistics based optimization method requires fewer features such as triple pattern and node priority etc., for finding the relevant results. The LUBM dataset contains the semantic queries and this dataset is used to measure the efficiency of the proposed Statistical based optimization method. The SPARQL queries are used to plot the query graph and triple scores are extracted from the graph. The cost value of the triple scores is measured and given as input to the proposed statistics method. The execution time of the statistics based optimization method for the query is 35 ms while the existing method has 48 ms.

Motion rank: applying page rank to motion data search

As the uses of motion capture data increase, the amount of available motion data on the web also grows. In this paper, we investigate a new method to retrieve and visualize motion data in a similar manner to Google Image Search. The main idea is to represent raw motion data into a series of short animated clip arts, called motion clip arts. Short animated clip arts can be quickly browsed and understood by people even though many of them appear at the same time on the screen. We first temporally segment the raw motion data files into short yet semantically meaningful motion segments. Then, we convert the motion segments into motion clip arts in a way that emphasizes the main motion and minimizes the data size for the efficient transmitting and processing on the web. When a user input query is received, our system first retrieves all the relevant motion clip arts by considering the input keywords and similarity between motions. Then, the retrieved results are re-ranked by our ranking algorithm developed based on the Google ImageRank algorithm. To prove the usability of our method, we build a web-based motion search system with the entire data collections of the CMU motion database. The experimental results show significant improvement, in terms of relevancy, in comparison with the simple keyword-based search interface.

ACSIR: ANOVA Cosine Similarity Image Recommendation in vertical search

In today’s world, online shopping is very attractive and grown exponentially due to revolution in digitization. It is a crucial demand to provide recommendation for all the search engine to identify users’ need. In this paper, we have proposed a ANOVA Cosine Similarity Image Recommendation (ACSIR) framework for vertical image search where text and visual features are integrated to fill the semantic gap. Visual synonyms of each term are computed using ANOVA p value by considering image visual features on text-based search. Expanded queries are generated for user input query, and text-based search is performed to get the initial result set. Pair-wise image cosine similarity is computed for recommendation of images. Experiments are conducted on product images crawled from domain-specific site. Experiment results show that the ACSIR outperforms iLike method by providing more relevant products to the user input query.

IR_URFS_VF: image recommendation with user relevance feedback session and visual features in vertical image search

In recent years, online shopping has grown exponentially and huge number of images are available online. Hence, it is necessary to recommend various product images to aid the user in effortless and efficient access to the desired products. In this paper, we present image recommendation framework with user relevance feedback session and visual features (IR_URFS_VF) to extract relevant images based on user inputs. User feedback is retrieved from image search history with clicked and un-clicked images. Image features are computed off-line and later used to find relevance between images. The relevance between images is determined by cosine similarity and are ranked based on clicked frequency and similarity score between images. Experiments results show that IR_URFS_VF outperforms CBIR method by providing more relevant ranked images to the user input query.

Query Categorization from Web Search Logs Using Machine Learning Algorithms

This paper presents a data-driven methodology to disambiguate a query by suggesting relevant subcategories within a specific domain. This is achieved by finding correlations between the user’s search history and the context of the current search keyword. We apply automatic categorization on each query to identify a list of categories which can describe the query given. To predict the categories of a user input query, we employed machine learning algorithms. We present the preliminary evaluation results and conclude with future work.

Image recommendation based on keyword relevance using absorbing Markov chain and image features

Image recommendation is an important feature of search engine, as tremendous amount of images are available online. It is necessary to retrieve relevant images to meet the user’s requirement. In this paper, we present an algorithm image recommendation with absorbing Markov chain (IRAbMC) to retrieve relevant images for a user’s input query. Images are ranked by calculating keyword relevance probability between annotated keywords from log and keywords of user input query. Keyword relevance is computed using absorbing Markov chain. Images are reranked using image visual features. Experimental results show that the IRAbMC algorithm outperforms Markovian semantic indexing (MSI) method with improved relevance score of retrieved ranked images.

Image Recommendation Based on ANOVA Cosine Similarity

Online shopping is very popular and has grown exponentially due to revolution in digitization. It is a fundamental requirement of all the search engines to provide recommendation to identify user preferences. In this paper, we have proposed an algorithm to recommend images based on ANOVA Cosine Similarity where text and visual features are integrated to fill the semantic gap. Visual synonyms of each term are computed using ANOVA p-value by considering image visual features on text-based search. Expanded queries are generated for user input query and text based search is performed to get initial result set. Pair-wise image cosine similarity is computed for recommendation of images. Experiments are conducted on product images crawled from domain specific site. Experiments results shows that the ACSIR outperforms iLike method by providing more relevant products to the user input query.

K-NEAREST NEIGHBOR CLASSIFICATION OVER SEMANTICALLY SECURE ENCRYPTED RELATIONAL DATA

Data Mining has wide applications in many areas such as banking, medicine, scientific research and among government agencies. Classification is one of the commonly used tasks in data mining applications. For the past decade, due to the rise of various privacy issues, many theoretical and practical solutions to the classification problem have been proposed under different security models. However, with the recent popularity of cloud computing, users now have the opportunity to outsource their data, in encrypted form, as well as the data mining tasks to the cloud. Since the data on the cloud is in encrypted form, existing privacy preserving classification techniques are not applicable. In this paper, we focus on solving the classification problem over encrypted data. In particular, we propose a secure k-NN classifier over encrypted data in the cloud. The proposed k-NN protocol protects the confidentiality of the data, user's input query, and data access patterns. To the best of our knowledge, our work is the first to develop a secure k-NN classifier over encrypted data under the semi-honest model. Also, we empirically analyze the efficiency of our solution through various experiments.

<italic>k</italic>-Nearest Neighbor Classification over Semantically Secure Encrypted Relational Data

Data Mining has wide applications in many areas such as banking, medicine, scientific research and among government agencies. Classification is one of the commonly used tasks in data mining applications. For the past decade, due to the rise of various privacy issues, many theoretical and practical solutions to the classification problem have been proposed under different security models. However, with the recent popularity of cloud computing, users now have the opportunity to outsource their data, in encrypted form, as well as the data mining tasks to the cloud. Since the data on the cloud is in encrypted form, existing privacy-preserving classification techniques are not applicable. In this paper, we focus on solving the classification problem over encrypted data. In particular, we propose a secure k-NN classifier over encrypted data in the cloud. The proposed protocol protects the confidentiality of data, privacy of user's input query, and hides the data access patterns. To the best of our knowledge, our work is the first to develop a secure k-NN classifier over encrypted data under the semi-honest model. Also, we empirically analyze the efficiency of our proposed protocol using a real-world dataset under different parameter settings.

Triple Patterns Extraction for Accessing Data on Ontology

Some web applications use the ontology to integrate multiple data sources because ontology-based data integration is the ideal solution to handle the semantic conflict. For accessing the data on ontology, ontology understanding query such as SPARQL is needed. However, end users enter the unstructured sentence (words, statements, etc.) as an input when they wanted to search the required information on the web. So, it is needed to extract the triplets (i.e. subjects, predicates and objects) from the input query to build the ontology browsing query SPARQL. Although there are many triplet extraction algorithms, either they can't fully define all triple patterns from the incoming query or they are time consuming process. The proposed algorithm presented in this paper can handle this triplet's incompleteness problem and the aim of this system is to extract the specific triplets from incoming query and to add the necessary information for supporting SPARQL query generating process in a time-saving manner. triple patterns from the user input query. Then the mediator automatically generates the SPARQL query to browse the ontology by using the triplets obtained from the triplet extraction algorithm. After the SPARQL query has been generated, the data included in the ontology can be retrieved and returned to the user. There are three main processing phases in ontology based data integration system such as ontology creation, ontology mapping and query service. Extracting the triplets from the user query is one of the most important roles of query service as describe in above. When the user query is submitted to the system, this query is needed to translate to the ontology understanding query SPARQL. The query language SPARQL has a graph-based structure and can be built by combining triple patterns extracted from the user input query (2). If the user input query is unstructured query, it will be more difficult to define whether which words are subjects or objects. The proposed algorithm is able to identify the subjects, predicates and objects in the unstructured query exactly. Currently there are many triplets extraction algorithm for supporting the triple patterns to assist the necessary application. They implement the triplet extraction process based on the parse tree generated by the parser. In this proposed algorithm, triple patterns included in the user input query are extracted with the help of domain specific ontology and machine readable dictionary WorldNet instead of using the parser. So, the processing time of this algorithm is lesser than other parser-based approaches. This paper is organized as follows. In Section II, this paper presents the related work. The system design detail is presented in Section III. It gives in Section IV the description of proposed triplet extraction algorithm. Section V demonstrates the sample query testing and results of proposed algorithm by comparing the result of other competitive algorithms. Section VI concludes with some remarks.

Self-improvement of voice interface with user-input spoken query at early stage of commercialization

This paper concerns the self-improvement of voice interface by using acoustic model re-training with user-input spoken query at early stage of commercialization, when the conventional confidence measure-based acoustic model re-training is not reliable. This paper analyzes error patterns in user-input spoken queries, categorizes these error patterns, defines a quantitative measurement for each category of error patterns and proposes a filter-based approach over this quantitative measurement. The proposed filter-based method includes four distinctive filters: filter over environmental noise level, filter over non-pitch ratio within utterance, filter over average phoneme duration function score and filter over clipped frame composition ratio. For the evaluation, the initial performance of the acoustic model was measured at 66.1% in terms of speech recognition rate. The overall performance is demonstrated as 73.8% when all of the proposed filters are applied for the re-training of the acoustic model. This result demonstrates 3.1% better recognition rate than a confidence measure-based acoustic model re-training method. Our proposed method is applicable to other data-driven classification services of consumer electronic products in other mediums (e.g. image) at their early stage of commercialization.

Smart: Semantically Mashup Rest Web Services

A mashup is a combination of information from more than one source, mixed up in a way to create something new, or at least useful. Anyone can find mashups on the internet, but these are always specifically designed for a predefined purpose. To change that fact, we implemented a new platform we called the SMART platform. SMART enables the user to make his own choices as for the REST web services he needs to call in order to build an intelligent personalized mashup, from a Google-like simple search interface, without needing any programming skills. In order to achieve this goal, we defined an ontology that can hold REST web services descriptions. These descriptions encapsulate mainly, the input type needed for a service, its output type, and the kind of relation that ties the input to the output. Then, by matching the user input query keywords, with the REST web services definitions in our ontology, we can find registered services individuals in this ontology, and construct the raw REST query for each service found. The wrap up from the keywords, into semantic definitions, in order to find the matching service individual, then the wrap down from the semantic service description of the found individual, to the raw REST call, and finally the wrap up of the result again into semantic individuals, is done for two main purposes: the first to let the user use simple keywords in order to build complex mashups, and the second to benefit from the ontology’s inference engine in a way, where services instances can be tied together into an intelligent mashup, simply by making each service output individuals, stand as the next service input.

An Intellectual Approach for Providing Secure Environment in Real World Web Application

In recent years, widespread adoption of the internet has resulted in rapid advancement in information technologies. Internet is used by the general population for the purposes such as financial transactions, educational activities and countless other activities. This development of the Internet use has unfortunately been accompanied by a growth of malicious activity in the web application. Every organization uses web application for accessing their data from the database, these applications use user inputs to create a query for storing, retrieving data from the database. Although all the organizations are concerned about their data security, the attackers can still corrupt the data by using the techniques like SQL injection, Hijack future session attack, Privilege Escalation attack, which are commonly referred to as intrusion, it also includes another Application layer attack called DDoS attack. In this paper we present a novel approach for detecting all the intrusions and Application layer attack by analyzing user behavior and analyzing the user input queries using low interaction Honeypot technique.

Query Recommendation for Improving Search Engine Results

As web contents grow, the importance of search engines become more critical and at the same time user satisfaction decreases. Query recommendation is a new approach to improve search results in web. In this paper a method is proposed that, given a query submitted to a search engine, suggests a list of queries that are related to the user input query. The related queries are based on previously issued queries, and can be issued by the user to the search engine to tune or redirect the search process. The proposed method is based on clustering processes in which groups of semantically similar queries are detected. The clustering process uses the content of historical preferences of users registered in the query log of the search engine. This facility provides queries that are related to the ones submitted by users in order to direct them toward their required information. This method not only discovers the related queries but also ranks them according to a similarity measure. The method has been evaluated using real data sets from the search engine query log.

시맨틱 웹 서비스 환경에서 시맨틱 질의 어댑터의 설계 및 구현

시맨틱 웹 서비스는 최근에 많은 연구가 수행되고 있는 시맨틱 웹 기술을 기반으로 웹 서비스를 지원하는 웹 기술이다. 지금까지 시맨틱 웹 서비스와 관련된 연구는 시맨틱 웹 문서 저장 기법과 시맨틱 질의 처리를 위한 추론 엔진 개발 둥에 많이 집중되어 왔다. 그러나 근본적으로 시맨틱 웹 서비스 환경을 지원하기 위해서는 사용자 혹은 에이전트가 시맨틱 정보를 질의할 수 있는 효과적인 질의 인터페이스의 제공이 필수적이다. 이에 관하여, 본 논문에서는 복잡한 시맨틱 정보에 대해 일반 사용자의 높은 질의 투명성을 제공하기 위한 시맨틱 질의 어댑터(SQA)를 제안한다. 먼저 BAML-S Profile의 요소를 분석하여 그래픽 기반의 절차적 사용자 질의 인터페이스를 설계한다. 그리고 사용자 인터페이스로 입력된 질의는 시맨틱 질의어인 RDQL로 변환하도록 구현하였다. 이때 RDQL 술어의 결합(disjunctive) 질의 문제를 해결하기 위한 시맨틱 질의어 다중 생성 프로세서를 제시한다. The Semantic Web Services is a next-generation Web technology that supports Web services, based on the semantic Web technologies. Until now, the researches on semantic Web services may be foiled on the semantic Web document management and the inference engine to efficiently process the semantic Queries. However, in order to realize the principle semantic Web environment it is necessary to provide a semantic query interface though which users and/or agents can efficiently request semantic information. In this regard, we propose the Semantic Query Adapter(SQA) to provide a high query transparency with users, especially when querying about a complex semantic information. We first design the procedural user query interface based on a graphic view, by analyzing DAML-S Profile documents. And then, we builds a module which a user input query transforms its corresponding RDQL. We also propose the multiple semantic query generating procedure as a new method to solve the disjunctive query problem of the RDQL primitive.

Domain-specific web search with keyword spices

Domain-specific Web search engines are effective tools for reducing the difficulty experienced when acquiring information from the Web. Existing methods for building domain-specific Web search engines require human expertise or specific facilities. However, we can build a domain-specific search engine simply by adding domain-specific keywords, called "keyword spices," to the user's input query and forwarding it to a general-purpose Web search engine. Keyword spices can be effectively discovered from Web documents using machine learning technologies. The paper describes domain-specific Web search engines that use keyword spices for locating recipes, restaurants, and used cars.