Method and device for establishing chain brand word bank and category word bank
A brand and chain technology, applied in the field of geographic information, can solve problems such as low work efficiency and inability to update the lexicon in time, and achieve the effect of improving efficiency, improving efficiency and speed
- Summary
- Abstract
- Description
- Claims
- Application Information
AI Technical Summary
Problems solved by technology
Method used
Image
Examples
Example Embodiment
[0045] Example one
[0046] In this embodiment, a chain brand word recognizer is trained based on the POI data in the POI database. The chain brand word recognizer can divide the name stems derived from the POI data into chain brand words and non-chain brand words, and filter the classification results Out is the backbone of the name of the chain brand word, and it is stored in the chain brand word database. See figure 1 , Which is a flowchart of a method for establishing a chain brand word database of the present invention, the method includes the following steps:
[0047] Step 101: Aggregate POI data with the same name backbone in the same city POI database into a POI data group, wherein the POI data group corresponds to the name backbone;
[0048] "Name backbone" refers to the part of the POI data name after removing the branch and address and other subsidiary information. The way of distinguishing the name backbone and subsidiary information is related to the POI data format. I...
Example Embodiment
[0098] Example two
[0099] The difference between the second embodiment and the first embodiment is that after the chain brand word recognizer is obtained, the recognition accuracy of the chain brand word recognizer is further checked. If the recognition accuracy of the chain brand word recognizer is checked, the recognition accuracy of the chain brand word The recognizer is adjusted, and then another test is performed, and the check and adjustment are repeated continuously until the recognition accuracy of the chain brand word recognizer meets the requirements. See figure 2 , Which is a flowchart of another method for establishing a chain brand word database of the present invention, the method includes the following steps:
[0100] Step 201: Aggregate POI data with the same name backbone in the POI database of points of interest in the same city into a POI data group, where the POI data group corresponds to the name backbone;
[0101] Step 202: Extract the identification feature...
Example Embodiment
[0125] Example three
[0126] The first two embodiments are mining chain brand words from the POI database. Since the data in the POI database uses terms, the chain brand words mined are basically standardized names, which may be related to the user's usage habits. Does not match. For example, the standardized name of a chain pharmacy is "**大药房", and the user may be accustomed to call it "**pharmacy". If the user enters the query term "**pharmacy", it will be concluded that the query term is not a chain brand The wrong result of the word. In addition, since the name of POI data is rarely a category word, it is also difficult to mine category words from the POI database.
[0127] In this embodiment, a recognizer that can recognize chain brand words, category words, and common words is trained based on the query words recorded in the user query log and the clicked POI data corresponding to the query words, and the recognizer is used to check the data recorded in the user query lo...
PUM
Abstract
Description
Claims
Application Information
- R&D Engineer
- R&D Manager
- IP Professional
- Industry Leading Data Capabilities
- Powerful AI technology
- Patent DNA Extraction
Browse by: Latest US Patents, China's latest patents, Technical Efficacy Thesaurus, Application Domain, Technology Topic.
© 2024 PatSnap. All rights reserved.Legal|Privacy policy|Modern Slavery Act Transparency Statement|Sitemap