Recognition method and device for core product word in title

Abstract

The invention discloses a method and a device for identifying core product words in a title, and relates to the technical field of computers. In the present invention, when the word2vec model generates the word vector of a word, it will refer to the adjacent or similar words of the word in the title. Therefore, the word vectors generated by the same word in different contexts are different. At the same time, the product word The n-gram feature of the product word contains the adjacent or similar words of the product word in the title, which can further indicate the context of the product word. Therefore, the n-gram feature of the same word in different contexts Furthermore, the expression of n-gram features in word vectors is also different, and the results of recognition through the core product word recognition model are also different, which improves the ambiguity problem of core product word recognition in the title and improves the accuracy rate.

Description

technical field [0001] The invention relates to the field of computer technology, in particular to a method and device for identifying core product words in a title. Background technique [0002] Sentence component analysis refers to the use of various methods to mark the basic components of sentences. Sentence component analysis is one of the basic problems in natural language processing and has a wide range of applications. E-commerce title component analysis is a branch of sentence component analysis, which is widely used in intent recognition, personalized ranking and other fields. But because the title is a pile of words (no subject, predicate verb, etc.), it is more complicated. [0003] The core product word in the title refers to the specific product involved in the title. For example, the core product word in the Korean version of casual pants men’s clothing is trousers. Identifying the core product word in the title is a major method for sentence composition anal...

AI Technical Summary

Problems solved by technology

[0004] However, there is a problem of inaccurate recognition in the method of using the vocabulary to analyze the core product words in the title. For example, the word "Xiaomi" has different meanings in different contexts: one is the brand word "Xiaomi mobile phone", and the other is Product word "millet porridge"

It is impossible to distinguish these two meanings simply by using a vocabulary, so the recognition of core product words is inaccurate

Images