TL;DR
This paper introduces a novel coding scheme for Chinese characters that combines phonetic spelling with ideographic properties, enabling efficient input and storage using conventional keyboards. The scheme achieves a high degree of uniqueness (99.5%) for a dictionary of 10,000 characters and can be represented internally using just 32 bits.
What Problem Does This Paper Address?
Inputting Chinese characters into computers has historically been a challenge due to the vast number of characters and the limitations of standard keyboards. Traditional methods often rely on complex input schemes or specialized hardware. This paper addresses the need for a user-friendly and efficient system that leverages existing keyboard infrastructure while accurately representing the unique nature of Chinese characters, which are a combination of phonetic and semantic (ideographic) components. The authors aim to create a coding scheme that is both practical for input and compact for internal computer storage.
Methods
The proposed coding scheme is a two-part system:
- Ideographic Property String: This part consists of three symbols derived from the character's radicals. Radicals are fundamental graphical components of Chinese characters. The paper details a process of grouping 456 radicals into 50 sets, each associated with a keyboard key. The first three radicals of a character determine this part of the code.
- Spelling String: This part uses the phonetic spelling of the character, represented by Latin alphabets. The authors opt for Latin alphabets over Pinyin for compatibility with standard keyboards. They also introduce a forgiving approach to spelling, allowing for variations (e.g., 'C' vs. 'CH', 'S' vs. 'SH', 'Z' vs. 'ZH') to simplify input, which is then addressed by remedial measures for uniqueness.
The combination of these two strings forms the unique code for each character. For internal representation, the scheme packs the ideographic part into 17 bits and the spelling part into 15 bits, totaling 32 bits per character.
Datasets
The primary "dataset" used in this paper is the dictionary "XianDai HanYu CiDian," which contains approximately ten thousand Chinese characters. The paper focuses on developing a coding scheme for these characters and evaluating its uniqueness and efficiency based on this dictionary. No new benchmark datasets were introduced.
Key Insights and Ideas
- Synergy of Phonetics and Ideographics: The core insight is that combining phonetic spelling with ideographic (radical) information provides a robust and unique way to represent Chinese characters, overcoming the limitations of using either alone.
- Keyboard Compatibility: By using Latin alphabets for spelling and mapping radical groups to keys, the scheme seamlessly integrates with standard typewriter keyboards, making it accessible to a wider user base.
- Forgiving Input for Uniqueness: The allowance for phonetic variations (e.g., 'C' vs. 'CH') simplifies user input, and the paper demonstrates how this non-uniqueness can be resolved through simple remedial measures, such as deleting less common characters or modifying their codes.
- Efficient Internal Representation: The 32-bit internal code is a significant achievement, allowing for compact storage and efficient processing of Chinese characters within computer systems.
Key Results
- High Uniqueness: The coding scheme achieves 99.5% uniqueness for the ten thousand characters in the "XianDai HanYu CiDian" dictionary. The remaining 0.5% (26 pairs) are addressed with suggested remedies to ensure unique codes.
- Average Code Length: Input codes require an average of 5.5 keys, with a maximum of seven keys.
- Internal Representation: Each code is packed into 32 bits for internal storage, comprising 17 bits for the ideographic component and 15 bits for the spelling component.
- Collating Sequence: The internal representation naturally leads to a logical collating sequence, where spelling primarily determines order, followed by the radical string.
Limitations and Future Work
- Arbitrary Radical Grouping: The paper acknowledges that the grouping of radicals into sets and their association with keyboard keys (Table 3) is somewhat arbitrary. This has a subtle impact on the collating sequence, suggesting that further research in this area is necessary to optimize it.
- User Evaluation: The authors plan to evaluate the effectiveness of the coding scheme through experiments with different users to measure coding efficiency, training time, and user reaction. This user acceptance is highlighted as the ultimate measure of the invention's success.
- Further Research on Collating Sequence: More research is needed to refine the set-key association to ensure a more optimal and practical collating sequence.
Summary updated