Text inputting method, apparatus and system based on a cache-based language model and a universal language model

A text inputting method comprising:

• obtaining a user input;
• generating a candidate sentence list according to the user input;
• for each candidate sentence in the candidate sentence list, respectively calculating a standard conditional probability of each word in the candidate sentence according to a universal language model;
• respectively calculating a cache conditional probability of each word in the candidate sentence according to a preconfigured modeling policy, the user input and a pre-cached user input;
• calculating a mixed conditional probability of each word according to the standard conditional probability and the cache conditional probability;
• obtaining an on-screen probability of the candidate sentence according to the mixed conditional probability;
• sorting candidate sentences in the candidate sentence list according to their on-screen probabilities;
• and outputting the sorted candidate sentence list;
• wherein:
• in the sorted candidate sentence list, a sentence with a largest on-screen probability is selected and outputted, and the sentence with the largest on-screen probability is cached;
• the process of respectively calculating a standard conditional probability of each word in the candidate sentence according to the universal language model comprises:
• respectively calculating the standard conditional probability of each word in the candidate sentence according to a pre-created standard Ngram language model, and calculating the cache conditional probability of an ith word in the candidate sentence comprises:
• obtaining a number of times ki that a word sequence which includes the ith word and a predefined constant number of words emerges before the ith word in cached training material;
• obtaining a number of times ki−1 that a word sequence which includes a predefined constant number of words emerges before the ith word in cached training material;
• obtaining a time function value of the ith word;
• and
• calculation a ratio of ki to ki−1, multiplying the ratio with the time function value of the ith word to obtain the cache conditional probability of the ith word in the user input.

A text processing apparatus, comprising:

• one or more processors;
• memory;
• and
• one or more program modules stored in the memory and to be executed by the one or more processors, the one or more program modules including:
• a universal language model module, a cache module, a cache-based language modeling module and a mixed model module;
• wherein the universal language model module is to receive a user input, generate a candidate sentence list according to the user input, calculate a standard conditional probability of each word in the each candidate sentence respectively and output the standard conditional probability to the mixed model module;
• the cache module is to cache a sentence outputted by the mixed model module;
• the cache-based language modeling module is to respectively calculate a cache conditional probability of each word in each candidate sentence according to a preconfigured cache-based language modeling policy, the user input and the sentences cached in the cache module, and output the cache conditional probability to the mixed model module;
• the mixed model module is to calculate a mixed conditional probability of each word according to the standard conditional probability and the cache conditional probability, calculate a sentence probability of each candidate sentence according to the mixed conditional probability, and select and output the sentence with the largest sentence probability;
• and
• the cache-based language modeling module is further configured to calculate a time function value of each word, wherein the cache-based conditional probability of each word is positively related to the time function value of the word, and the time function value is related to a time when the word enters the cache module;
• wherein the time function value is a result obtained by dividing a preconfigured constant by a time interval between an ith word enters the cache area and a word currently inputted by the user.