The University of Texas System

Washington University

International Business Machines Corporation

Johns Hopkins University

Gautam Das

Nan Zhang

Mukesh Kumar Mohania

Sandor Szalay

Ullas Balan Nambiar

Hao Hua Huang

As file systems reach the petabytes scale, users and administrators are increasingly interested in acquiring high-level analytical information for file management and analysis. Two particularly important tasks are the processing of aggregate and top-k queries which, unfortunately, cannot be quickly answered by hierarchical file systems such as ext3 and NTFS. Existing pre-processing based solutions, e.g., file system crawling and index building, consume a significant amount of time and space (for generating and maintaining the indexes) which in many cases cannot be justified by the infrequent usage of such solutions. User interests can often be sufficiently satisfied by approximate (i.e., statistically accurate) answers. A just-in-time sampling-based system can, after consuming a small number of disk accesses, produce extremely accurate answers for a broad class of aggregate and top-k queries over a file system without the requirement of any prior knowledge. The system is efficient, accurate and scalable.

Just-in-time analytics on large file systems

A just-in-time sampling-based system can, after consuming a small number of disk accesses or queries, produce extremely accurate answers for a broad class of aggregate and top-k queries over a file system or database without the requirement of prior knowledge. The system is efficient, accurate, and scalable. The system performs aggregate estimations of a hidden database through its web interface by employing techniques that use a small number of queries to produce unbiased estimates with small variance. It conducts domain discovery over a hidden database through its web interface by employing techniques which provide effective guarantees on the effectiveness of domain discovery. Systems and methods enhance forms used by mobile devices to access hidden databases. It employs data analytics to improve the usage of form fields, including providing context-sensitive auto-completion suggestions, highlighting selections in drop-down boxes and eliminating suggestions in drop-down boxes.

Just-in-time analytics on large file systems and hidden databases

Methods, systems and computer products perform cost estimate to determine an efficient approach to answer a query according to one of several unified query plans. One unified query plan involves querying an unstructured database, referencing a unified index, and probing a structured database based on matches discovered in the unified index. The results of the unstructured database query are used to lookup entries in a unified index associated with the structured database. Then the structured database is probed by querying only the subset of the structured database gleaned from the unstructured database query.

Supporting unified querying over autonomous unstructured and structured databases

Growing community of inventors

Irving, TX, United States of America

Gautam Das