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 obtaining an output banner associated with a Chinese language input couplet, the method comprising:
receiving the Chinese language input couplet, the Chinese language input couplet including a first sentence and a second sentence, each of the first and the second sentences having a same number of words, the words in the first sentence being different than the words in the second sentence, each of the first and the second sentences having a meaning that is different than the other sentence;
utilizing a computer processor that is a component of a computer to identify a set of existing banners that are related to the input couplet; each of the banners in the set of existing banners having a number of words that is different than the number of words in the Chinese language input couplet sentences. each of the banners in the set of existing banners having a meaning that is different than the meaning of either the first sentence or the second sentence but that is complimentary to the meanings of the first and the second sentences;
generating a set of banners that are related to the input couplet, each of the banners in the set of generated banners having a number of words that is different than the number of words in the Chinese language input couplet sentences, each of the banners in the set of generated banners having a meaning that is different than the meaning of either the first sentence or the second sentence but that is complimentary to the meanings of the first and the second sentences;
generating a ranked banner set by applying a ranking model to rank the set of existing banners and the generated set of banners relative to one another based on relevancy to the input couple; and
selecting the output banner from either the set of existing banners or the generated set of banners wherein selecting the output banner comprises selecting the output banner from the ranked banner set.
2. The method of claim 1 , further comprising considering the contents of both the generated and existing sets in a determination of which banner should be the output banner.
3. The method of claim 1 , wherein identifying a set of existing banners comprises generating a feature vector for each banner in the set of existing banners.
4. The method of claim 3 , wherein identifying a set of existing banners further comprises constructing a banner taxonomy based on the set of existing banners and based on the feature vectors.
5. The method of claim 4 , wherein identifying a set of existing banners further comprises generating a feature vector for the input couplet.
6. The method of claim 5 , wherein identifying a set of existing banners further comprises comparing, through the banner taxonomy, the input couplet to the set of existing banners.
7. The method of claim 6 , wherein comparing the input couplet to the set of existing banners comprises calculating a distance between the input couplet feature vector, and each of the feature vectors for the banners in the set of existing banners.
8. The method of claim 7 , wherein identifying a set of existing banners further comprises identifying at least one banner in the set of existing banners as being a better match candidate banner than another of the banners in the set of existing banners.
9. The method of claim 8 , wherein constructing a banner taxonomy comprises classifying the set of existing banners based on category.
10. The method of claim 9 , wherein constructing a banner taxonomy further comprises dividing the banners in a category into subcategories.
11. The method of claim 10 , wherein constructing a banner taxonomy further comprises creating a centroid feature vector for each subcategory.
12. The method of claim 11 , wherein comparing the input couplet to the set of existing banners further comprises categorizing the input couplet.