Information Query Research Articles

SwipeCut: Interactive Segmentation via Seed Grouping

Interactive image segmentation algorithms rely on the user to provide annotations as the guidance. When the task of interactive segmentation is performed on a small touchscreen device, the requirement of providing precise annotations could be cumbersome to the user. We design a new interaction mechanism that actively queries seeds for guiding the user to label. Our method enforces sparsity and diversity criteria on the selection of query seeds, and at each round of interaction, the user is only presented with a small number of informative query seeds that are far apart from each other. Therefore, the user merely has to swipe through on the ROI-relevant query seeds for checking which ones of the query seeds are inside the region of interest (ROI). This kind of interaction should be easy since those gestures are commonly used on a touchscreen. As a result, we are able to derive a user-friendly interaction mechanism for annotation on small touchscreen devices. The performance of our algorithm is evaluated on six publicly available datasets. The evaluation results show that our algorithm achieves high segmentation accuracy, with short computational time and less user feedback.

A Methodology of Constructing a Distributed Information System for Searching for Scientific and Technical Information Based on an Object Data Model

A methodology of organizing the search for scientific and technical information and the creation of an appropriate information retrieval system is considered. The methodological basis is the three-level hierarchical structure of ontologies. The first level forms the conceptual apparatus of the subject area, the second level details the terminology of a specific area of scientific knowledge. Ontologies of the third level are focused on a specific scientific direction, they allow one to structure objects and research methods, as well as the scientific result and practical value of a scientific publication. An object data model is used. The basics of the language of information queries are proposed.

Gene Teller: an extensible Alexa Skill for gene-relevant databases.

SummaryVoice assistants have become increasingly embedded in consumer electronics, as the quality of their interaction improves and the cost of hardware continues to drop. Despite their ubiquity, these assistants remain underutilized as a means of accessing biological research data. Gene Teller is a voice assistant service based on the Alexa Skills Kit and Amazon Lambda functions that enables scientists to query for gene-centric information in an intuitive manner. It includes several features, such as synonym disambiguation and short-term memory, that enable a natural conversational interaction, and is extensible to include new resources. The underlying architecture, based on Simple Storage Service and Amazon Web Services Lambda, is cost efficient and scalable.Availability and implementationA publicly accessible version of Gene Teller is available as an Alexa Skill from the Amazon Marketplace at https://www.amazon.com/dp/B08BRD8SS8. The source code is freely available on GitHub at https://github.com/solinvicta/geneTeller.

Research on Application of Three-Dimensional Design Standardization and Three-Dimensional Design Review Method for Transmission and Transformation Project

This paper systematically introduces a three-dimensional transmission and transformation engineering auxiliary design software, and analyses its function, structure and technical characteristics. The main feature of this software is that it has a relatively powerful three-dimensional auxiliary design capability, which can help designers quickly generate a three-dimensional model of power transmission and transformation projects. The three-dimensional model can not only give the project investor an overall image and visual effect of the project during the design bidding stage, but also serve as a reference for the design of the relevant drawings in the later period. At the same time, after having an accurate model, the designer can verify the correctness and reliability of the design, can directly obtain some spatial information, and can dynamically modify, adjust and verify the model as needed. The software also provides related information statistics, query, modification, retrieval and other functions, and provides corresponding drawing management tools according to drawing management specifications. (With the actual application of the system, the design level and technical capabilities of designers will be further improved to ensure that the design work can be completed with high quality and efficiency.

Learning semantic information from Internet Domain Names using word embeddings

Word embeddings is a well-known set of techniques widely used in Natural Language Processing (NLP). These techniques are able to learn words’ semantic based on the distributional hypothesis which states that words that are used and occur in the same contexts tend to purport similar meanings.This paper explores the usage of word embeddings in a new scenario to create a Vector Space Model (VSM) for Internet Domain Names (DNS). Our goal is to find semantically similar domains only using information of DNS queries without any knowledge about the content of those domains.The results presented here have practical applications in many engineering activities including websites recommendations, identification of fraudulent or risky sites, parental-control systems and anomaly detection in network traffic analysis (among others).We use the distributional hypothesis to learn the semantic of domain names from users’ web navigation patterns, validating empirically that domain names that occur in the same web sessions tend to have similar semantic.We also test different word embeddings techniques: word2vec, app2vec (considering time intervals between DNS queries), and fastText (which includes sub-word information).Due to the characteristics of domain names, we found fastText as the best option for building a VSM for DNS, being 10.5% superior than word2vec with Skip-Gram which was the next best technique considering the Mean Average Precision at k (MAP@k) metric, which compares the most similar domains in our VSM with the most similar domains provided by a third party source, namely, similar sites service offered by Alexa Internet, Inc.

Analysis of Various Multilingual Document Clustering

Today’s world is heading towards data science era. As the volume of the data is increasing extremely at exponential rate and data is produced and circulated all over the world, not only in English languages but in every regional language too. Since the data’s are in multilingual it is extremely difficult to manage such huge amount of variant data. Hence there is a scope for research work on multilingual document clustering. By document clustering we can retrieve the information of user query. This technique is to divide a given set of documents into a certain number of clusters. The aim is to create a multilingual document clustering that are related internally, but substantially different from each other. Main challenge faced while creating MDC, is the quality and stability of the cluster which swirls rapidly with document sets. Multilingual Document clustering has to be represented in a form of matrix which is done by either Vector Space Model or TF–IDF method. Where each word is given a value representing particular document. News articles are retrieved from relevant news explorer using proper search engines. Word in news articles are represented in one dimensional error using mathematical vector. Keywords in every articles are selected using term frequency (tf) and evaluated with inverse document frequency (idf). It uses discriminative power of keyword’s over articles. It encourages users to opt for cluster based browsing which is acceptable for processing the results. Big Data tools work efficiently in distributed environment which gives a significant analysis of our retrieved information. Many document clustering works well with small set of data but fails to deal with the large set of document. In this paper we concentrate on bring out the various problems that raise during multilingual document clustering and possible solution to overcome those problems.

Cost and Pricing of Differential Privacy in Demand Reporting for Smart Grids

In smart grid, the highly dynamics of electrical loads introduce uncertainty to the system, which also brings new challenges for the power grid control and optimization. Reporting demand information by customers in solving the economic dispatch control (EDC) problem in power grids is a promising approach, but it raises important privacy concerns. One solution is to add random noise to aggregate queries of demand reports can provide differential privacy (DP) for the individual customers. However, the noisy query information can adversely degrade the EDC's optimality. In this paper, we first analyze the cost in demand reporting in terms of how DP-induced noise will increase the total generation cost. We show that the noise amounts for different customers are intricately coupled with one another in determining the total cost. To deal with the coupling, we apply the principle of Shapley value to attribute fair shares of the total cost to the power grid buses and study the properties of this cost sharing approach. For efficient sharing of the privacy cost, in a manner scalable to large power systems with many buses, we propose heuristic algorithms to approximate the Shapley value. Moreover, we study important network effects of the per-customer DP cost within a privacy group attached to the same bus, in that this cost depends on how many other customers choose the same privacy group. Accordingly, we design a pricing scheme for the customers that we prove to be truthful, i.e., customers have no incentives to deviate from the group that represents their true privacy needs. We also discuss the impact of the high-demand customer on other customers’ cost and propose an approach in the pricing scheme to charge higher cost on the high-demand customer. Trace-driven simulations based on a 5-bus power system model validate our analysis and illustrate the performance of the proposed cost sharing algorithms.

Ubiquitous and Low Power Vehicles Speed Monitoring for Intelligent Transport Systems

This paper presents a high scalability real-time intelligent traffic monitoring system, based on Radio Frequency Identification (RFID). The main features of this system are low cost, low power consumption, traffic monitoring, and connectivity. The system’s architecture includes an RFID reader, a passive tag, and a Raspberry Pi. Our solution collects vehicle information from the labels and stores the data into a database by employing only one antenna. The main challenge is that the RFID module is not robust enough to recognize the information in the vehicle’s RFID tag on each information query, while the label is in the reading zone. This instability does not allow us to know precisely when and where a vehicle enters or leaves the sensing zone. What is more, the high random error in the power signal and the complexity of its characteristic curve pattern add difficulty to the speed calculation, when we reduce the number of antennas to one. For this reason, an innovative approach has been designed, using customized modular neural network (MNN). This method fits the collected data (power signal vs time) affected by acute random noise, to the characteristic correspondence function among the signal power and the position of the terminal, which domains are dimensionally different. As a result, we can estimate the vehicle speeds and obtain the whole vehicle information. Under this novel method, we are able to reduce the hardware, in comparison with previous approaches, making it cheaper and decreasing power consumption.

Research on the Construction of the General Control Platform for Yunnan Rare and Precious Metal Materials Genetic Engineering

In this paper, we initially construct a master control platform of Yunnan rare and precious metal materials (RAPMMs) genetic engineering framework based on the idea of material gene engineering, which includes a master material data management sub-platform, a high-throughput computing sub-platform and a high-throughput material preparation and characterization sub-platform, and connects the data information in the sub-platform. The database covers the standard performance data of RAPMMs from all over the world, and the initial amount of data is 2.2 million items, which provides data resources for deep learning of material big data in the future. In addition, we have researched the fusion and convergence technology of multi-source heterogeneous RAPMMs data, which support the management, processing and storage of all kinds of materials professional data. In order to reduce the researchers’ need for precise retrieval of large amounts of data, the deep search technology based on rare metal data is investigated to achieve precise search in a large data environment and provide professional visualization and statistical analysis techniques. Through the technical research on the interface of the open high-throughput computing platform, the online generation and submission of the operations of the first-principles computing software VASP are realized, and the automatic data identification and import of the calculation result files are achieved. Finally, it realizes the functions of controlling and controlling the high-throughput calculation platform, high-throughput preparation and characterization platform, and data platform, with the functions of resource coordination and scheduling, information sharing, public database management, and relevant information retrieval and query, etc., and establishes a common database structure for rare and precious metals, which lays a solid foundation for the future development of material genetic engineering.

Semantic Service Search in IT Crowdsourcing Platform: A Knowledge Graph-Based Approach

Understanding user’s search intent in vertical websites like IT service crowdsourcing platform relies heavily on domain knowledge. Meanwhile, searching for services accurately on crowdsourcing platforms is still difficult, because these platforms do not contain enough information to support high-performance search. To solve these problems, we build and leverage a knowledge graph named ITServiceKG to enhance search performance of crowdsourcing IT services. The main ideas are to (1) build an IT service knowledge graph from Wikipedia, Baidupedia, CN-DBpedia, StuQ and data in IT service crowdsourcing platforms, (2) use properties and relations of entities in the knowledge graph to expand user query and service information, and (3) apply a listwise approach with relevance features and topic features to re-rank the search results. The results of our experiments indicate that our approach outperforms the traditional search approaches.

Differences in forum communication of residents and visitors in MOOCS

Facilitation of social interactions in Massive Open Online Courses can benefit from conceptualizing forum sub-populations through a networked lens. Such a lens implies that forum interactions represent a network of learners with heterogeneous levels of commitment to forum activity. A networked lens helps capture forum communities, so-called forum residents, as well as the rest of the crowd, also known as forum visitors. The current study empirically demonstrates the differences between these forum sub-populations. Building on a large dataset of manually labelled discussion threads in four edX MOOCs, our findings uncover two distinct patterns in the communication of posters with different length of forum commitment. As the courses progressed, all types of communication decreased for forum visitors, but both socio-cognitive and informational queries increased for forum residents. We find that the communication dynamics of committed forum posters cannot be observed when the entire forum population is examined in its entirety. Further, the study profiles learners around discussion type sequences. We show that communication topics of forum visitors appear narrow and topical when compared to the diversity of topics by resident posters. This paper offers a foundation towards the scaling of social teaching practices for personalised learning of both residents and visitors as well as community development in massive open online courses.

On the equivalence of optimal recommendation sets and myopically optimal query sets

Preference elicitation is an important component in many AI applications, including decision support and recommender systems. Such systems must assess user preferences, based on interactions with their users, and make recommendations using (possibly incomplete and imprecise) beliefs about those preferences. Mechanisms for explicit preference elicitation—asking users to answer direct queries about their preferences—can be of great value; but due to the cognitive and time cost imposed on users, it is important to minimize the number of queries by asking those that have high (expected) value of information.An alternative approach is to simply make recommendations and have users provide feedback (e.g., accept a recommendation or critique it in some way) and use this more indirect feedback to gradually improve the quality of the recommendations. Due to inherent uncertainty about a user's true preferences, often a set of recommendations is presented to the user at each stage. Conceptually, a set of recommendations can also be viewed as choice query, in which the user indicates which option is most preferred from that set.Because of the potential tension between making a good set recommendation and asking an informative choice query, we explore the connection between the two. We consider two different models of preference uncertainty and optimization: (a) a Bayesian framework in which a posterior over user utility functions is maintained, optimal recommendations are assessed using expected utility, and queries are assessed using expected value of information; and (b) a minimax-regret framework in which user utility uncertainty is strict (represented by a polytope), recommendations are made using the minimax-regret robustness criterion, and queries are assessed using worst-case regret reduction. We show that, somewhat surprisingly, in both cases, there is no tradeoff to be made between good recommendations and good queries: we prove that the optimal recommendation set of size k is also an optimal choice query of size k. We also examine the case where user responses to choice queries are error prone (using both constant and mixed multinomial logit noise models) showing the results are robust to this form of noise.In both frameworks, our theoretical results have practical consequences for the design of interactive recommenders. Our results also allow us to design efficient algorithms to compute optimal query/recommendation sets. We develop several such algorithms (both exact and approximate) for both settings and provide empirical validation of their performance.

Creation of Query Expansion Based on the Subject Domain Thesaurus in the Ontology of Knowledge of the Semantic Library

Possibilities of query expansion with subject area thesaurus are discussed. The role of the context defined by thesaurus term links is both to refine the query and to increase the size of the sample on the query. Of particular importance is the process of expanding the query for scientific subject areas where the search based on special terminology. In this case, thesauruses of subject areas must be used to minimize the occurrence of information noise. The proposed approach takes into account the application of similar terminology in various subject areas. Examples of the use of thesaurus of separate sections of equations of mathematical physics and related fields demonstrate the effectiveness of the chosen approach of research. By linking to concepts of information resources of other areas of knowledge, the extension of the information query captures search fields of remote subject areas and various types of data, texts, symbolic, audio and video archives. Research shows that expanding the query based on context semantics improves the search quality of scientific publications in digital information and increases the effectiveness of scientific interdisciplinary research.

Privacy preserving search services against online attack

Searchable functionality is provided in many online services such as mail services or outsourced data storage. To protect users privacy, data in these services is usually stored after being encrypted using searchable encryption. This enables the data user to securely search encrypted data from a remote server without leaking data and query information. Public key encryption with keyword search is one of the research branches of searchable encryption; this provides privacy-preserving searchable functionality for applications such as encrypted email systems. However, it has an inherent vulnerability in that the information of a query may be leaked using a keyword guessing attack. Most of existing works aim to make the system resistant to offline keyword guessing, but this does not protect against online attacks on real world services. In this paper, we move a step forward to present a generic framework able to resist online keyword guessing attack using a server-assisted model. Specifically, we design a novel primitive C mirrored all-but-one lossy encryption, which can prevent a specific user from generating valid encryptions. This primitive can be seen as an access control on encryption ability. Combining searchable encryption technique with the new primitive makes online keyword guessing attack impossible for the specified user, even if the attack is launched online. We further give formal security analysis for the generic framework, and a concrete implementation with efficiency analysis to show that our design is practical.

Relevance-Promoting Language Model for Short-Text Conversation

Despite the effectiveness of sequence-to-sequence framework on the task of Short-Text Conversation (STC), the issue of under-exploitation of training data (i.e., the supervision signals from query text is ignored) still remains unresolved. Also, the adopted maximization-based decoding strategies, inclined to generating the generic responses or responses with repetition, are unsuited to the STC task. In this paper, we propose to formulate the STC task as a language modeling problem and tailor-make a training strategy to adapt a language model for response generation. To enhance generation performance, we design a relevance-promoting transformer language model, which performs additional supervised source attention after the self-attention to increase the importance of informative query tokens in calculating the token-level representation. The model further refines the query representation with relevance clues inferred from its multiple references during training. In testing, we adopt a randomization-over-maximization strategy to reduce the generation of generic responses. Experimental results on a large Chinese STC dataset demonstrate the superiority of the proposed model on relevance metrics and diversity metrics.1

Big Data Analysis of the Ecological Environment in the Dianchi Watershed

To analyze the massive data on the ecological environment in Dianchi Watershed scientifically, understand its situation and dynamic trend over time, we plan to establish a feature-based four-dimensional space-time data model and build an integrated space-time database. The SuperMap map database is used to establish a space-time visual query and analysis system of the ecological environment information based on WebGIS with the B/S structure. The ecological environment information of the Dianchi Watershed is taken as the research object for scientific and comprehensive analysis and management of the ecological environment situation, thereby providing intuitive and reliable data information support for the decision-making of the ecological environment construction.

Assembly Information Model Based on Knowledge Graph

There are heterogeneous problems between the CAD model and the assembly process document. In the planning stage of assembly process, these heterogeneous problems can decrease the efficiency of information interaction. Based on knowledge graph, this paper proposes an assembly information model (KGAM) to integrate geometric information from CAD model, non-geometric information and semantic information from assembly process document. KGAM describes the integrated assembly process information as a knowledge graph in the form of “entity-relationship-entity” and “entity-attribute-value”, which can improve the efficiency of information interaction. Taking the trial assembly stage of a certain type of aero-engine compressor rotor component as an example, KGAM is used to get its assembly process knowledge graph. The trial data show the query and update rate of assembly attribute information is improved by more than once. And the query and update rate of assembly semantic information is improved by more than twice. In conclusion, KGAM can solve the heterogeneous problems between the CAD model and the assembly process document and improve the information interaction efficiency.

Application of Cloud Computing in Geological Hazard Early Warning System

China’s geomorphology is complex and changeable. Geological hazards occur frequently. Landslides, mudslides and other geological hazards seriously threaten the lives and property of the people. Monitoring and early warning of geological hazards to reduce the impact of disasters is an urgent problem at this stage. In order to ensure the safety of people’s lives and property, the design and development of real-time monitoring and early warning systems are the most effective means of preventing and controlling disasters. This article is based on the Microsoft cloud computing platform. The geological data collected by the automatic monitoring station is transmitted to the cloud server through GPRS DTU and stored in SQL Server 2008. Generate Web Service through VS2010 platform, use ASP.NET technology to build dynamic website to display early warning details, complete Web Service call through Android smartphone platform, conduct real-time geological disaster warning information query, and provide users with cloud platform-based geological disaster warning system. After analyzing the causes of geological hazards, a model of geological hazard influencing factors, a model for calculating the probability of hazards in geological structures, and a process for processing geographic information data were designed. Using the data of the extracted GIS spatial database to predict geological disasters.

UAV-based Regional Environmental Monitoring IoT System

The UAV regional environmental monitoring system has the characteristics of various functions and simple operation. Relevant staff can ensure effective monitoring of regional temperature and humidity by means of specific mobile networks and flight control of drones. Based on past work experience, this paper summarizes the content of program demonstration and system design, and discusses the software flow design of regional environmental monitoring IoT system from five aspects: process design review, overall structure, topology diagram module, information query display module and system test.

Design of Flood Control Hydrological Information Measurement and Forecast System in a Certain Area

In recent years, hydrological information in this area has played an important role in flood control and drought relief, and achieved good social, economic and ecological benefits, but there is still a certain gap with the latest requirements of the state in hydrological. With the rapid expansion of hydrological information collection stations and the rapid increase of station network density, considering the current needs of local hydrological information management, hydrological management departments and flood control and Drought Relief Command departments, this paper uses software engineering method to design a hydrological information measurement and reporting system. The system realizes centralized management of water information, real-time information query and analysis, dynamic monitoring and alarm, auxiliary decision-making, remote data collection and other hydrological business functions and provides decision support for leaders of all departments. It plays an important role in the success of flood control and drought relief work to minimize the loss of disasters and ensure the safety of people’s lives and properties with the least human and material resources.