ソフトウェアセキュリティとは ソフトウェアセキュリティのリスク管理 技術の選択 オープンソースとクローズドソース ソフトウェアセキュリティのガイドライン ソフトウェア監査 バッファオーバーフロー アクセス制御 競合状態 ランダム性と決定性 暗号技術の利用 信頼性管理と入力検証 パスワード認証 データベースセキュリティ クライアントサイドセキュリティ ファイアウォールのセキュリティ 暗号技術の基礎

Building Secure Software : ソフトウェアセキュリティについて開発者が知っているべきこと

Neural Machine Translation (NMT) is a new technique for machine translation that has led to remarkable improvements compared to rule-based and statistical machine translation (SMT) techniques, by overcoming many of the weaknesses in the conventional techniques. We study and apply NMT techniques to create a system with multiple models which we then apply for six Indian language pairs. We compare the performances of our NMT models with our system using automatic evaluation metrics such as UNK Count, METEOR, F-Measure, and BLEU. We find that NMT techniques are very effective for machine translations of Indian language pairs. We then demonstrate that we can achieve good accuracy even using a shallow network; on comparing the performance of Google Translate on our test dataset, our best model outperformed Google Translate by a margin of 17 BLEU points on Urdu-Hindi, 29 BLEU points on Punjabi-Hindi, and 30 BLEU points on Gujarati-Hindi translations.

Neural Machine Translation of Indian Languages

Crime is one of the most predominant and alarming aspects in our society and its prevention is a vital task. Crime analysis is a systematic way of detecting and investigating patterns and trends in crime. The aim of this model is to increase the efficiency of crime investigation systems. This model detects crime patterns from inferences collected from the crime scene and predicts the description of the perpetrator who is likely suspected to commit the crime. This work has two major aspects: Crime Analysis and Prediction of perpetrator identity. The Crime Analysis phase identifies the number of unsolved crimes, and analyses the influence of various factors like year, month, and weapon on the unsolved crimes. The prediction phase estimates the description of the perpetrators like, their age, sex and relationship with the victim. These predictions are done based on the evidences collected from the crime scene. The system predicts the description of the perpetrator using algorithms like, Multilinear Regression, K- Neighbors Classifier and Neural Networks. It was trained and tested using San Francisco Homicide dataset (1981–2014) and implemented using python.

Crime Data Analysis and Prediction of Perpetrator Identity Using Machine Learning Approach

In the modern digital era, software systems are extensively adapted and have become an integral component of human society. Such wide use of software systems consists of large and more critical data that inevitably needs to be secured. It is imperative to make sure that these software systems not only satisfy the users’ needs or functional requirements, but it is equally important to make sure the security of these software systems. However, recent research shows that many software development methods do not explicitly include software security measures during software development as they move from demand engineering to their final losses. Integrating software security at each stage of the software development life cycle (SDLC) has become an urgent need. Tackling software security, various methods, techniques, and models have been suggested and developed, however, only a few of them provide strong evidence for building secure software applications. The main purpose of this research is to study security measures in the context of the development of secure software (SSD) during the study of systematic mapping (SMS). Based on the inclusion and exclusion criteria, 116 studies were selected. After the data extraction from the selected 116 papers, these were classified based on the quality assessment, software security method, SDLC phases, publication venue, and SWOT analysis. The results indicate that this domain is still immature and sufficient research work needs to be carried out particularly on empirically evaluated solutions.

Systematic Mapping Study on Security Approaches in Secure Software Engineering

A huge amount of valuable resources is available on the web in English, which are often translated into local languages to facilitate knowledge sharing among local people who are not much familiar with English. However, translating such content manually is very tedious, costly, and time-consuming process. To this end, machine translation is an efficient approach to translate text without any human involvement. Neural machine translation (NMT) is one of the most recent and effective translation technique amongst all existing machine translation systems. In this paper, we apply NMT for English-Tamil language pair. We propose a novel neural machine translation technique using word-embedding along with Byte-Pair-Encoding (BPE) to develop an efficient translation system that overcomes the OOV (Out Of Vocabulary) problem for languages which do not have much translations available online. We use the BLEU score for evaluating the system performance. Experimental results confirm that our proposed MIDAS translator (8.33 BLEU score) outperforms Google translator (3.75 BLEU score).

/pdf/neural-machine-translation-for-english-tamil-fbmqtgup02.pdf

Neural Machine Translation for English-Tamil

Information Present in Different language and Structure gives Rise to language as barrier in information retrieval. Informative Document on queen Elizabeth is been writing by foreign language English which makes its difficult for a Marathi reader to understand and seek History of England, on Similar lines Literature Work on Shivaji is mostly documented in Marathi which makes foreign Historians difficult to gain know, in both case user is at times unknown of facts due to language and may lose interest on information. Vital information on current happing on village and taluka level are been published in newspaper with local language which setback information spread among other Masses. Many Time government documents and forms are been presented in English Language where a lay man from Marathi language background finds difficulty to understand information and even avoid such procedures therefore it highly urges for need of automated Software based Translation system which would assist in cross Domain information Retrieval. Machine Translation assist to translate Information presented in one language to other language. Information can be present in form of text, speech and image translating this information helps for sharing of information and ultimately information gain. A lot of work has been done on Translation of English to Hindi, Tamil Bangla and other foreign languages also. Machine Translation is challenging Research Area with numerous issues due to language ambiguity like grammar, Structure and even fluency of use. Numerous Methodologies have been proposed and developed which have uplifts and downfalls also, Although statistical and rule based at core with each having limitations. rule based produce accurate mapped translation and are trainable system but costly, whereas statistical produce fluent translation but lack accuracy and sense. Hybrid is combine approach which integrated approach and helps to optimize translation output. The research manuscript we present hybrid machine Translator for English to Marathi language which translated Web pages, text Documents on Agriculture (crops fruits for farmer), Medical reports in Marathi and tourism related information. Proposed System consists of Parallel Multi-Engines which process statistical and rule based Translation for same input document and produce a optimized result by performing statistical over rule based which give fluent language sense outputs. Mapper algorithm is been used in rule based Translation, with Agriculture corpus, medical and tourism corpus for statistical evaluation. Marathi wordnet has been implemented to enhance dictionary and incorporate better translation result Currently System has been proposed for text Document which can be extended to speech and voice. Comparative analysis for point view in one dimension of only limited set of Queries is done with Google Translator. Holding hybrid approach as better methodology. A Systematic survey of only 10 key articles used in research has been done. This research article is extension of our previous research surveys and partial implementations. And innovative S-measure has been new parameter proposed and evaluated by our research team.

Hybrid machine translation for English to Marathi: A research evaluation in Machine Translation: (Hybrid translator)

Test case prioritization includes testing experiments in a request that builds the viability in accomplishing some execution objectives. The importance amongst the most imperative testing objectives is the fast rate of fault recognition. Test case ought to run in a request that extends the likelihood of fault discovery furthermore that detects the most serious issues at the early stage of testing life cycle. In this paper, we develop and prove the necessity of Component-Based Software testing prioritization framework which plans to uncover more extreme bugs at an early stage and to enhance software product deliverable quality utilizing Genetic Algorithm (GA) with java decoding technique. For this, we propose a set of prioritization keys to plan the proposed Component-Based Software java framework. In our proposed method, we allude to these keys as Prioritization Keys (PK). These keys may be project size, scope of the code, information stream, and bug inclination and impact of fault or bug on overall system, which prioritizes the Component-Based Software framework testing. The integrity of these keys was measured with implementation of key assessment metric called KAM that will likewise be ascertained. This paper demonstrates how software testing can be efficient with management of data integrity factor to avoid major security issues. One of the main advantages of our approach is that domain specific semantics can be integrated with the data quality test cases prioritization, thus being able to discover test feed data quality problems beyond conventional quality measures.

Component-Based Software System Test Case Prioritization with Genetic Algorithm Decoding Technique Using Java Platform

Search Engines are based on bivalent logic and probability theory with lack of real world conceptual knowledge, higher precision and reasoning ability. Ranked links or snippets are less effective with big data Analysis and web of information. State of art Search Engine Google endeavor to retrieve good quality document, Msn Engine i.e. Bing endeavor decision boosting. Advanced algorithms account in user intents semantics and societal patterns on Web with recommendation system as product (recommendation Engines). Search engines have advanced from Text based to voice based (Dialogue based) to Image based (multimedia as input Question to search). Major advancement analysis of search Engine Enhancement, search Engines have advanced from traditional data base retrieval machine to web based machines from horizontal engines to Vertical Engines (Naukari.com, Quickr.com) with dedicated crawler Technology at heart. Information of web is structured instructed and posses a huge challenge to big data analytics. All though Retrieval results are optimistic yet they Lack in ability to interpret user Question. Precise answering from relevant Informatics is Search Engine Enhancement. Time complexity and memory are Algorithmic and Machine design parameters that support optimized search result. Question Answering Search Engine (QA Engine) is machine with deductive reasoning capability ability to amalgamate information from various knowledge datasets. QA Engine is front linear area in advanced information retrieval techniques, state of art technology to future of search engines, expert systems. QA search engines have advanced from Shallow Technique (keyword technique) to template based Structured Knowledge processing Engines, profile based engines, and context based machines to cross language machines to Multimedia QA Engines. Community based QA Yahoo answer, stack overflow to specialized search engines like ask.com, qura.com, true knowledge. Question Answering system have been embedded in Google Search Engine(Google Question Answering). IBM's Watson, a cognitive machine Thinking machine like Humans is decision support Engine developed under Deep QA project with advanced Natural language processing Information Retrieval with deep mining, a machine learning model that learns over time with reasoning model at base. With advent of android platform web based services like Apple Siri, Android Assistant questioner technology is art that assists. Enhancement future search engines with analysis (thinking), cognitive ability. This manuscript I present in Architecture on Question answering Search Engine that is search Engine Enhancement with Finite State Machine which facilitate answering question to time complexity The abstract, methodology, algorithmic analysis, of 20 research paper facilitate a future research and integration of proposed Architecture of Question Answering Search Engine.

Question Answering Search engine short review and road-map to future QA Search Engine

This paper focuses on a method to obtain sentiments of people from twitter and using it to analyze whether they can become threat to particular person or society. Machine learning is one of the technologies which is used for decision making and also make predictions by learning more from dataset. Since huge amount of data is streaming very fast, it needs to be classified in short duration. In this paper we aim to achieve faster classification by using Big Data with Naive Bayes Classifier. We have demonstrated this by using HIVE to store and process data.

Crime mitigation at Twitter using Big Data analytics and risk modelling

File System is Data-Structure with Logic (method) to handle efficiently a group of information on disk. File systems come are classified according to their structure method of operation, speed of operation, scalability and flexibility size and security. A file system framework is an intends to sort out data anticipated that would be retained after a program terminates by giving methodology to store, retrieve and update information, and also deal with the accessible space on the device(s) which contain it. A file system framework sorts out information in a proficient way and is tuned to the particular attributes of the device. Individual device partitions can be setup utilizing one of the wide ranges of file systems. Each has its own particular features, impediments, and unique characteristics. This paper concentrates on the components of versatile and adaptable parallel file system framework named Lustre. A null hypothesis work has been presented on Luster file system which is a step in HPC(high performance computing). The hypothesis accepts research scope from survey of articles from various portals and presents in better approach developed by our team in research domain of file system to handle issues and built a better scalable file system like luster. A sequential survey analysis has been done on file system for articles from last three decades in File system research. The research scope unwrapped by scholars is been highlighted in this article and implementation work of innovative scalable file system is been presents which solves many research Questions and space in file system research domain. Performance evaluation helps in assisting File system in current scenario. A systematic literature survey has been incorporated in manuscript for finding research directions. Twenty research articles have ben breakdown in abstract methodology and research scope to find in challenges and subsequent methods to overcome those challenges.

Shashank Joshi

Papers

Hybrid machine translation for English to Marathi: A research evaluation in Machine Translation: (Hybrid translator)

Component-Based Software System Test Case Prioritization with Genetic Algorithm Decoding Technique Using Java Platform

Question Answering Search engine short review and road-map to future QA Search Engine

Crime mitigation at Twitter using Big Data analytics and risk modelling

Luster a scalable architecture file system: A research implementation on active storage array framework with Luster file system