Mohammad Abid Khan

Selecting predicate logic for knowledge representation by comparative study of knowledge representation schemes

In Artificial Intelligence, knowledge representation is a combination of data structures and interpretive procedures that leads to knowledgeable behavior. Therefore, it is required to investigate such knowledge representation technique in which knowledge can be easily and efficiently represented in computer. This research paper compares various knowledge representation techniques and proves that predicate logic is a more efficient and more accurate knowledge representation scheme. The algorithm in this paper splits the English text/sentences into phrases/constituents and then represents these in predicate logic. This algorithm also generates the original sentences from the representation in order to check the accuracy of representation. The algorithm has been tested on real text/sentences of English. The algorithm has achieved an accuracy of 80%. If the text is in simple discourse units, then the algorithm accurately represents it in predicate logic. The algorithm also accurately retrieves the original text/sentences from such representation.

Lexical-semantic divergence in Urdu-to-English Example Based Machine Translation

Divergence is one of the common obstacles in Example Based Machine Translation (EBMT). It needs to be identified, categorized and resolved to obtain correct translation for any pair of language. The focus in this research is on lexical semantic divergence and six different types are identified and generalizations are made on the basis of examples, for Urdu to English translation. Strategies are also presented for the identification of these types. This work helps in efficient retrieval of the examples from a bilingual corpus, which in turn will help in the alignment and recombination stages of EBMT, to produce good quality translation.

Reciprocal anaphora resolution in Pashto discourse

It is very important for an effective machine translation system to transform the multi-sentential discourses, of a source language, into simple and cohesion free discourses. Anaphora resolution is the predicament in computational linguistics that is responsible for the alleged problem. This paper is about the lexical (reciprocal) anaphora resolution in Pashto language and presents an algorithm that is based on the rules to resolve the reciprocal pronouns (RecipPro) of Pashto language. The algorithm has been tested on real world text taken from Pashto stories, novels, newspapers, Web-pages and showed an accuracy of 94.45%.

Strong Personal Anaphora Resolution in Pashto Discourse

Treatment of natural languages by computer is a difficult task because natural languages are complex, context sensitive and abstract. Rationale behind anaphora resolution is to make languages easier for computer to comprehend and process them further. This paper is focused on an algorithm for personal anaphora resolution and some important factors that play a major role in the resolution process. The algorithm resolves strong 1st person (PI), 2nd person (P2) and 3rd person (P3) pronouns in Pashto language in their direct, oblique and possessive cases. The algorithm has been tested on real world text taken from novels and newspapers and showed an accuracy of 80%.

Treatment of Pronominal Anaphoric Devices in Urdu Discourse

Handling of human language by computer is a very intricate and complex task. In natural languages, the sentences are usually part of discourse units just as words are part of sentences. Anaphora resolution plays a significant role in discourse analysis for chopping larger discourse units into smaller ones. This process is done for the purpose of better understanding and making the processing of text easier by computer. This paper is focused on the discussion of various factors that play an important role in pronominal anaphora resolution in Urdu discourses. An algorithm is developed that resolves pronominal anaphoric devices with 86% success rate

Rule based personal references resolution in pashto discourse for better machine translation

Pashto is one of the richest languages of the world and consists of a large number of personal pronouns that are categorized into strong, weak and directional pronouns. In this paper a computational approach for resolving these pronouns is proposed and an algorithm is developed. This algorithm is the combination of two sub-algorithms that resolves these pronouns. The algorithm has been tested on real world text taken from novels and stories which showed 78% accuracy.

Effective Use of Evaluation Measures for the Validation of Best Classifier in Urdu Sentiment Analysis

Sentiment analysis (SA) can help in decision making, drawing conclusion, or recommending appropriate solution for different business, political, or other problems. At the same time reliable ways are also required to verify the results that are achieved after SA. In the frame of biologically inspired approaches for machine learning, getting reliable result is challenging but important. Properly verified and validated results are always appreciated and preferred by the research community. The strategy of achieving reliable result is adopted in this research by using three standard evaluation measures. First, SA of Urdu is performed. After collection and annotation of data, five classifiers, i.e., PART, Naives Bayes mutinomial Text, Lib SVM (support vector machine), decision tree (J48), and k nearest neighbor (KNN, IBK) are employed using Weka. After using 10-fold cross-validation, three top most classifiers, i.e., Lib SVM, J48, and IBK are selected on the basis of high accuracy, precision, recall, and F-measure. Further, IBK resulted as the best classifier among the three. For verification of this result, labels of the sentences (positive, negative, or neutral) are predicted by using training and test data, followed by the application of the three standard evaluation measures, i.e., McNemar’s test, kappa statistic, and root mean squared error. IBK performs much better than the other two classifiers. To make this result more reliable, a number of steps are taken including the use of three evaluation measures for getting a confirmed and validated result which is the main contribution of this research. It is concluded with confidence that IBK is the best classifier in this case.

Knowledge representation of Urdu text using predicate logic

Knowledge representation is a key area of research in artificial intelligence which deals with the proper storage and retrieval of knowledge for various useful applications. This research paper proves that knowledge can be easily and efficiently represented in predicate logic. The algorithm in this paper splits the Urdu text/sentences into phrases/constituents and then represents these in predicate logic. This algorithm also generates the original sentences from the representation in order to check the accuracy of representation. The algorithm has been tested on real text/sentences of Urdu. The algorithm has achieved an accuracy of 88%. As the algorithms works on pre-tagged input file, so if the tagging is done correctly then the algorithm achieves high level of accuracy. Therefore it is required that there should be proper rules by the help of which one can correctly tag the input text into phrases/ constituents. The algorithm accurately represents such text in predicate logic. The algorithm also accurately retrieves the original text/sentences from such representation.

Implementation of the rule-based approach for the resolution of strong personal anaphora in Pashto discourse

The rationale behind anaphora resolution in Pashto language is to make it easier for computer to comprehend and further process Pashto text accurately. This paper is focused on the implementation of the algorithm (Ali et al., 2007) and modifying few of its rules. It also adds some new rules and eliminates an undesirable one to enhance the accuracy and efficiency. The algorithm uses a rule-based approach for the resolution of strong personal pronouns of Pashto language in their direct, oblique, and possessive cases. It has been implemented in Visual C#.NET and tested on real world text taken from different genres. The modified algorithm increases the accuracy from 80% to 85.14%.

Ellipses resolution in WH-constructions in Pashto

The constructions in Pashto, having a question involving a WH-word equivalent and an answer or response to that question, are called WH-constructions. In most of the cases, in both written and spoken WH-constructions, there occur ellipses among them. This paper presents the first ever rulebased algorithm and the factors involved in ellipses resolution in WH-constructions in Pashto text. The algorithm resolves ellipses occurring in the answers to WH-questions by using the morphological and syntactic information of both the question and its corresponding answer. The algorithm accurately resolves the ellipses occurring in the answers to the questions having WH-words i.e. (Who-Dir), (Who-Obl), (What), (Which), (Where) and (How much) among them. The algorithm has been tested on real world text of Pashto taken from various genres such as books, novels, newspapers and online magazines. The algorithm achieved an accuracy of 76%.