Data preparation is the first and a very important step towards any Large Language Model (LLM) development. This paper introduces an easy-to-use, extensible, and scale-flexible open-source data preparation toolkit called Data Prep Kit (DPK). DPK is architected and designed to enable users to scale their data preparation to their needs. With DPK they can prepare data on a local machine or effortlessly scale to run on a cluster with thousands of CPU Cores. DPK comes with a highly scalable, yet extensible set of modules that transform natural language and code data. If the user needs additional transforms, they can be easily developed using extensive DPK support for transform creation. These modules can be used independently or pipelined to perform a series of operations. In this paper, we describe DPK architecture and show its performance from a small scale to a very large number of CPUs. The modules from DPK have been used for the preparation of Granite Models [1] [2]. We believe DPK is a valuable contribution to the AI community to easily prepare data to enhance the performance of their LLM models or to fine-tune models with Retrieval-Augmented Generation (RAG).

Recently, the needs of data required for data analysis are becoming more diversified, and research on data extraction and analysis methods has been continuously made in order to effectively respond to various needs. Process mining is a solution that analyzes various system logs built by companies or healthcare institutions so that they can be used for process improvement. From the process model extracted from the system logs, it is possible not only to grasp the exact flow of the current business process, but also to acquire additional information such as repetitive execution of activities in the process where the bottleneck occurs in the business process flow. The manufacturing industry has made great efforts to improve the process management, and as many companies are paying attention to big data these days, various data-related technologies are emerging in the healthcare industry as well to properly provide patients with the care needed. Process mining tools allow users to pull data by programming in a process mining query language using the APIs provided with the process mining tool, or by manually creating reusable analytical documents using user friendly tool. However, these tasks require the users to be familiar with the query language APIs and understand the data model and its relationships with respect to creating analytical documents. This paper proposes a methodology that allows users to easily extract desired data through natural language interface, which relieves nonprofessional users of the burden of programming in a process mining query language. The process mining query engine with natural language interface presented in this paper consists of four major components. Among them, the natural language processing pipeline that not only extracts intermediate representation of entities used when constructing a process mining query language report from natural language queries, but also effectively extracts a query hint from the context of natural language query. The query hint is used to select a process-specific function from the library that fits the context of the user query while transforming a natural language query into a process mining query report. The method proposed in this study has the advantage of being able to roughly grasp the process state for the user just by entering a query in natural language. The proposed system provides users with four query process options. That is, the user 1) retrieves intermediate representation of entities and query hints from the NLP pipeline, 2) retrieves the process mining query language from the query language generator, 3) submits the query language to the process mining engine and execute the query, 4) retrieves description of intermediate representation of entity and query hints in natural language to confirm that the query is processed correctly. The contents proposed in this paper were constructed and executed, and the query reports in process mining query language programmatically generated by the proposed query engine were also executed in a process mining engine and the query results were verified.

Database management systems offer an efficient way of managing huge amount of data such as financial and healthcare data and he data retrieval from databases requires knowledge of Structured Query Language (SQL). In this paper, an Automatic SQL Generation System is proposed to help users who are inexperienced in querying database with SQL. The proposed SQL generation system reads formatted data items in the query report from the user and converts the data items into SQL statements programmatically with the help of a data model that is pulled from a database. The SQL generation system can handle simple queries composed of a query block with a SELECT statement as well as complex queries composed of multiple query blocks containing multiple SELECT statements. The proposed system is integrated with an NLIDB (Natural Language Interface for Database) system to translate data items (or tokens) extracted from queries in natural languages into SQL query language, and the system is also integrated and adapted with various types of databases and use cases that include financial and healthcare use cases. The experiment results show that the proposed system correctly handles user queries in natural language just like any other neural model based system and more importantly, the proposed SQL generation engine generates SQL queries without syntactic problems with various databases for all queries.

A large number of time series forecasting models including traditional statistical models, machine learning models and more recently deep learning have been proposed in the literature. However, choosing the right model along with good parameter values that performs well on a given data is still challenging. Automatically providing a good set of models to users for a given dataset saves both time and effort from using trial-and-error approaches with a wide variety of available models along with parameter optimization. We present AutoAI for Time Series Forecasting (AutoAI-TS) that provides users with a zero configuration (zero-conf) system to efficiently train, optimize and choose best forecasting model among various classes of models for the given dataset. With its flexible zero-conf design, AutoAI-TS automatically performs all the data preparation, model creation, parameter optimization, training and model selection for users and provides a trained model that is ready to use. For given data, AutoAI-TS utilizes a wide variety of models including classical statistical models, Machine Learning (ML) models, statistical-ML hybrid models and deep learning models along with various transformations to create forecasting pipelines. It then evaluates and ranks pipelines using the proposed T-Daub mechanism to choose the best pipeline. The paper describe in detail all the technical aspects of AutoAI-TS along with extensive benchmarking on a variety of real world data sets for various use-cases. Benchmark results show that AutoAI-TS, with no manual configuration from the user, automatically trains and selects pipelines that on average outperform existing state-of-the-art time series forecasting toolkits.

Discovering temporal lagged and inter-dependencies in multivariate time series data is an important task. However, in many real-world applications with big data, such as commercial cloud management or predictive maintenance in manufacturing, such dependencies can be time-variant and non-linear, which makes it more challenging to extract such dependencies through traditional methods like Granger causality or statistical models. In this work, we present a novel deep learning model that uses multiple layers of adapted gated recurrent units (GRUs) for discovering both time lagged behaviors and inter-timeseries dependencies, representing them in the form of directed weighted graphs. Each individual time series is first analyzed by a pair of encoding-decoding GRUs in order to discover the time lagged dependencies and representing its samples as high dimensional vectors. Such vectors collected from all component time series are then analyzed by a decoding network component to discover inter-dependencies across all time series while forecasting their next values in the multivariate time series. Though the discovery of two types of dependencies are separated at two levels of our neural network, they are tightly connected and jointly trained in an end-to-end manner. With this joint training, improvement in learning of one type of dependency immediately impacts the learning process of the other one, leading to the overall highly accurate dependencies discovery. We empirically test our model on synthetic time series data in which the exact form of dependencies are known. We also practically evaluate its performance on two real-world applications, (i) dynamic multivariate performance monitoring data with high volatility from a commercial cloud provider and, (ii) multivariate time series generated by sensors for a manufacturing plant. We show how our approach is capable of capturing these dependency behaviors via intuitive and interpretable dependency graphs and use them to generate forecasting values.

Analysis of legal and contract documents often requires both the discovery of document structure, as well as the accurate identification of important elements such as party (buyer, supplier), nature (obligation, right) and category (warranties, delivery, etc.). Hence, exploring novel features that lead to better element classification accuracy as well as better document structure discovery is particularly important. In this paper, we develop and present novel unsupervised learning techniques to analyze a large scale corpus of contract documents with the goal of learning and deriving new features to enhance classification accuracy over the elements of interest, and to extract relevant features leading to meaningful clusterings over contract structures. Particularly, we propose a novel t-threshold autoencoder neural network that flexibly controls the number of active neurons in response to sentences of different lengths at the network's input. Such an adaptive sparseness threshold enforces competition and specialization among encoding neurons and hence results in better features learning. We also present an extension of the convolutional neural network classifier that allows for the incorporation of these novel augmented features and show that higher classification accuracies over various classes of contract elements can be achieved. We further present a practical pipeline of deriving features from contract documents along with a clustering solution based on the K-means algorithm that leads to the separation among different types of sentences in the contract documents. We empirically demonstrate the performance of our developed techniques on a novel data corpus of Software Procurement contracts.

Change detection in system behavior and its root cause detection is essential for many large-scale systems such as, manufacturing plants, in order to keep systems running uninterrupted and avoid costly machine breakdown via predictive maintenance. In this paper, we present a novel graph based technique that uses time variant interdependencies and lagged dependencies among different components of a system to detect changes in the system behavior. We further find the root causes for these detected changes by pointing out the component and its historical values that are responsible for initiating and changing the system to the new state. The proposed mechanism extracts these time variant dependencies using a deep learning system and converts them into weighted directed graphs and applies graph based techniques for change detection. For each detected change, our system uses graph theoretic techniques to uncover the root causes for the change. Such a mechanism provides us with valuable insights about the inner workings of a system from a different perspective as opposed to traditional techniques for root cause analysis that directly apply statistical models to the time series data for analysis. Experimental results on real manufacturing data show, that we can detect changes in system behavior and accurately identify the root causes in almost 71% of the cases for which we have the ground truth. For synthetic data, our system can correctly identify root causes in 87% of the cases.

Automated Ontology Learning systems are nowadays practical and used in a variety of domains. By using these systems, subject matter experts (SMEs) and ontology designers can readily construct very large ontologies consisting of tens of thousands of concepts and their relations based on a corpus. However, ontologies of this size make it extremely challenging for such SMEs to understand and further tune these ontologies. Prior studies have proposed techniques for concept ranking based solely on the analysis of the structure of the ontology graphs. In this paper, we propose a novel approach, which further exploits a word-level summarization technique applied to the source documents used to generate the ontology. Using the document summarization technique, we devise features that measure concept importance based on source documents where concepts are extracted. We demonstrate the effectiveness of our approach by comparing with existing ranking methods and by devising a scalable evaluation process inspired from the document retrieval domain.

Performance monitoring of cloud-native applications that consist of several micro-services involves the analysis of time series data collected from the infrastructure, platform, and application layers of the cloud software stack. The analysis of the runtime dependencies amongst the component microservices is an essential step towards performing cloud resource management, detecting anomalous behavior of cloud applications, and meeting customer Service Level Agreements (SLAs). Finding such dependencies is challenging due to the non-linear nature of interactions, aberrant data measurements and lack of domain knowledge. In this paper, we propose a novel use of the modeling capability of Long-Short Term Memory (LSTM) recurrent neural networks, which excel in capturing temporal relationships in multi-variate time series data and being resilient to noisy pattern representations. Our proposed technique looks into the LSTM model structure, to uncover dependencies amongst performance metrics, which were learned during training. We further apply this technique in three monitoring use cases, namely finding the strongest performance predictors, discovering lagged/temporal dependencies, and improving the accuracy of forecasting for a given metric. We demonstrate the viability of our approach, by comparing the results of our proposed method in the three use cases with those obtained from previously proposed methods, such as Granger causality and the classical statistical time series analysis models, such as ARIMA and Holt-Winters. For our experiments and analysis, we use performance monitoring data collected from two sources: a controlled experiment involving a sample cloud application that we deployed in a public cloud infrastructure and cloud monitoring data collected from the monitoring service of an operational, public cloud service provider.

Apache Spark enables fast computations and greatly accelerates analytics applications by efficiently utilizing the main memory and caching data for later use. At its core Apache Spark uses data structures called RDDs (Resilient Distributed Datasets) to give a unified view to the distributed data. However, the data represented in the RDDs remain unencrypted which can result in leakage of confidential data produced or processed by applications. Apache Spark persists (unencrypted) RDDs to the disk storage under various circumstances including but not limited to caching, RDD checkpointing and data spill during the data shuffling operations, etc. This lack of security makes Apache Spark unsuitable for processing of sensitive information that should be secured at all times. Moreover, RDDs stored in the main memory are prone to main-memory attacks such as RAM-scrapping. In this paper, we propose and develop solutions to fill-up such security lapses in the current Apache Spark framework. We present three different approaches to incorporate security in the Apache Spark framework. These approaches are designed to limit the exposure of unencrypted data during data processing, caching and data spill to disk. We use combination of cryptographic splitting and encryption to secure data stored and spilled by Apache Spark, both to the disk as well as to the main memory. Our approaches provide strong security by incorporating combination of Information Dispersal Algorithm (IDA) and Shamir's Perfect Secret Sharing (PSS). Extensive experimentation show that with appropriately chosen parameters our security approaches provide high security at a performance penalty between 10%-25%.

Cloud service providers are offering the popular Hadoop analytics platform following an "as-a-service" model, i.e. clusters of machines in their cloud infrastructures pre-configured with Hadoop software. Such offerings lower the cost and complexity of deploying a comparable system on-premises, however security considerations and in particular data confidentiality hamper wider adoption of such services by enterprises that handle data of sensitive nature. In this paper, we describe our efforts in providing security for data-at-rest (i.e. data that is stored) when Hadoop is offered as a cloud service. We analyze the requirements and architecture for such service and further describe a new distributed file system that we developed for Hadoop called SDFS, towards supporting this premise. We analyze parameter tuning for SDFS and through experiments on a real test-bed we evaluate its performance. We further present simulation results that explore the parameter space and can guide tuning.