Generating Chinese language banners
View Patent ↗Embodiments are disclosed for automatically generating a banner given a first scroll sentence and a second scroll sentence of a Chinese couplet. The first and/or second scroll sentence can be generated by an automatic computer system or by a human (e.g., manually generated and then provided as input to an automated banner generation system) or obtained from any source (e.g., a book) and provided as input. In one embodiment, an information retrieval process is utilized to identify banner candidates that best match the first and second scroll sentences. In one embodiment, candidate banners are automatically generated. In one embodiment, a ranking model is applied in order to rank banner candidates derived from the banner search and generation processes. One or more banners are then selected from the ranked banner candidates.
1. A computer-implemented method for generating an output banner associated with a Chinese language input couplet comprising first and second sentences, the method comprising:
identifying a set of related words, each related word being identified based on a number of times that the related word and a word in the input couplet occur at a same position in a couplet corpus, each of the words in the set of related words and the input couplet being in a Chinese language;
calculating an association strength between each of the related words in the set of related words and the input couplet;
utilizing a computer processor that is a component of a computer to combine related words in the set of related words in a plurality of different combinations to create a set of banner candidates, the related words in the set of related words being combined based on their association strengths;
selecting the output banner from the set of banner candidates; and
providing an output indicative of the output banner.
2. The method of claim 1 , wherein identifying comprises identifying based on analysis conducted relative to a translation model.
3. The method of claim 1 , wherein combining related words in the set of related words comprises arbitrarily combining related words in the set of related words so as to form a plurality of 4-character banner candidates.
4. The method of claim 1 , wherein selecting the output banner comprises ranking the banner candidates contained within the set of banner candidates.
5. The method of claim 1 , wherein selecting the output banner comprises ranking the banner candidates contained within the set of banner candidates relative to banner candidates not contained within the set of banner candidates.
6. The method of claim 2 , wherein the translation model includes a statistical framework and is trained on the couplet corpus.
7. The method of claim 1 , wherein a related word is selected into the set of related words based on a probability of the related word given a word in the input couplet.
8. The method of claim 1 , wherein the input couplet comprises first and second scroll sentences.
9. The method of claim 8 , wherein the first and second sentences have an equal number of words.
10. The method of claim 1 , and further comprising a set of existing banners that are related to the input couplet.
11. A computer-implemented method comprising:
selecting, from a collection of existing banners, one or more existing banners that are related to a Chinese language input couplet, the Chinese language input couplet comprising first and second sentences;
using one or more computer processors to identify a set of related words based on words in the input couplet, each of the words in the set of related words and the input couplet being in a Chinese language, and to generate one or more banners by combining words in the set of related words into a plurality of different combinations;
ranking the one or more generated banners relative to the one or more existing banners; and
selecting an output banner based on the ranking.
12. The method of claim 11 , wherein selecting comprises:
selecting the one or more existing banners from a database of existing banners, based on features of the existing banners and based on features of the input couplet.
13. The method of claim 12 , wherein providing an output comprises displaying a selected one of the existing banners.
14. The method of claim 11 , wherein each of the existing banners and the set of generated banners has a meaning that is different than the meaning of either the first and second sentence but that is complimentary to the meanings of the first and second sentence.