580 California St., Suite 400
San Francisco, CA, 94104
Key research themes
1. How do self-organizing encoding schemes improve compression performance by adapting to locality of reference?
This research theme investigates encoding schemes that dynamically adjust encoding based on local frequency changes in the data, aiming to outperform traditional static schemes like Huffman coding especially in scenarios exhibiting locality of reference. These schemes embody principles from self-organizing data structures, such as move-to-front heuristics, to adaptively reduce code length for frequently accessed words or symbols over short intervals, thus potentially achieving better compression ratios and processing efficiency.
Key finding: Introduced a self-organizing data compression scheme applying the move-to-front heuristic on word lists, exploiting locality of reference whereby frequently used words are moved to the front and encoded with shorter integers.... Read more
Key finding: Discussed various baseband digital encoding techniques, including Manchester encoding which inherently supports self-synchronization by embedding clocking information within the data transitions, thus improving signal... Read more
Key finding: Provided an integrated finite state machine-based implementation and comparative evaluation of encoding methods such as Manchester, Miller, and FM0, emphasizing their operation in maintaining DC balance, self-clocked data... Read more
2. What are effective algorithmic advancements in lossless text compression exploiting transform and pattern matching techniques?
This theme synthesizes cutting-edge lossless text compression methods that improve compression ratios by combining advanced preprocessing transforms such as Burrows-Wheeler Transform (BWT) with pattern matching and entropy coding like Huffman. It centers on algorithmic strategies that leverage data redundancy patterns, repeated substrings, and statistical symbol probabilities to optimize coding efficiency and reduce output size without loss, relevant to growing demands in bandwidth and storage.
Key finding: Proposed a lossless text compression approach combining BWT with a two-key pattern reduction method targeting frequent consecutive character repetitions, followed by pattern detection and Huffman coding. The method achieves a... Read more
Key finding: Empirically compared coding algorithms including Huffman, arithmetic, and other coders across various alphabet sizes and probability distributions, highlighting trade-offs in compression effectiveness, speed, and memory. The... Read more
Key finding: Developed a formula-based arithmetic coding method that obviates interval recalculation typically required, demonstrating equivalence to conventional approaches through practical compression and decompression of text. This... Read more
3. How can encoding schemes leverage data structure representations to achieve space-efficiency while maintaining query capabilities?
This research domain addresses encoding data structures in a manner that compresses their representation close to the theoretical minimum information content (effective entropy) while still enabling efficient query answering operations. Encoding data structures consider the minimal subset of information needed to correctly respond to queries, potentially allowing original data to be discarded and reducing space requirements drastically, an important avenue for big data and succinct data structure research.
Key finding: Formalized the concept of encoding data structures where only an equivalence class representative encoding is stored sufficient for query answering, reducing the storage space from that needed for the original data. The work... Read more
Key finding: Presented a numeric encoding scheme for concept hierarchies in transactional databases that decreases both time and space complexity for mining association rules. This encoding allows higher abstraction level data... Read more
Key finding: Explained various encoding methods that convert digital data into forms suitable for different transmission channels, addressing signal synchronization and decoding challenges. The coverage includes both baseband digital... Read more