A method and device for constructing an urban knowledge map
A technology of knowledge graph and city, applied in the field of big data, can solve the problems that urban planners cannot provide intuitive reference, and it is inconvenient to expand the content of smart city system
- Summary
- Abstract
- Description
- Claims
- Application Information
AI Technical Summary
Problems solved by technology
Method used
Image
Examples
Embodiment 1
[0066] Embodiment 1 of the present invention discloses a method for constructing an urban knowledge map, the method flow chart is as follows figure 1 shown, including the following steps:
[0067] S101, performing word segmentation on the city data corpus to obtain at least one phrase;
[0068] In the process of executing step S101, optionally, firstly segment the city data corpus by sentence segmentation to obtain all the sentences contained in the city data corpus; then perform Chinese word segmentation on each sentence obtained by segmenting the sentence to obtain each phrase.
[0069] S102, extracting features and constructing corresponding feature vectors for each phrase;
[0070] In the process of executing step S102, for each phrase, the features contained therein are extracted; optionally, the features can be one or more of word features, part-of-speech features, context window features, dictionary features and statistical features, Specific features can be selected ...
Embodiment 2
[0081] In combination with the method for constructing a city knowledge map disclosed in Embodiment 1 of the present invention, as figure 1 In the shown step S103, the specific process of generating the city entity recognition model in advance, such as image 3 shown, including the following steps:
[0082] S301, performing word segmentation on the city data entity training corpus to obtain at least one entity training phrase;
[0083] S302, extracting features and constructing corresponding entity training feature vectors for each entity training phrase;
[0084] S303, using each entity training feature vector as the input data of the first initial deep belief network for identifying urban entities, and performing layer-by-layer unsupervised pre-training on the first initial deep belief network, the first initial deep belief network consists of at least one Restricted Boltzmann machine layers are stacked;
[0085] In the process of executing step S303, the first initial de...
Embodiment 3
[0104] Based on the urban knowledge map construction method disclosed in the first and second embodiments above, the third embodiment of the present invention provides a corresponding device for implementing the above urban knowledge map construction method, and its structural diagram is as follows Figure 5 As shown, the urban knowledge map construction device 100 includes: a word segmentation module 101, a feature vector construction module 102, an entity recognition module 103, an entity relationship identification module 104 and a city knowledge map construction module 105; the entity recognition module 103 includes a city entity recognition model generation unit 1031, the entity relationship identification module 104 includes a city entity relationship identification model generating unit 1041;
[0105] The word segmentation model 101 is used to perform word segmentation on the urban data corpus to obtain at least one phrase;
[0106] Feature vector construction module 10...
PUM
Login to View More Abstract
Description
Claims
Application Information
Login to View More