With over forty years of use and refinement, access control, often in the form of access control rules (ACRs), remains a significant and widely-used control mechanism for information security. ACRs regulate who can perform specific actions on specific resources within a software-intensive system and are considered a critical component to ensure both confidentiality and integrity. Although software can and does implement access control at the application layer, failure to enforce data access controls at the persistence layer may allow uncontrolled data access when individuals bypass application controls or the ACRs are inconsistently implemented. Our research goal is to improve security and compliance by ensuring access control rules explicitly and implicitly defined within unconstrained naturallanguage product artifacts are appropriately enforced within a system's relational database. Access control implemented in both the application and persistence layers strongly supports a defense in depth strategy. We specify a tool-based process to 1) parse existing, unconstrained naturallanguage product artifacts; 2) classify whether or not a sentence in the product artifact implies access control and whether or not the sentence implies database model elements; and, as appropriate, 3) extract ACR elements; 4) extract database model elements; 5) map extracted data model to a database schema; and 6) implement role-based access control (RBAC) within a relational database. To validate our process, we examined these steps in three studies.
Mr. Elworthy D. Proposed the unprecedented large volume of semi-structured data has exacerbated the need for an easy-to-use query interface for semi-structured data sources. Naturallanguage interfaces and keyword search techniques that take advantage of the data set structure make it very easy for ordinary users to access the data. In this paper, it is introduced that important challenges that lie in the way of building an elective and efficient keyword and/or naturallanguage query interface for semi-structured databases. It shows that the current approaches to this problem rely heavily on heuristics that are intuitively appealing but ultimately ad hoc. Their assumptions are valid for some domains, database designs, and/or schema structures but they are not correct in general. Thus, these often retrieves false. Positive answers, overlook correct answers, and cannot rank answers appropriately.
order. By using tree parsing algorithms, we can easily carry out the comparison process. B-tree is implemented in such a way that we can construct the b-tree of any order based on the number of keywords to be stored. Any new keywords to be added will take a place in an alphabetical order. Searching takes place by finding out the path to the keywords to be searched. This reduces the time by making the search only in either left or right of the root it belongs to recursively. The system will proceed to process the query only if database or table name given in the naturallanguage query exists in the knowledge base. Semantic Analysis:
NATURAL LAGUAGE UPDATES COLING 82, J Horeclc)) (ed ) North Holland Publishing Company ? Academia, 1982 NATNRAL LANGUAGE UPDATES* Sharon C Salveter David Maier Computer Science Department State Univers[.]
Design Tool Combining Keyword Analyzer and Case based Parser for Developing Natural Language Database Interfaces Design Tool Combining Keyword Analyzer and Case based Parser for Developing Natural Lan[.]
The present system is a console-based system for each and every database separately. The user has to learn how to use that console to connect to a particular database and how to work on console to execute query statements. In the real time project development each and every user has a need to connect to different kinds of the databases that are at different locations (systems). Whenever the user wants to connect to a database then the client libraries that are required to connect to that database server has to be installed at the client. This process repeats in each and every user system. In the current system the queries are in high level languages like SQL. The person who is using that system must learn the SQL and write the queries in the High level languages. The existing system works a lot of man-hours both for learning how to use console to connect to different kinds of query statements. To connect to a database from a user system needs specific database client libraries has to install at the user system. Learn and study about a particular database console is time taking process and that console is used to connect to one database only. The drawbacks in the existing system are as follows
Abstract. Despite the large number of NaturalLanguage Interfaces to Databases (NLIDB) that have been implemented, they do not guarantee to provide a correct response in 100% of the queries. In this paper, we present a way of semantic modelling the elements that integrate the knowledge of a NLIDB with the aim of increasing the number of correctly-answered queries. We design semantic representations in order to: a) model any relational database schema and its relationship with the naturallanguage and b) add metadata to naturallanguage words to enable our NLIDB to interpret naturallanguage queries that contain superlatives. We configured our NLIDB in a relational database that we migrated from Geobase and used the Geoquery250 corpus to evaluate its performance. We compare its performance with the interfaces ELF, Freya and NLP-Reduce. The results indicate that our proposal allowed our NLIDB to obtain the best performance.
A Database is a collection of data that is organized to process future accessibility. A database management system is a collection of programs that enables users to create and maintain the database system for access in the future. There are different models to store data, such as the relational database model, in which data is stored as records and fields in rows and columns of tables. Data can be related to each other by defining relationships between tables. Another popular method of storing data that is studied in this thesis is the triplestore method. A Triple consists of a “subject”, a “predicate” and an “object”. The triplestore has some advantages compared with relational stores. Although the relational model is more mature and popular than the triplestore model, it suffers some limitations and deficiencies. A major aspect of the relational model is that it uses a schema for its organization, thus, the designer or database administrator has the responsibility to recognize what type of questions will be asked in advance and builds a schema based on that. Therefore, the relational data has to have schema for its organization. Problems shows up when the database schema needs to be changed because the relational database is not flexible and its schema remains static while the stored data type is not known in advance in some applications, like criminal investigations. In order to provide continuous changes to apply to database schema and queries, some researchers believe that it would be easier
According to comparative genomics studies, human candidate reproduction genes can be used as reference counterpart of that in pigs. By mining and analyzing the biomedical literature database using naturallanguage processing technology, we builded the ReCGiP, which provides candidate genes related to six main reproductive processes including spermatogenesis, oogenesis, fertiliza- tion, preimplantation development, embryo implantation and placental development. Other genes related informa- tion, such as associated literatures, KEGG pathway, GO annotation and OMIM information. The gene-gene co- occurrence networks [9-11] were also included where a line was drawn between two genes if two genes were co- cited in one Pubmed abstract. ReCGiP provides genes which are associated with the reproductive processes and the reproductive traits, and related literature information. The database will facilitate the researchers to choose their interesting genes for the experimental design.
Abstract. In most information systems, databases are accessed and manipulated typically through systems developed to tailor-fit the company’s needs. The usual problem in these cases is the limitation on data accessibility because the users are constrained to the forms created for the system. Another way of accessing the database is through Structured Query Language (SQL), a language that is not familiar to end users, thus still limiting the access to the data. NaturalLanguageDatabase Interfaces were developed to address limited database accessibility for end users. This paper presents AlLaDIn, a web-based naturallanguage interface to the Community-Based Monitoring System of a city in the Philippines.
The Aneesah system has been implemented using a novel PM engine. The PM engine controls user utterance matching against scripts in the system’s knowledge base. The PM engine works based on a two tier approach (Tier 1 and Tier 2) approach. Tier 1 deals with user utterance matching against information stored in domain database (i.e. sales history database etc) tables to capture co-occurrence of attribute or key records leading to the formulation of a query based response, and Tier 2 deals with utterance matching against hand scripted patterns stored in Frequently Ask Questions (FAQ) and General Chat domains. The PM engine has been designed to work with rule based and non- rule based response handling. A rule based response can be described as a scripted textual response, executed following a successful utterance matching in either FAQ domain or General Chat domain. A non-rule based response relates to the formulation of a database query following successful utterance processing against Domain Database Scripts. The PM engine works on principle of pattern matching approach, which partially resembles an approach implemented in InfoChat system . The PM engine works in conjunction with sentence similarity strength function to determine an appropriate response, and also interact with the user to resolve any ambiguity during conversation. The PM engine is equipped to deal with more than one response match situation. In the case, where a user utterance has attracted duplicate responses from different knowledge base domains, the PM engine uses sentence similarity calculation to execute the highest match value response. The user utterance is categorized once a match has been found in the knowledge base. The relevant domain (Domain Database Scripts, FAQ Domain or General Chat Domain) is activated once a user utterance is matched in the knowledge base. The domain activation is used by the PM engine to conversely engage with user and staying relevant to the conversing topic.
Abstract—This paper is concerned with the application of a conversational agent and expert system to provide a naturallanguage interface to a database. Typically, naturallanguagedatabase interfaces (NLDI's) use grammatical and/or statistical parsing. Conversational agents take a different approach, capturing key elements of user input which then trigger pre-determined output templates. It is assumed that the type of naturallanguage questions which could be asked of a specific relational database will contain a limited number of key words (attributes), which could be captured by a conversational agent. In the proposed system, once a conversational agent has identified all relevant attributes and their values, an expert system would then apply rule based reasoning on these attributes to construct an SQL query. The knowledge base of the expert system would contain information on the database structure (metadata) and on the different possible structures of SQL queries. This would result in a real time system, which could extract both database attributes and attribute values from the user input and automatically apply a rule based reasoning system to determine the answer the user’s query.
With the IR in place, the process examines the graph to see if any of the words have previously been detected in prior sentences from an access control perspective. Previously used words in part of an access control tuple are marked with a domain flag for the subject, action, or resource. The tool maintains a domain dictionary consisting of four entry types. The first entry type is an action verb, which represents an action the user takes within the application. For each action verb, we track the corresponding database permission (delete, insert, update, or select) required to perform the action within the database. For example, the “edit” verb has the permissions “select” and “update.” The “select” permission is required as the user must have access to view the record and then “update” to modify the record in the database. We also track whether or not the verb implies a negative permission, which used to revoke the corresponding access from a role. Subjects form the second entry type in the dictionary. Subjects correspond to database roles. The third entry type is a resource, which corresponds to a table or column within the database. The fourth word type is a list of domain words relevant to the current application domain. Since the user provides these domain words, they are populated into the dictionary at the start of the process. Domain ontologies, taxonomies, and thesauri may serve as sources for these words.
In most databases and information systems, cooperation with the users is uncommon. Users of databases and information systems need to know the underlying database structure and compose an accurate and in-depth SQL queries in other to generate direct answers. The need to develop databases and information systems that can predict and accept users’ queries in simplified naturallanguage and SQL has led to many findings and development. The well-known works concerned with NLP and cooperative systems are as follows:
tables which are properly normalized. Now if the user in global wishes to access the data from the table in the database, he/she has to be technically proficient in the SQL language to make a query for the UNIVERSITY database. Our system eliminates this part and enables the end user to access the tables in his/her own language. Let us take an example: Suppose we want to view information such as year of establishment of department, and code of department which department name equals “Department of Economics and Management” from the department table of the UNIVERSITY database, we use the following SQL statement (query): SELECT year-of-establishment-of- department, code-of-department FROM department WHERE department-name =’Department of Economics and Management’. But a person, who doesn’t know MySQL database syntax, will not be able to access the UNIVERSITY database unless he/she knows the SQL syntax of firing a query to the database. But using NLP, this task of accessing the database will be much simpler. So the above query will be rewritten using NLP as a question in the web user interface as: What is the year of establishment of the department and code of the department which department name equals “Department of Economics and Management”? Both the SQL statement and NLP statement to access the department table in the UNIVERSITY database would result in the same output by making query not n a SQL like query language, but simply in English like naturallanguage.
This paper is proposing a solution to solve both the language nuance   and performance issues with the use of shallow parsing , which does not re- quire an understanding of language nuances. The shallow parsing approach be- ing proposed by this paper is the use of keywords  to identify characteristics that are important for the search. This paper will introduce the use of an index file containing keywords that can be used to enhance the performance. Jwalapu- ram & Mamidi  are among a number of authors who have carried out re- search into using keywords to enable NLIDB based systems to perform searches. The keyword searching proposed in this paper unlike Jwalapuram & Mamidi  uses Part of Speech (POS)  processing and an index file which allows for individual words to be extracted from the naturallanguage query. The indivi- dually extracted words can then be used to create the query for the NLIDB solu- tion.
PROCESSING COMPLEX NOUN PHRASES IN A NATURAL LANGUAGE INTERFACE TO A STATISCAL DATABASE PROCESSING COMPLEX NOUN PHRASES IN A NATURAL LANGUAGE INTERFACE TO A STATISTICAL DATABASE F r e d P O P O W l C[.]
Interaction with a database based application requires either a GUI or expertise in database query language. A simple conversational system in naturallanguage will be a great boon to the vast majority of application users. Naturallanguage interfaces to databases (NLIDB) systems allow a user to communicate with the database directly by entering the query in the form of a naturallanguage question. The NLIDB system maps the naturallanguage query to the appropriate SQL by processing the information in the query and correlation with the system and domain metadata. NLIDB can be considered as a classical problem in the field of naturallanguage processing. Although, the earliest research has started since the late sixties, NLIDB remains as an open research problem. The solution to the NLIDB problem can be obtained in two stages i.e. Linguistic processing and Database processing. In the first stage of linguistic processing, the naturallanguage query is mapped and translated into the corresponding SQL query by appropriate mapping functions. In the second stage of database processing, database management and access is performed and the SQL query is executed by the system.