Character-String Auto-Completion Task

An Character-String Auto-Completion Task is a string generation task that completes the remainder of a character string.

• Context:
  • Input: Text Data.
  • output: Autocompletion Suggestions.
  • Task Requirement(s):
    • a Language Model or a Structured Prediction Model,
    • a Rule-based Expert System (optional) .
  • It can be solved by a Auto-Completion System that implements Auto-Completion Algorithm.
  • It can range from being a Syntactic Autocompletion Task, to being a Semantic Autocompletion Task, to being a Contextual Autocompletion Task.
  • It can range from an Online Autocompletion Task to being an Offline Autocompletion Task.
  • It can range from being a Language Model Based Autocompletion Task to being a Structured Prediction Based Autocompletion Task, to being a Sequence-Based Autocompletion Task.
  • …
• Example(s):
  • a Text Auto-Completion Task such as:
    • a Academic Writing Auto-Completion Task,
    • a Typing Assistant Auto-Completion Task;
  • a Software Code Auto-Completion Task such as:
    • Intelligent Code Completion Task (Bruch et al., 2009),
    • Qt Creator Auto-Completion Task;
  • a Query Auto-Completion Task;
  • a WikiText Auto-Completion Task;
  • a Hand-Drawn Symbol Auto-Completion Task (e.g. Costagliola et al., 2013);
  • an Error-Tolerant Auto-Completion Task (Chaudhuri & Kaushik, 2009).
  • …
• Counter-Example(s):
  • an Artificial Grammar Learning Task,
  • an Autoreplace Task,
  • an Auto-Encoding Task,
  • a Spelling Error Correction Task,
  • a Parsing Task,
  • a Text Wikification Task.
• See: Human-Computer Interaction, Editing System, Language Model, Natural Language Inference System, Natural Language Processing System.

---

---

References

2019

• (Wikipedia, 2019) ⇒ https://en.wikipedia.org/wiki/Autocomplete Retrieved:2019-8-13.
  • Autocomplete, or word completion, is a feature in which an application predicts the rest of a word a user is typing. In graphical user interfaces, users can typically press the tab key to accept a suggestion or the down arrow key to accept one of several.

    Autocomplete speeds up human-computer interactions when it correctly predicts the word a user intends to enter after only a few characters have been typed into a text input field. It works best in domains with a limited number of possible words (such as in command line interpreters), when some words are much more common (such as when addressing an e-mail), or writing structured and predictable text (as in source code editors).

    Many autocomplete algorithms learn new words after the user has written them a few times, and can suggest alternatives based on the learned habits of the individual user.

2013

• (Costagliola et al., 2013) ⇒ Gennaro Costagliola, Mattia De Rosa, and Vittorio Fuccella. (2013). "Investigating Human Performance in Hand-Drawn Symbol Autocompletion". In: Proceedings of the 2013 IEEE International Conference on Systems, Man, and Cybernetics. ISBN:978-1-4799-0652-9 doi:10.1109/SMC.2013.54
  • QUOTE: Auto-completion has proven effective in various text-based applications. Conversely, no experiments have been carried out on the automatic completion of hand-drawn symbols and only a few systems have been presented in the literature for this purpose. Nevertheless, such a feature might be useful in different domains, e.g. for accelerating symbol retrieval and for launching complex gestural commands. In this paper we present a user study aimed at evaluating the human performance in hand-drawn symbol auto-completion.

2009a

• (Chaudhuri & Kaushik, 2009) ⇒ Surajit Chaudhuri, and Raghav Kaushik. (2009). “Extending Autocompletion to Tolerate Errors.” In: Proceedings of the 2009 ACM SIGMOD International Conference on Management of data. ISBN:978-1-60558-551-2 doi:10.1145/1559845.1559919
  • QUOTE: At its core, autocompletion is about suggesting valid completions of a partially entered lookup string with the intention of minimizing and guiding the users typing. The concept of autocompletion is not new. There are various approaches studied in prior work on autocompletion. There is a vast body of work on predictive autocompletion 4, 8, 16 where the idea is to use information retrieval techniques, language models and learning to suggest potential completions.

    In contrast, our focus is on the scenario where there is a table T of strings being looked up and thus completions are suggested based on matches in T. This is the form of autocompletion supported by http://www.amazon.com and Yahoo Finance for instance. The most common form of such autocompletion is based on exact matching. In this section, we formalize some of the key concepts involved in exact autocompletion.

2009b

• (Bruch et al., 2009) ⇒ Marcel Bruch, Martin Monperrus, and Mira Mezini. (2009). “Learning from Examples to Improve Code Completion Systems.” In: Proceedings of the 7th joint meeting of the European software engineering conference and the ACM SIGSOFT symposium on The foundations of software engineering. ISBN:978-1-60558-001-2 doi:10.1145/1595696.1595728

2006

• (Hyvonen & Makela, 2006) ⇒ (2006, September). "Semantic autocompletion". In Asian Semantic Web Conference (pp. 739-751). Springer, Berlin, Heidelberg. DOI:10.1007/11836025_72
  • QUOTE: The idea of autocompletion is to predict what the user is typing in, and to complete the work automatically. The benefits of this simple idea are manyfold: First, the computer helps the user in memorizing the right vocabulary used. Second, typing errors in the input can be minimized. Third, autocompletion speeds up the interaction. A side effect of the idea is that it encourages the usage of long descriptive names and commands that are more understandable to the users. An idea related to autocompletion is autoreplace, where the idea is to use predefined abbreviations in typing and the system automatically replaces these with full-blown strings.

    In order to make the prediction right and as early as possible, the underlying vocabulary must be known, be limited, and the words in the lexicon should differ from each other in terms of the leading characters. These conditions hold in many applications, such as operating system shells, email programs, browsers, etc.

     Autocompletion is used, e.g., in Microsoft’s Intellisense feature of the Visual Studio, where the idea is applied to source code editing (...)

    Traditional autocompletion is based on matching input strings with a list of usable words in a vocabulary. This paper generalizes this approach onto the semantic level(...)

---