Browse Prior Art Database

Automatic Completion and Suggestion of Literature

IP.com Disclosure Number: IPCOM000019256D
Original Publication Date: 2003-Sep-08
Included in the Prior Art Database: 2003-Sep-08
Document File: 1 page(s) / 42K

Publishing Venue

IBM

Abstract

Often times, a user is hindered in writing their documents by the sluggishness of their typing speed. Autocomplete mechanisms are in place that attempt to guess the intent of a user's writing and provide shortcuts to speed up the process of typing common words or phrases. This mechanism is limited, however, in the phrases that are recognized. Described herein is an approach towards remedying that limitation. By providing a full-fledged dictionary, phrasebook and grammar module as resources to an autocomplete methodology, mixed with a little heuristic analysis, an application can perpetually guess at or suggest words and phrases in an attempt to assist the user in completing their document more quickly.

This text was extracted from a PDF file.
This is the abbreviated version, containing approximately 52% of the total text.

Page 1 of 1

Automatic Completion and Suggestion of Literature

      Upon typing out a word, this mechanism will analyze the input and perform a comparison against the contents of its dictionary module. This is feasible because this comparison is made via a running filter against a pre-indexed database. For example, the user intends to type the word "quiet." When the Q key is typed, only words starting with a Q will be examined. Each following key press will further narrow the list of words that can be formed until there are only a few possible choices that remain. For example, after "qui", the dictionary may only have entries for "quiet", "quit", "quiver" and "quick." In this way, a list of possible word selections is narrowed to be below some implementation specific threshold.

    Once the word selection threshold is reached, grammar can be analyzed to determine what would be legal words to use at the given point in the sentence. For example, if the sentence already states "I demand qui", then the dictionary can see that this word will likely be a noun. As such it can eliminate the verb "quit" and the adjective "quick."

    The number of word selections at this point has further been limited. If there is more than one selection still available, then more analysis can be performed. A search within the locality of the document, for example, for synonyms and antonyms or related words can be executed. For example, if the word "arrow" or "bow" were to be found in the previous paragraph or so, then perhaps "quiver" or "quivers" would be suggested. However, if the word "loud", "scream", "noise", etc. were found, then perhaps "quiet" would be suggested....