Browsing by Author "Dr. Rao Muhammad Adeel Nawab"

Now showing 1 - 20 of 22

A Benchmark Corpus and Methods for the Urdu Qur’anic Question-Answering System
(Library Information Services, COMSATS University Islamabad, Lahore Campus, 2025) Abida Ashraf; SP23-RCS-002; Dr. Rao Muhammad Adeel Nawab; LHR TP 9497
Question Answering System aims to provide answers to a question posed in natural language. A question can be factoid and non-factoid. With the rapid increase in technologies, the QA task become one of the most challenging and important tasks of Natural Language Processing (NLP). In previous studies, the problem of the QA system for the Holy Quran has been explored for Arabic, English, and Indonesian languages. Urdu has more than 170 million speakers worldwide yet considered low low-resource language. However, the problem of the QA system for the Holy Quran has not been explored for the Urdu language due to the unavailability of the dataset. To overcome this research gap, this study developed a benchmark corpus of the Holy Quran in the Urdu language. We developed the corpus by carefully mapping the QRCD corpus which was originally in Arabic language into Urdu. For the mapping of answers, we take the expertise of data annotators who have a deep knowledge of the Holy Quran in both Arabic and Urdu language. We conducted experiments in a state-of-the-art BERT based LLMs. We evaluate the proposed techniques using Exact Match, partial Reciprocal Rank, and F1 score, and expert judgment.
Author Region Identification for Urdu Language
(Library Information Services, COMSATS University Islamabad, Lahore Campus, 2020) Rehman Akram Chughtai; FA18-RCS-007; LHR TP 6407; Dr. Rao Muhammad Adeel Nawab
Author region identification is a subtask of author profiling which intends to identify the region of the author, based on the written text. Identification of author region may improve content recommendation, security, reduction in cyber-crimes due to its different applications in fake profile detection, content recommendation, sales and marketing, and forensic linguistics. Lot of work has been done on several author profiling tasks in English, Arabic and other European languages but no significant work found on author region identification task using Urdu language. To fill this gap, we proposed a large benchmark corpus and methods to predict author’s region using Urdu language articles. The proposed corpus consists of 12,100 Urdu articles written by authors from different regions. These articles are categorized into three different regions i.e. Pakistan, India and others. There are 5000 Urdu articles for Pakistan region, 5000 Urdu articles for India region and 2100 articles for Others category. For each article, article web links, author name and domain of each article is mentioned for future research purposes. To demonstrate how our proposed corpus can be used for the development and evaluation of author region identification for Urdu, I applied different classical machine learning and deep learning approaches. For classical machine learning, I used word n-grams technique for feature extractions and then applied classifiers i.e. Logistic Regression, SVM, Naïve Bayes, Random Forest, Adaboost and Gradient boosting for multi classification task. For deep learning algorithms, I applied CNN, RNN, LSTM, Bi-LSTM, and GRU on proposed corpus. Evaluation was carried out using accuracy, weighted precision, weighted recall and weighted F1-measure as evaluation measures. Experiments show that the best results with accuracy of 85 % is achieved by classical machine learning approach using SVM classifier on 3000 features with word n-gram [n ranges from 1 to 3], while for deep learning approach CNN with 1 layer outclass other classifiers with accuracy of 96 %. Deep learning CNN method performed best among all classical machine learning and deep learning methods for author region identification task on proposed Urdu articles corpus.
Cross Genre Author Profiling Using Semantic Tagger Based Approach
(Library Information Services, COMSATS University Islamabad, Lahore Campus, 2017) Humaira Muqdes; FA15-RCS-023; LHR TP 7000; Dr. Rao Muhammad Adeel Nawab
Author profiling is a task to predict one or more traits of an author (e.g. age, gender etc.), from his/her written text. The collaborative environments have significantly increased the chance of fake author profiles, particularly over the social media. The field of automatically detecting an author’s profile from written text has potential applications in marketing, forensics, security, detecting fake profiling and harassment cases. This research work aims to explore the problem of cross genre author profiling, in which training dataset is in one genre and test dataset is in another genre. We explored three different methods for cross-genre author profiling: (1) stylometry based approach, (2) content based approach, (3) semantic tags based approach and (4) combination of previous three approaches. As far as we are aware semantic tags based approach and combination of approaches has not been previously used for cross genre author profiling problem. Evaluation was carried out using four benchmark author profiling corpora: (1) PAN-AP-14 Social Media Corpus, (2) PAN-AP-14 Hotel Reviews corpus, (3) PAN-AP-14 Blogs Corpus and (4) PAN-AP-16 Twitter Corpus. Accuracy was used an evaluation measure. Above mentioned techniques were applied for both same and cross genre problems. For same genre, best results were obtained using content based technique on PAN-AP 14-Hotel Reviews Corpus (Accuracy =55.26) for Age (Accuracy= 66.59) for gender content based techniques for PAN-AP-Blogs Corpus (Accuracy = 67.57) for age, Semantic tagger based technique (Accuracy = 78.38) for gender. Content based technique on PAN-AP-14-social media Corpus (Accuracy =55.26) for Age (Accuracy= 78.95) for gender. Content based technique on PAN-AP-16-Twitter Corpus (Accuracy =54.76) for Age (Accuracy= 74.42) for gender. For Cross Genre Training Corpus PAN-14-Blogs and testing Corpus PAN-14-Hotel reviews for age (Accuracy = 56.96) for gender training Corpus is PAN-16-Twitter (Accuracy= 58.97). Content Based Technique for Training Corpus PAN-14-Hotel reviews and testing Corpus PAN-14-Blogs for age (Accuracy = 40.81) for gender Stylistic based for training Corpus is PAN-14-Blogs (Accuracy= 59.86). Content based technique for Training Corpus PAN-14-Social Media and testing Corpus PAN-14-Blogs for age (Accuracy = 40.81) for gender Content based technique training Corpus is PAN-14-Blogs (Accuracy= 59.86). Stylistic based technique for PAN-16- Twitter training corpus and testing corpus PAN-14-Blogs (Accuracy = 40.81) for age. Content based technique for gender where testing corpus is PAN-14-Blogs (61.22) for gender.
Cross-genre Multi-label Emotion Classification on Mono-lingual and Code-mixed Texts
(Library Information Services, COMSATS University Islamabad, Lahore Campus, 2023) Ahmad Mahmood; LHR TP 8472; FA21-RCS-008; Dr. Rao Muhammad Adeel Nawab
This research focuses on This research focuses on Emotion Classification (EC) in textual data and explores the problem of Cross-genre Multi-Label Emotion Classification (CGMLEC). EC involves detecting the emotional attitude conveyed by text, while Multi-Label Emotion Classification aims to assign relevant emotion labels that accurately reflect the author's state of mind. In previous studies most of the work has been done using Same-genre Multi-label Emotion Classification [2], [14]–[16]. Whereas there is only one article that solves the problem of CGMLEC, but the language is the same. However, the problem of CGMLEC has not explored using mono-lingual (English tweets) and code-mixed (SMS messages). Secondly, this research study has Developed, Applied, Evaluated, and Compared Classical Machine Learning, Deep Learning and Transfer Learning based methods. Whereas in Transfer Learning methods we have implied 5 distinct sentence transformers (all-distilroberta-v1, all-MiniLM-L12-v2, all-mpnet-base-v2, facebook-dpr-question_encoder-single-nq-base, and LaBSE), Also the Feature Fussion based approach is applied which combines the features of all the transformers and then the Machine Learning models are applied. The proposed transformers-based approach outperforms with an F1 score of 0.3333 using all-distilroberta-v1 transformer. ) in textual data and explores the problem of Cross-genre Multi-Label Emotion Classification (CGMLEC). EC involves detecting the emotional attitude conveyed by text, while Multi-Label Emotion Classification aims to assign relevant emotion labels that accurately reflect the author's state of mind. In previous studies most of the work has been done using Same-genre Multi-label Emotion Classification [2], [14]–[16]. Whereas there is only one article that solves the problem of CGMLEC, but the language is the same. However, the problem of CGMLEC has not explored using mono-lingual (English tweets) and code-mixed (SMS messages). Secondly, this research study has Developed, Applied, Evaluated, and Compared Classical Machine Learning, Deep Learning and Transfer Learning based methods. Whereas in Transfer Learning methods we have implied 5 distinct sentence transformers (all-distilroberta-v1, all-MiniLM-L12-v2, all-mpnet-base-v2, facebook-dpr-question_encoder-single-nq-base, and LaBSE), Also the Feature Fussion based approach is applied which combines the features of all the transformers and then the Machine Learning models are applied. The proposed transformers-based approach outperforms with an F1 score of 0.3333 using all-distilroberta-v1 transformer.
Cross-Lingual Abstractive Text Summarization for English-Urdu Language Pair
(Library Information Services, COMSATS University Islamabad, Lahore Campus, 2021) Duaa-e-Nadeem; SP20-RCS-001; LHR TP 7600; Dr. Rao Muhammad Adeel Nawab
Cross-Lingual Abstractive Text Summarization (CLATS) is a process to generate an abstractive summary in Language B (target language) of the given document in Language A (source language). CLATS has the potential to allow cross-lingual information access for millions of people around the world who do not speak or understand languages with a broad online presence and it makes documents originally written in native languages available to a wide audience. The popularity of digital documents has resulted in a rapid increase in the amount of information available online. This massive volume of data, however, cannot be manually analyzed. A summary of a big written document allows individuals to quickly grasp the topic's material without reading the entire document. CLATS plays a vital role in news summarization, literature summarization, timeline/event summarization and summarization of Wikipedia or any information source articles for better understanding. This research aims at exploring the problem of CLATS. Previously, CLATS systems have been developed for different language pair like Chinese-English, German-English, Arabic-English and vice versa but this problem is not explored for English-Urdu language pair, even though Urdu is a widely spoken South Asian language. The twofold aim of this research is to fill the gap: (1) develop a large benchmark corpus CLATS for English-Urdu language pair and (2) develop, apply, evaluate and compare state-of the-art deep learning techniques on proposed corpus for CLATS for English-Urdu language pair. For developing CLATS-EU corpus, News Summary and All the News are used as base corpus. The following steps were taken to create the proposed corpus: (1) automatic translation of base corpus (2) manual inspection and correction of testing data of base corpus (3) corpus standardization After development of CLATS-EU corpus, Deep learning sequence to sequence techniques (RNN, LSTM, GRU, BRNN, BLSTM and BGRU) are developed by using CLATS-EU corpus. Then RNN, LSTM and GRU models are implemented by applying attention mechanism. ROUGE1, ROUGE2 and ROUGEL are used as evaluation x measures to evaluate the performance of the techniques that we have developed. For CLATS, GRU gave the best results on our proposed CLATS-EU corpus
Cross-Lingual Paraphrase Detection for English Urdu Language Pa
(Library Information Services, COMSATS University Islamabad, Lahore Campus, 2021) Nida Waheed; SP19-RCS-025; LHR TP 7598; Dr. Rao Muhammad Adeel Nawab
The main aim of the Cross-Lingual Paraphrase Detection (CLPD) task is to identify whether a text pair (text 01 in language A and text 02 in language B) are paraphrased of one another or not. With the massive increase in technological advancements in digital media and internet facilities, the availability of digital content across multiple languages has become free as well as easy to access. Besides paraphrasing, there has been a rise in cross-lingual paraphrasing since the development of advanced text processing tools. Moreover, the facility of translating existing texts from the original language into any other desired language using freely available translation tools has worsened the problems and triggered an increase in cross-lingual paraphrasing. Previously, there has been plenty of work done for the cross-lingual paraphrase detection in different languages but the downside is that under-resourced languages are still neglected. Urdu is a widely spoken language, despite the fact, the work done for the Urdu language in the CLPD domain is notably less because of the complexity and challenges of the Urdu language. It still comes in the category of low-resource languages. The research in the text analysis domain is mostly focused on well-resourced languages such as English and other European languages. The CLPD problem has formerly been explored for English-Czech, English German, English-Spanish, English Chinese, etc. language pairs. However, the CLPD problem has not yet been studied and reported for the English-Urdu language pair. Although, Cross-Lingual Paraphrasing is now being used enormously in various domains such as content writing, blog posts, and academics, especially in research works related to different languages. To overcome this research gap, a benchmark corpus and approaches are presented through this study for CLPD. The first objective of this research work is to develop a large benchmark Cross-Lingual Paraphrase Detection Corpus (CLPD-EU) for English-Urdu language pair at the sentence level by using a semi-automatic translation approach. For the creation of the CLPD-EU corpus, an existing corpus named Microsoft Research Paraphrase x Corpus (MRPC) is used. The second chief objective of this research work is to make a thorough comparison of classical machine learning (Bilingual Dictionary Based Approaches), Cross-lingual Word Embedding Based Approaches, Transfer Learning (Sentence Transformers Based Approaches) Approaches, and Feature Fusion Approaches on our proposed corpus. The results exhibited that our proposed Feature Fusion Approach 1 has shown notable performance for CLPD. The evaluation of all these approaches is carried out using the weighted average of Precision, Recall, and F1 scores.
Detecting Urdu Semantic Textual Similarity through Word and Sentence Embedding Techniques
(Library Information Services, COMSATS University Islamabad, Lahore Campus, 2019) Muhammad Farhat Ullah; FA17-RCS-019; LHR TP 5625; Dr. Rao Muhammad Adeel Nawab
Semantic Textual Similarity (STS) evaluates the degree to which two sentences or short texts are semantically proportional to one another. STS is one of the significant prob lems in the field of Natural Language Processing (NLP). Text reuse and plagiarism detection are famous examples of STS. STS has three types; (1) Monolingual STS: if the source and suspicious short texts are in the same language. (2) Multilingual STS: if the source is in one language and sus picious short text is more than two languages. (3) Cross-lingual STS: if the source in one language and suspect is in the other language, often translation of each other. STS could be found several levels, for example, word, sentence, paragraph, and document level. Urdu is one of the low resource languages. It’s the National Language of Pakistan, also widely spoken and used in electronic, print media of Pakistan, India, and Bangladesh. The main aim of this thesis is to develop techniques that measure STS as paragraph level for the Urdu language. Our thesis aims to develop and investigate the new feature extraction techniques to ad dress the problem of STS for Urdu. We divide it into three layers, (1) Train Word and Sentence Embedding models on Urdu datasets, (2) Apply these new feature extraction techniques to extract feature from Urdu short text pairs, and (3) Apply machine learn ing classification algorithms for Urdu STS. In the first step we train Word and Sentence Embedding models on Urdu datasets. By using these embedding models, we extract the word, and sentence embedding features from pre-processed Urdu Short Text Reuse Corpus (USTRC) short texts. After that, we find the cosine similarity between these extracted feature vectors, then apply classification algorithms on similarity to classify short texts into verbatim, paraphrased and independently written. In word embedding techniques, we used Word2Vec, GloVe, and FastText with Addition, Average, and Mul tiplication Functions also we explored Smooth Inverse (SI) and Term Frequency (TF) weighted word embedding techniques. In sentence embedding techniques we have used two unsupervised (sent2vec and LASER) and two supervised (InferSent and BERT) ix techniques. We have used seven machine learning algorithms to classify similarity score including, Naive Bayes (NB), Support Vector Machine (SVM), Logistic Regression (LR), Random Forest (RF), K-Nearest Neighbour (KNN), Gradient Boost (GB), Multi-Layer Percep tron (MLP). To evaluate these classifiers, we apply F1 measure. We got best F1 measure = 0.68, 0.75, 0.92, 0.70 by using sent2vec sentence embedding technique with GB and MLP classifiers.
Developing a Gold Standard Benchmark Corpus and Methods for Multi-label Emotion Classification for Urdu Language
(Library Information Services, COMSATS University Islamabad, Lahore Campus, 2023) Azqa Abdul Ghani; SP22-RCS-025; LHR TP 8685; Dr. Rao Muhammad Adeel Nawab
Emotion Classification (EC) involves detecting the emotional tone conveyed by text, while Multi-Label Emotion Classification (MLEC) aims to assign relevant emotion labels that accurately reflect the author's state of mind. This research focuses on MLEC in Urdu text. In previous studies, only one work has been done for the problem of MLEC, including the development of a dataset based on Urdu Tweets using six emotions [5]. However, the problem of MLEC has not been explored for 12 emotions and diverse genres. To fulfill this research gap, this research study has developed a large benchmark corpus of 16,011 Urdu reviews, manually annotated with 12 emotions and gathered from 9 different genres. Furthermore, this research study has Developed, Applied, Evaluated, and Compared Classical Machine Learning, Deep Learning and Transfer Learning based methods. Whereas in Transfer Learning methods, we have implied 6 distinct sentence transformers (xlm-r-100langs-bert-base-nli-stsb-mean-tokens, xlm-r-100langs-bert-base nli-mean-tokens, gtr-t5-large, all-mpnet-base-v2, distiluse-base-multilingual-cased-v2, and LaBSE), Also the Combined Feature-based approach is applied which averages the features of all the columns of the sentence transformer model and then the Machine Learning models are applied. The proposed transformers-based approach outperforms with an F1 score of 0.4562 using the ‘xlm-r-100langs-bert-base-nli-stsb-mean-tokens’ sentence transformer
Developing a Performance Evaluation and Course Recommendation System for Student
(Library Information Services, COMSATS University Islamabad, Lahore Campus, 2021) Chaudary Atif Raza Warraich , Asghar Afzal , Muham; SP17-BSE-128 , SP17-BSE-050 , SP17-BSE-059; Dr. Rao Muhammad Adeel Nawab; LHR TP 7120
Recommenders systems are widely used in a number of field, especially in e-commerce. In past few years, these systems also occupy the major areas of e-learning systems. We will be carrying out detailed analysis on student performance and course recommendation in this final year project. The input to our system is the student marks and our system will fetch all the relative facts and figures that will lead to check the student performance and on this performance, we will recommend a course to student. We will apply various pre-processing techniques to prepare the data for detailed analytics. An efficient and effective Dashboard will be developed to show student performance and recommended courses to the student.
Developing a Speech Recognition System for the Urdu Language
(Library Information Services, COMSATS University Islamabad, Lahore Campus, 2021) Muhammad Ibtesam Arshad; FA17-BSE-061; Dr. Rao Muhammad Adeel Nawab; LHR TP 7015
Automatic Speech Recognition Systems (ASR) are used to convert the acoustic signals which are caught through the microphones into the sequence of words. The Automatic Speech Recognition Systems (ASR) empower the machines to react correctly, reliably, and effectively to human speech or voice and offer helpful and important services to the users. The interaction with computers or gadgets is faster, simpler, and easier, through voice or speech instead of typing through the keyboard or console, so the people will prefer Speech Recognition Systems. Speech Recognition Systems will facilitate the users in workplaces, in marketing, in banking, in health care, in language learning, and in the education field also. This project aims to design and the implementation a Speech Recognition System for the Urdu Language. In this first phase, we have collected data for the Urdu Language. To build a general Speech Recognizer, a huge amount of data is needed. In the second phase, we have implemented our System. In the final phase, we have trained and tested our Model. This model is Similar to Deep Speech 2 Models. The evaluation is carried out using Word Error Rate and Character Error Rate.
Developing a Urdu Lemmatizer Using Dictionary Look-up Approach
(Library Information Services, COMSATS University Islamabad, Lahore Campus, 2020) Saima Shaukat; FA18-RCS-009; LHR TP 5987; Dr. Rao Muhammad Adeel Nawab
Lemmatization is the process of finding the root form of a word. Lemmatization is used as preprocessing step for most of the NLP applications. Lemmatization is im portant part of NLP applications which involve the use of root form instead of using derivational forms of words. Text mining, information retrieval systems, document clustering, search engines and text categorizations are few of the applications where lemmatization process is used. Urdu is a widely spoken language in the world, but very less work has been done on developing basic NLP tools for this language, one of them is Urdu lemmatizer. Since Urdu is a morphologically rich language and has words with many inflectional and derivational forms, development of an efficient lemmatizer is a challenging task and it will be useful for many Urdu NLP applications. These challenges help us to choose the customized and more appropriate approach for Urdu lemmatization which is dictionary-based approach. The main aim of this research work is to develop an Urdu lemmatizer. To develop the efficient Urdu lemmatizer, we used dictionary look-up approach. Majority of exist ing lemmatizers are of in English and other European languages. Most of the existing lemmatizers are rule based and also the information about parts of speech tags is ig nored. Development of large dictionaries for Urdu Language is also missing in existing work. To fulfill this gap, the main goals of the study are: (1) Develop a large bench mark corpus for the Urdu language and (2) Explore the relationship between parts of speech tag and lemmatizer (3) Developed the standard approach for Urdu lemmatizer. The first major contribution of this study is the development of a large benchmark Urdu dictionary. The proposed corpus was developed in the following steps: (1) In the first step we collected data from two different sources: urmono corpus and Wikipedia dump (2) Preprocess and tokenize the collected data (3) extract the frequency of each word using python script (4) Selection of most frequent words (5) Assign parts of speech ix tags to selected words (6) manually annotate data (assign lemma to each word) (7) De velop a benchmark corpus and save in a standard csv format. Second major contribution of this research work is exploring the relationship be tween POS tag and lemma of a word. Train the POS tagger on collected Wikipedia dump and then used in the development of Urdu lemmatizer because user can enter input with or without POS tags. If user will enter input without POS tag, tagger will assign most frequent used tag to the word before generating lemma of a word. As a third major contribution of this research work, proposed a dictionary-based approach for Urdu lemmatizer. As the Urdu is morphological rich language; it also includes many Persian, Turkish and Arabic words. So, for the efficient lemmatizer, dictionary-based approach is more suitable. The evaluation was carried out using two techniques (1) With-POS-DLA (2) Without-POS-DLA. We achieved accuracy of 66.79% and 76.44% respectively. The best result is achieved by using Without-POS DLA technique.
Development of Large Scale English-Urdu Machine Translation Corpus for Statistical and Neural Machine Translation Systems
(Library Information Services, COMSATS University Islamabad, Lahore Campus, 2019) Moodser Hussain; FA17-RCS-005; LHR TP 5623; Dr. Rao Muhammad Adeel Nawab
The process of automatically converting the text of one natural language into an other is known as Machine Translation (MT). The two state-of-the-art machine translation techniques are Statistical Machine Translation (SMT) and Neural Ma chine Translation (NMT). In both MT techniques, a large-scale parallel aligned corpus of source & target language text is the most basic requirement to develop quality models. For the Urdu language, there are minimal parallel resources de veloped until now, and these existing resources are not enough to produce some quality translation systems. This research study is majorly focusing on: (1) the development of a large-scale semi-automatically aligned English-Urdu parallel corpus of more than 150k sen tences by collecting the data from different domains. (2) the compilation of large sized monolingual data for English and Urdu language by merging existing cor pora. (3) the development of a character-separated parallel corpus for supervised transliteration (by using Wikipedia titles) (4) the evaluation of impact generated by language model and monolingual corpus on translation quality (by using differ ent data and word-order in language models training). (5) development of Phrase based SMT (PB-SMT) and Recurrent Neural Network-based NMT (RNN-NMT) systems to evaluate and compare their performance on the proposed parallel cor pus. (6) the usage of supervised transliteration and ‘minimum error rate training’ to enhance the quality of SMT. To evaluate the performance of MT systems, ‘Bilingual Evaluation Understudy (BLEU)’ is used as an evaluation measure. The best-achieved scores for English to-Urdu translations are 37.39 and 32.70 by using PBSMT and RNN-NMT, re spectively. While for Urdu-to-English translations, PBSMT and RNN-NMT yield the best score of 28.20 and 30.63 respectively
Image Caption Generation for Urdu Language – Data and Methods
(Library Information Services, COMSATS University Islamabad, Lahore Campus, 2021) Nimra Shafiq; SP20-RCS-010; LHR TP 7602; Dr. Rao Muhammad Adeel Nawab
Image Captioning (IC) generates a suitable caption for the given query image. The focus of IC is understanding the contextual meaning of visual contents in an image and then describing them in semantically and syntactically correct description in the natural language. The most salient part of an image is described in a sentence to increase the visual understanding of people. It has many applications like navigation guidance for blind persons, image-based optimized search engines, visual context understanding, early babyhood education, image description in an e-commerce site, and image indexing. Previously most of the work done on Image Captioning was in English and other high resource languages like Chinese, Arabic, and Hindi, etc. To the best of our knowledge, the Urdu language has not been explored previously for Urdu Image Captioning (UIC). The primary reason for this is the unavailability of benchmark corpora for UIC. So, the foremost aim of this research is to develop a benchmark Image Captioning corpus for the Urdu language (known as UIC-21 corpus). The proposed corpus is based on the Fliker-8k English corpus. It is developed by translating the English image caption into Urdu image captions using a semi-automatic translation approach. In this approach, automated translation tools are used to automatically translate the image captions in Urdu language and then a manual inspection and correction approach is used to improve the quality of the captions. To demonstrate how UIC systems can be developed and evaluated on our proposed UIC-21 corpus, we applied deep learning techniques (CNN for images and RNN, LSTM, and GRU for Urdu textual data) and state-of-the-art transfer learning techniques (InceptionV3 with ImageNet weights and VGG16 with ImageNet weights for images, and pre-trained word2vec urduvec and W2V models for textual data) for UIC systems, which are further evaluated using BELU-1, BLEU-2, BLEU-3, and BLEU-4. Their experimental results show that the transfer learning InceptionV3-GRU with the pretrained word embedding method achieved the highest BLEU-1 score of 0.60 on our proposed UIC-21 corpus
Information Retrieval System For Urdu Language
(Library Information Services, COMSATS University Islamabad, Lahore Campus, 2021) HASSAN NAWAZ , SAJID DIN MUHAMMAD; SP17-BCS-125 , SP17-BCS-059; Dr. Rao Muhammad Adeel Nawab; LHR TP 7188
Internet Searches in the learning environment differs from traditional “Query” or “known objects” research: Students or common person are typically looking for or master a specific topic, and search engines help them to find that information. This report presents IRS (Information Retrieval System), an exploratory search engine for learning resources based on Wikipedia articles: provides set of topics in Urdu language and relationship between Wikipedia articles provides suggestion to Urdu native speakers to deepen research in a specific field. IRS indexes several hundred thousand Wikipedia articles. This report also contains the results of evaluation of IRS show that the exploratory research paradigm using Wikipedia knowledge is a promising method that can be used to create information retrieval systems for learning environments.
Multi-Label Author Profiling on Multi-Lingual Text
(Library Information Services, COMSATS University Islamabad, Lahore Campus, 2020) Samra Kanwal; SP19-RCS-011; LHR TP 6418; Dr. Rao Muhammad Adeel Nawab
Author profiling is the task of author attributes classification where the main aim is to predict the profile and demographic features of an author which includes age group, gender, region, personality, etc., by examining the written content of the author. There are different promising applications of author profiling including security, forensic analysis, and identification of harassing text messages, marketing intelligence, and fake profile identification. In literature, the majority of the studies have been carried for single-label author profiling i.e., predicting only one single label at a time. There are very few studies available on multi-label author profiling on mono-lingual text, i.e., predicting more than one label at a time. However, the problem of multi-label author profiling has not been completely explored for multi-lingual text. The main objective of this research work is to explore the problem of multi-label author profiling on multi lingual text (English and Roman Urdu). For this purpose, the aim is to predict four author traits including gender, age, education, and language as a multi-label task using three state-of-the-art methods: (1) Content based Methods (N-gram models for both word and character), (2) Deep Learning Approaches (CNN, LSTM, BI-LSTM, GRU, and BI-GRU) and (3) Transfer Learning Approaches (BERT, and XLNET). The evaluations were carried out on three benchmark multi-lingual datasets, RUEN-AP-17, SMS–AP–18, and BT-AP-19. After extensive experimentation and comparison, the results show that the content-based method outperforms the deep learning and transfer learning methods for multi-label author profiling tasks on all multi-lingual corpora used in this study. On the RUEN AP-17 corpus the best results (Accuracy = 0.71, F1-measure = 0.65) were obtained using the word tri-gram model with the Naïve Bayes classifier. On SMS–AP–18 corpus the best results (Accuracy = 0.74, F1-measure = 0.69) were obtained using word uni gram model using support vector machine with one-vs-rest and one-vs-one classifiers, and on BT-AP-19 corpus the best results (Accuracy = 0.74, F1-measure = 0.69) were obtained using word bi-gram model using support vector machine with one-vs-rest and one-vs-one classifiers
Smart Reply Suggestion for an Email in Multiple Languages – Data and Metho
(Library Information Services, COMSATS University Islamabad, Lahore Campus, 2021) Nazia Jehan; FA19-RCS-002; LHR TP 7295; Dr. Rao Muhammad Adeel Nawab
Smart Reply Suggestion (SRS) is a novel end-to-end system that refers toward the method of suggesting the three short email responses. These responses are themselves complete short email replies, suggested in diverse form for (monolingual) email i.e. English. In literature, mostly the model is specifically designed for monolingual emails like English. There is a paucity of state-of-the-art techniques that handle both monolingual and multilingual email at a time. Moreover, there is a paucity of benchmark corpus in multiple languages-based emails. This research has tried to overcome the limitations of an earlier study of the SRS system by proposing a novel method for the SRS system which handle both the monolingual and multilingual email at a time. In this research, we treat this problem as a multilabel text classification problem. we developed the " (ML-SRS) Email corpus collected from academia, consisting of three different languages including English, Roman Urdu and a combination of Roman Urdu and English emails (text) and developed the list of smart response prepared to annotate that dataset manually the inter-annotator agreement calculated and then standardized the dataset in .CSV format. For our proposed SRS system, we considered two approaches. In the first approach, we considered content-based n-gram approach at word and character level with the combination of multi-label classifier i.e. One Vs Rest, Label Powerset, with eight classical machine learning classifiers i.e. Linear SVC, Logistic Regression etc. In the second approach, we consider LSTM, CNN, GRU, Bi GRU, Bi-LSTM models. Both approaches are evaluated on Micro Precision, Micro Recall, Micro F1 score and Hamming loss. However, Label Powerset with the combination of Linear SVC performed well as compared to others
Text Reuse Detection for the Urdu Language at Lexical Level
(Library Information Services, COMSATS University Islamabad, Lahore Campus, 2023) Ayesha Noreen; FA21-RCS-026; LHR TP 8461; Dr. Rao Muhammad Adeel Nawab
Text reuse is a process of creating new text(s) from pre-existing text(s). In recent years, Urdu Text Reuse Detection (UTRD) has gained the attention of researchers because the text is readily available in digital format all over the internet and can be copied or paraphrased from another source without proper attribution, which makes it easier to reuse but hard to detect. In previous studies, the problem of UTRD has been explored at the sentence level, sentence/passage level, and document level, along with benchmark corpora and methods. However, the problem of UTRD has not been explored at the Lexical level with respect to corpora and methods. To fulfill this research gap, this research study has developed a large benchmark manually annotated corpus of 22,184 text pairs at two levels of rewrite: 1) Derived = 8,660 and 2) Non-Derived = 13,524. Secondly, this research study has developed, applied, evaluated, and compared baseline approaches (Uni-gram Overlap and Word Embedding-based approaches) with proposed transformer-based approaches on the proposed UTRD-Lex-22 corpus. The proposed transformers-based approach (using a combination of 7 different Sentence Transformers including: distiluse-base-multilingual-cased-v1, LaBSE, paraphrase-multilingual-MiniLM-L12-v2, paraphrase-multilingual-mpnet-base-v2, xlm-r-100langs-bert-base-nli-mean-tokens, xlm-r-bert- base-nli-stsb-mean-tokens, xlm-r-distilroberta-base-paraphrase-v1) outperforms with an F1 score of 0.69 compared to the best results obtained using feature fusion approach (F1 = 0.70).
Toxic Comment Classification of Roman Urdu Text
(Library Information Services, COMSATS University Islamabad, Lahore Campus, 2019) Waheed Abbas; FA17-RCS-013; LHR TP 5624; Dr. Rao Muhammad Adeel Nawab
We are living in an era of technology where most of us have easy access to the inter net. Due to internet bloom, the use of social media and discussion forum, in particular for communication, increase tremendously in recent years. But this advancement also opens doors for trolls who poison these social media and forums by their abusive behav ior toward others. Detection of toxic language online is becoming an important issue in recent years. A toxic comment is defined as rude, unreasonable, or disrespectful com ments which likely to make you leave a conversation. Current methods for handling toxicity online are often heavily dependent on manual moderation, and not scalable enough to handle growing users day by day. In this thesis, we investigated the problem of toxic language detection for Roman Urdu (transliteration of Urdu in English) text because Roman Urdu is widely used to communicate online in Pakistani and Indian community. Various work has been done to detect toxicity in the English language, but due to unavailability of large Roman Urdu corpora, there is no work done so far in Roman Urdu. The aim of this thesis work is two folds: first is to develop large scale corpus for detection of toxic language in Roman Urdu, second is to apply different proven classical machine learning and state-of-the-art deep learning methods on the developed corpus. In the first step, we collected data from the YouTube comment section to develop Roman Urdu corpus. The final corpus consists of 10,863 unique comments where we have 5,887 toxic and 4,976 non-toxic comments. We performed manual annotation of corpus and got it reviewed from another annotator. We applied proven classical ma chine learning algorithms like Logistic Regression, K-Nearest Neighbors, Naive Bayes, Random Forest, Support Vector Machine and Multi-layer Perceptron along with state of-the-art deep learning methods like Long Short-Term Memory, Bidirectional Long Short-Term Memory, and Convolutional Neural Networks. Best results obtained from Random Forest with character 4-gram (Accuracy = 0.964, Precision = 0.980, Recall = 0.953, F1 = 0.966).
Toxicity Detection in Speech
(Library Information Services, COMSATS University Islamabad, Lahore Campus, 2021) Mohsin Nazakat , Samiya Ijaz, Ghulam Rasool; FA18 BCS-052, FA18 BCS-143, FA18 BCS-112; Dr. Rao Muhammad Adeel Nawab
Speech is a fundamental source of communication. We express our ideas, thoughts, and sentiments through speech. What we say (our speech) has a very significant effect on our personalities and the personalities of our listeners. A positive speech spreads positivity in society, gives peace to its speaker and pleasure to the listeners. On the other hand, negative or toxic speech cause great damage to our society. It not only has a negatively effects the minds of the speaker and listener but also spoils their personalities. Ultimately toxic conversation reshapes human personality in such a way that might leads speakers and listeners to get involved in toxic actions. So, it is momentous to put a stop to these toxic conversations in our society. According to the best of our knowledge, there is no existing system that can classify a digital audio conversation (digital speech) as toxic or normal. To fulfil this gap, we have built an administrative android mobile app that will notify the parent whenever an audio speech is classified as toxic. This will allow parent to guide their children to avoid having such toxic conversations. Using the digital audio data of day-to-day conversations, we have trained a Machine Learning model, and using the REST API call, we will send audio to the model and in response, we get a prediction of whether the audio is toxic or normal. In the case of Toxicity, a notification is dispatched to the parent to notify them that a toxic conversation is detected.
Urdu Short Text Reuse Detection Using Sentence Transformers
(Library Information Services, COMSATS University Islamabad, Lahore Campus, 2021) Hamza Hafeez; SP19-RCS-014; LHR TP 7589; Dr. Rao Muhammad Adeel Nawab
The process of creating new text(s) using an existing one is called Text Reuse. The goal of text reuse detection is to identify whether a new text is created using a source text or not. A source text can be reused at different rewrite levels to generate either partially (Paraphrased) or fully (Verbatim) reused texts pairs. The Paraphrased Text Reuse shows the relationship between two text pairs conveying the same idea differently. Paraphrased text reuse text is difficult to detect since different edit operations can be used to generate it which varies considerably according to the context where it is being used. In literature, the Paraphrase Text Reuse Detection problem has been explored for English as well as some other European languages, but it has not been investigated for the Urdu language because of the non-availability of the Urdu paraphrase text reuse detection corpus. Therefore, to fulfill this gap this study aims to develop a large-scale gold standard benchmark Urdu Paraphrase Text Reuse Corpus (UPTRC) for the Urdu text reuse detection with special emphasizes on paraphrase text reuse detection for Urdu language. In addition, we have applied the baseline approaches (N-Gram Based Approaches) and proposed other approaches including Word Embedding Based Approaches, Sentence Transformers Based Approaches, and other Feature Fusion Approaches for Urdu Paraphrase Text Reuse Detection (UPTRD). However, Urdu Short Text Reuse Detection (USTRD) has been explored at the sentence level, passage level and document level along with different approaches that have been previously applied on Urdu Short Text Reuse Corpus (USTRC). Nonetheless, the USTRD task has only been explored with lexical and semantic approaches on USTRC but has not been explored with state-of-the-art Sentence Transformers Based Approaches and Word Embedding Based Approaches. The Sentence Transformers Based Approaches are used in a variety of applications and natural language processing tasks, especially for semantic text similarity and duplicate pair detection. But the Sentence Transformers has not been explored for Urdu text reuse detection. So, there is a need to explore the Sentence Transformers Based Approaches on the USTRC corpus for the task of USTRD. In this study, along with the corpus construction, we have applied the N-Gram Based Approaches as baseline and proposed new approaches including Word Embedding Based Approaches, Sentence Transformers Based Approaches, and Feature Fusion Approaches for our proposed UPTRC corpus. We have proposed applied compared and evaluated same approaches for USTRC corpus as well as our proposed corpus to investigate the performance of our proposed approaches. The best results obtained were the weighted average of F1 = 0.855607 by our proposed Feature Fusion Approaches for the task of UPTRD on our proposed UPTRC corpus. In the case of the USTRC corpus, our proposed Feature Fusion Approaches still outperformed the baseline approaches for ternary classification with F1 = 0.715164. The performance of the proposed approaches for both corpora is evaluated and reported using the weighted average of Precision (P), Recall (R), and F1 evaluation measures.