arXiv subjects

Mourad Ouzzani

Publications and source records attributed to Mourad Ouzzani.

At least 19 recordsLinked to original sources

arXiv2026

Fanar 2.0: Arabic Generative AI Stack

We present Fanar 2.0, the second generation of Qatar's Arabic-centric Generative AI platform. Sovereignty is a first-class design principle: every component, from data pipelines to deployment infrastructure, was designed and operated entirely at QCRI, Hamad Bin Khalifa University. Fanar 2.0 is a story of resource-constrained excellence: the effort ran on 256 NVIDIA H100 GPUs, with Arabic having only ~0.5% of web data despite 400 million native speakers. Fanar 2.0 adopts a disciplined strategy of data quality over quantity, targeted continual pre-training, and model merging to achieve substantial gains within these constraints. At the core is Fanar-27B, continually pre-trained from a Gemma-3-27B backbone on a curated corpus of 120 billion high-quality tokens across three data recipes. Despite using 8x fewer pre-training tokens than Fanar 1.0, it delivers substantial benchmark improvements: Arabic knowledge (+9.1 pts), language (+7.3 pts), dialects (+3.5 pts), and English capability (+7.6 pts). Beyond the core LLM, Fanar 2.0 introduces a rich stack of new capabilities. FanarGuard is a state-of-the-art 4B bilingual moderation filter for Arabic safety and cultural alignment. The speech family Aura gains a long-form ASR model for hours-long audio. Oryx vision family adds Arabic-aware image and video understanding alongside culturally grounded image generation. An agentic tool-calling framework enables multi-step workflows. Fanar-Sadiq utilizes a multi-agent architecture for Islamic content. Fanar-Diwan provides classical Arabic poetry generation. FanarShaheen delivers LLM-powered bilingual translation. A redesigned multi-layer orchestrator coordinates all components through intent-aware routing and defense-in-depth safety validation. Taken together, Fanar 2.0 demonstrates that sovereign, resource-constrained AI development can produce systems competitive with those built at far greater scale.

cs.CL↗

arXiv2026

OMNIA: Closing the Loop by Leveraging LLMs for Knowledge Graph Completion

Knowledge Graphs (KGs) are widely used to represent structured knowledge, yet their automatic construction, especially with Large Language Models (LLMs), often results in incomplete or noisy outputs. Knowledge Graph Completion (KGC) aims to infer and add missing triples, but most existing methods either rely on structural embeddings that overlook semantics or language models that ignore the graph's structure and depend on external sources. In this work, we present OMNIA, a two-stage approach that bridges structural and semantic reasoning for KGC. It first generates candidate triples by clustering semantically related entities and relations within the KG, then validates them through lightweight embedding filtering followed by LLM-based semantic validation. OMNIA performs on the internal KG, without external sources, and specifically targets implicit semantics that are most frequent in LLM-generated graphs. Extensive experiments on multiple datasets demonstrate that OMNIA significantly improves F1-score compared to traditional embedding-based models. These results highlight OMNIA's effectiveness and efficiency, as its clustering and filtering stages reduce both search space and validation cost while maintaining high-quality completion.

cs.DB↗

arXiv2026

Fanar-Sadiq: A Multi-Agent Architecture for Grounded Islamic QA

Large language models (LLMs) can answer religious knowledge queries fluently, yet they often hallucinate and misattribute sources, which is especially consequential in Islamic settings where users expect grounding in canonical texts (Qur'an and Hadith) and jurisprudential (fiqh) nuance. Retrieval-augmented generation (RAG) improves grounding, however, a single retrieve-then-generate pipeline is insufficient for diverse Islamic queries, including verbatim scripture, citation-grounded guidance, and rule-constrained computations such as zakat and inheritance. To address these challenges, we present Fanar-Sadiq, a bilingual Arabic-English Islamic QA system built on a multi-agent, tool-augmented architecture. It is a core component of the Fanar AI platform. Fanar-Sadiq routes Islamic queries to specialized modules within an agentic tool architecture. It supports intent-aware routing, retrieval-grounded fiqh answers with normalized citations and verification traces, exact verse lookup with quotation validation, and deterministic Sunni zakat and inheritance calculators with madhhab-sensitive branching. We evaluate the end-to-end system on public Islamic QA benchmarks and show strong effectiveness and efficiency. It is publicly accessible through an API and Web application and has received over 1.9M accesses in less than a year (https://api.fanar.qa/docs).

cs.CL↗

arXiv2025

HCT-QA: A Benchmark for Question Answering on Human-Centric Tables

Tabular data embedded in PDF files, web pages, and other types of documents is prevalent in various domains. These tables, which we call human-centric tables (HCTs for short), are dense in information but often exhibit complex structural and semantic layouts. To query these HCTs, some existing solutions focus on transforming them into relational formats. However, they fail to handle the diverse and complex layouts of HCTs, making them not amenable to easy querying with SQL-based approaches. Another emerging option is to use Large Language Models (LLMs) and Vision Language Models (VLMs). However, there is a lack of standard evaluation benchmarks to measure and compare the performance of models to query HCTs using natural language. To address this gap, we propose the HumanCentric Tables Question-Answering extensive benchmark (HCTQA) consisting of thousands of HCTs with several thousands of natural language questions with their respective answers. More specifically, HCT-QA includes 1,880 real-world HCTs with 9,835 QA pairs in addition to 4,679 synthetic HCTs with 67.7K QA pairs. Also, we show through extensive experiments the performance of 25 and 9 different LLMS and VLMs, respectively, in an answering HCT-QA's questions. In addition, we show how finetuning an LLM on HCT-QA improves F1 scores by up to 25 percentage points compared to the off-the-shelf model. Compared to existing benchmarks, HCT-QA stands out for its broad complexity and diversity of covered HCTs and generated questions, its comprehensive metadata enabling deeper insight and analysis, and its novel synthetic data and QA generator.

cs.IR↗

arXiv2025

Fanar: An Arabic-Centric Multimodal Generative AI Platform

We present Fanar, a platform for Arabic-centric multimodal generative AI systems, that supports language, speech and image generation tasks. At the heart of Fanar are Fanar Star and Fanar Prime, two highly capable Arabic Large Language Models (LLMs) that are best in the class on well established benchmarks for similar sized models. Fanar Star is a 7B (billion) parameter model that was trained from scratch on nearly 1 trillion clean and deduplicated Arabic, English and Code tokens. Fanar Prime is a 9B parameter model continually trained on the Gemma-2 9B base model on the same 1 trillion token set. Both models are concurrently deployed and designed to address different types of prompts transparently routed through a custom-built orchestrator. The Fanar platform provides many other capabilities including a customized Islamic Retrieval Augmented Generation (RAG) system for handling religious prompts, a Recency RAG for summarizing information about current or recent events that have occurred after the pre-training data cut-off date. The platform provides additional cognitive capabilities including in-house bilingual speech recognition that supports multiple Arabic dialects, voice and image generation that is fine-tuned to better reflect regional characteristics. Finally, Fanar provides an attribution service that can be used to verify the authenticity of fact based generated content. The design, development, and implementation of Fanar was entirely undertaken at Hamad Bin Khalifa University's Qatar Computing Research Institute (QCRI) and was sponsored by Qatar's Ministry of Communications and Information Technology to enable sovereign AI technology development.

cs.CL↗

arXiv2023

RetClean: Retrieval-Based Data Cleaning Using Foundation Models and Data Lakes

Can foundation models (such as ChatGPT) clean your data? In this proposal, we demonstrate that indeed ChatGPT can assist in data cleaning by suggesting corrections for specific cells in a data table (scenario 1). However, ChatGPT may struggle with datasets it has never encountered before (e.g., local enterprise data) or when the user requires an explanation of the source of the suggested clean values. To address these issues, we developed a retrieval-based method that complements ChatGPT's power with a user-provided data lake. The data lake is first indexed, we then retrieve the top-k relevant tuples to the user's query tuple and finally leverage ChatGPT to infer the correct value (scenario 2). Nevertheless, sharing enterprise data with ChatGPT, an externally hosted model, might not be feasible for privacy reasons. To assist with this scenario, we developed a custom RoBERTa-based foundation model that can be locally deployed. By fine-tuning it on a small number of examples, it can effectively make value inferences based on the retrieved tuples (scenario 3). Our proposed system, RetClean, seamlessly supports all three scenarios and provides a user-friendly GUI that enables the VLDB audience to explore and experiment with the system.

cs.DB↗

arXiv2021

RPT: Relational Pre-trained Transformer Is Almost All You Need towards Democratizing Data Preparation

Can AI help automate human-easy but computer-hard data preparation tasks that burden data scientists, practitioners, and crowd workers? We answer this question by presenting RPT, a denoising auto-encoder for tuple-to-X models (X could be tuple, token, label, JSON, and so on). RPT is pre-trained for a tuple-to-tuple model by corrupting the input tuple and then learning a model to reconstruct the original tuple. It adopts a Transformer-based neural translation architecture that consists of a bidirectional encoder (similar to BERT) and a left-to-right autoregressive decoder (similar to GPT), leading to a generalization of both BERT and GPT. The pre-trained RPT can already support several common data preparation tasks such as data cleaning, auto-completion and schema matching. Better still, RPT can be fine-tuned on a wide range of data preparation tasks, such as value normalization, data transformation, data annotation, etc. To complement RPT, we also discuss several appealing techniques such as collaborative training and few-shot learning for entity resolution, and few-shot learning and NLP question-answering for information extraction. In addition, we identify a series of research opportunities to advance the field of data preparation.

cs.LG↗

arXiv2019

Dataset-On-Demand: Automatic View Search and Presentation for Data Discovery

Many data problems are solved when the right view of a combination of datasets is identified. Finding such a view is challenging because of the many tables spread across many databases, data lakes, and cloud storage in modern organizations. Finding relevant tables, and identifying how to combine them is a difficult and time-consuming process that hampers users' productivity. In this paper, we describe Dataset-On-Demand (DoD), a system that lets users specify the schema of the view they want, and have the system find views for them. With many underlying data sources, the number of resulting views for any given query is high, and the burden of choosing the right one is onerous to users. DoD uses a new method, 4C, to reduce the size of the view choice space for users. 4C classifies views into 4 classes: compatible views are exactly the same, contained views present a subsumption relationship, complementary views are unionable, and contradictory views have incompatible values that indicate fundamental differences between views. These 4 classes permit different presentation strategies to reduce the total number of views a user must consider. We evaluate DoD on two key metrics of interest: its ability to reduce the size of the choice space, and the end to end performance. DoD finds all views within minutes, and reduces the number of views presented to users by 2-10x.

cs.DB↗

arXiv2019

DeepER -- Deep Entity Resolution

Entity resolution (ER) is a key data integration problem. Despite the efforts in 70+ years in all aspects of ER, there is still a high demand for democratizing ER - humans are heavily involved in labeling data, performing feature engineering, tuning parameters, and defining blocking functions. With the recent advances in deep learning, in particular distributed representation of words (a.k.a. word embeddings), we present a novel ER system, called DeepER, that achieves good accuracy, high efficiency, as well as ease-of-use (i.e., much less human efforts). For accuracy, we use sophisticated composition methods, namely uni- and bi-directional recurrent neural networks (RNNs) with long short term memory (LSTM) hidden units, to convert each tuple to a distributed representation (i.e., a vector), which can in turn be used to effectively capture similarities between tuples. We consider both the case where pre-trained word embeddings are available as well the case where they are not; we present ways to learn and tune the distributed representations. For efficiency, we propose a locality sensitive hashing (LSH) based blocking approach that uses distributed representations of tuples; it takes all attributes of a tuple into consideration and produces much smaller blocks, compared with traditional methods that consider only a few attributes. For ease-of-use, DeepER requires much less human labeled data and does not need feature engineering, compared with traditional machine learning based approaches which require handcrafted features, and similarity functions along with their associated thresholds. We evaluate our algorithms on multiple datasets (including benchmarks, biomedical data, as well as multi-lingual data) and the extensive experimental results show that DeepER outperforms existing solutions.

cs.DB↗

arXiv2019

LocationSpark: In-memory Distributed Spatial Query Processing and Optimization

Due to the ubiquity of spatial data applications and the large amounts of spatial data that these applications generate and process, there is a pressing need for scalable spatial query processing. In this paper, we present new techniques for spatial query processing and optimization in an in-memory and distributed setup to address scalability. More specifically, we introduce new techniques for handling query skew, which is common in practice, and optimize communication costs accordingly. We propose a distributed query scheduler that use a new cost model to optimize the cost of spatial query processing. The scheduler generates query execution plans that minimize the effect of query skew. The query scheduler employs new spatial indexing techniques based on bitmap filters to forward queries to the appropriate local nodes. Each local computation node is responsible for optimizing and selecting its best local query execution plan based on the indexes and the nature of the spatial queries in that node. All the proposed spatial query processing and optimization techniques are prototyped inside Spark, a distributed memory-based computation system. The experimental study is based on real datasets and demonstrates that distributed spatial query processing can be enhanced by up to an order of magnitude over existing in-memory and distributed spatial systems.

cs.DB↗

arXiv2019

Technical Report: Optimizing Human Involvement for Entity Matching and Consolidation

An end-to-end data integration system requires human feedback in several phases, including collecting training data for entity matching, debugging the resulting clusters, confirming transformations applied on these clusters for data standardization, and finally, reducing each cluster to a single, canonical representation (or "golden record"). The traditional wisdom is to sequentially apply the human feedback, obtained by asking specific questions, within some budget in each phase. However, these questions are highly correlated; the answer to one can influence the outcome of any of the phases of the pipeline. Hence, interleaving them has the potential to offer significant benefits. In this paper, we propose a human-in-the-loop framework that interleaves different types of questions to optimize human involvement. We propose benefit models to measure the quality improvement from asking a question, and cost models to measure the human time it takes to answer a question. We develop a question scheduling framework that judiciously selects questions to maximize the accuracy of the final golden records. Experimental results on three real-world datasets show that our holistic method significantly improves the quality of golden records from 70% to 90%, compared with the state-of-the-art approaches.

cs.DB↗

arXiv2019

Data Curation with Deep Learning [Vision]

Data curation - the process of discovering, integrating, and cleaning data - is one of the oldest, hardest, yet inevitable data management problems. Despite decades of efforts from both researchers and practitioners, it is still one of the most time consuming and least enjoyable work of data scientists. In most organizations, data curation plays an important role so as to fully unlock the value of big data. Unfortunately, the current solutions are not keeping up with the ever-changing data ecosystem, because they often require substantially high human cost. Meanwhile, deep learning is making strides in achieving remarkable successes in multiple areas, such as image recognition, natural language processing, and speech recognition. In this vision paper, we explore how some of the fundamental innovations in deep learning could be leveraged to improve existing data curation solutions and to help build new ones. In particular, we provide a thorough overview of the current deep learning landscape, and identify interesting research opportunities and dispel common myths. We hope that the synthesis of these important domains will unleash a series of research activities that will lead to significantly improved solutions for many data curation tasks.

cs.DB↗

arXiv2018

Reuse and Adaptation for Entity Resolution through Transfer Learning

Entity resolution (ER) is one of the fundamental problems in data integration, where machine learning (ML) based classifiers often provide the state-of-the-art results. Considerable human effort goes into feature engineering and training data creation. In this paper, we investigate a new problem: Given a dataset D_T for ER with limited or no training data, is it possible to train a good ML classifier on D_T by reusing and adapting the training data of dataset D_S from same or related domain? Our major contributions include (1) a distributed representation based approach to encode each tuple from diverse datasets into a standard feature space; (2) identification of common scenarios where the reuse of training data can be beneficial; and (3) five algorithms for handling each of the aforementioned scenarios. We have performed comprehensive experiments on 12 datasets from 5 different domains (publications, movies, songs, restaurants, and books). Our experiments show that our algorithms provide significant benefits such as providing superior performance for a fixed training data size.

cs.DB↗

arXiv2018

Unsupervised String Transformation Learning for Entity Consolidation

Data integration has been a long-standing challenge in data management with many applications. A key step in data integration is entity consolidation. It takes a collection of clusters of duplicate records as input and produces a single "golden record" for each cluster, which contains the canonical value for each attribute. Truth discovery and data fusion methods, as well as Master Data Management (MDM) systems, can be used for entity consolidation. However, to achieve better results, the variant values (i.e., values that are logically the same with different formats) in the clusters need to be consolidated before applying these methods. For this purpose, we propose a data-driven method to standardize the variant values based on two observations: (1) the variant values usually can be transformed to the same representation (e.g., "Mary Lee" and "Lee, Mary") and (2) the same transformation often appears repeatedly across different clusters (e.g., transpose the first and last name). Our approach first uses an unsupervised method to generate groups of value pairs that can be transformed in the same way (i.e., they share a transformation). Then the groups are presented to a human for verification and the approved ones are used to standardize the data. In a real-world dataset with 17,497 records, our method achieved 75% recall and 99.5% precision in standardizing variant values by asking a human 100 yes/no questions, which completely outperformed a state of the art data wrangling tool.

cs.DB↗

arXiv2018

A large scale study of SVM based methods for abstract screening in systematic reviews

A major task in systematic reviews is abstract screening, i.e., excluding, often hundreds or thousand of, irrelevant citations returned from a database search based on titles and abstracts. Thus, a systematic review platform that can automate the abstract screening process is of huge importance. Several methods have been proposed for this task. However, it is very hard to clearly understand the applicability of these methods in a systematic review platform because of the following challenges: (1) the use of non-overlapping metrics for the evaluation of the proposed methods, (2) usage of features that are very hard to collect, (3) using a small set of reviews for the evaluation, and (4) no solid statistical testing or equivalence grouping of the methods. In this paper, we use feature representation that can be extracted per citation. We evaluate SVM-based methods (commonly used) on a large set of reviews ($61$) and metrics ($11$) to provide equivalence grouping of methods based on a solid statistical test. Our analysis also includes a strong variability of the metrics using $500$x$2$ cross validation. While some methods shine for different metrics and for different datasets, there is no single method that dominates the pack. Furthermore, we observe that in some cases relevant (included) citations can be found after screening only 15-20% of them via a certainty based sampling. A few included citations present outlying characteristics and can only be found after a very large number of screening steps. Finally, we present an ensemble algorithm for producing a $5$-star rating of citations based on their relevance. Such algorithm combines the best methods from our evaluation and through its $5$-star rating outputs a more easy-to-consume prediction.

cs.IR↗

arXiv2017

Human-Centric Data Cleaning [Vision]

Data Cleaning refers to the process of detecting and fixing errors in the data. Human involvement is instrumental at several stages of this process, e.g., to identify and repair errors, to validate computed repairs, etc. There is currently a plethora of data cleaning algorithms addressing a wide range of data errors (e.g., detecting duplicates, violations of integrity constraints, missing values, etc.). Many of these algorithms involve a human in the loop, however, this latter is usually coupled to the underlying cleaning algorithms. There is currently no end-to-end data cleaning framework that systematically involves humans in the cleaning pipeline regardless of the underlying cleaning algorithms. In this paper, we highlight key challenges that need to be addressed to realize such a framework. We present a design vision and discuss scenarios that motivate the need for such a framework to judiciously assist humans in the cleaning process. Finally, we present directions to implement such a framework.

cs.DB↗

arXiv2017

Pattern-Driven Data Cleaning

Data is inherently dirty and there has been a sustained effort to come up with different approaches to clean it. A large class of data repair algorithms rely on data-quality rules and integrity constraints to detect and repair the data. A well-studied class of integrity constraints is Functional Dependencies (FDs, for short) that specify dependencies among attributes in a relation. In this paper, we address three major challenges in data repairing: (1) Accuracy: Most existing techniques strive to produce repairs that minimize changes to the data. However, this process may produce incorrect combinations of attribute values (or patterns). In this work, we formalize the interaction of FD-induced patterns and select repairs that result in preserving frequent patterns found in the original data. This has the potential to yield a better repair quality both in terms of precision and recall. (2) Interpretability of repairs: Current data repair algorithms produce repairs in the form of data updates that are not necessarily understandable. This makes it hard to debug repair decisions and trace the chain of steps that produced them. To this end, we define a new formalism to declaratively express repairs that are easy for users to reason about. (3) Scalability: We propose a linear-time algorithm to compute repairs that outperforms state-of-the-art FD repairing algorithms by orders of magnitude in repair time. Our experiments using both real-world and synthetic data demonstrate that our new repair approach consistently outperforms existing techniques both in terms of repair quality and scalability.

cs.DB↗

arXiv2014

Similarity Group-by Operators for Multi-dimensional Relational Data

The SQL group-by operator plays an important role in summarizing and aggregating large datasets in a data analytic stack.While the standard group-by operator, which is based on equality, is useful in several applications, allowing similarity aware grouping provides a more realistic view on real-world data that could lead to better insights. The Similarity SQL-based Group-By operator (SGB, for short) extends the semantics of the standard SQL Group-by by grouping data with similar but not necessarily equal values. While existing similarity-based grouping operators efficiently materialize this approximate semantics, they primarily focus on one-dimensional attributes and treat multidimensional attributes independently. However, correlated attributes, such as in spatial data, are processed independently, and hence, groups in the multidimensional space are not detected properly. To address this problem, we introduce two new SGB operators for multidimensional data. The first operator is the clique (or distance-to-all) SGB, where all the tuples in a group are within some distance from each other. The second operator is the distance-to-any SGB, where a tuple belongs to a group if the tuple is within some distance from any other tuple in the group. We implement and test the new SGB operators and their algorithms inside PostgreSQL. The overhead introduced by these operators proves to be minimal and the execution times are comparable to those of the standard Group-by. The experimental study, based on TPC-H and a social check-in data, demonstrates that the proposed algorithms can achieve up to three orders of magnitude enhancement in performance over baseline methods developed to solve the same problem.

cs.DB↗