The invention provides an information recommendation method and device. After information to be recommended is obtained, information characteristic vectors of the information to be recommended are extracted, meanwhile user characteristic vectors of a user to be recommended are obtained; then, a well-trained recommended algorithm model based on a factorization mechanism is utilized to conduct fitting on the information characteristic vectors of the information to be recommended and the user characteristic vectors, user information matched with the information to be recommended is obtained, namely a user group possibly interested in the information to be recommended is obtained. Due to the fact that the well-trained recommended algorithm model based on the factorization mechanism gives consideration to the association of the characteristics during calculation, recommendation results of the information recommendation method are more accurate.

Information recommendation method and device

The invention provides an advertisement click-through rate (CTR) prediction method based on a factorization machine (FM). The method includes: carrying out analysis and abnormal-value elimination on click record data; encoding feature data; training the FFM (Field-aware Factorization Machine); and predicting a click-through rate. According to the method, an FFM model is adopted to carry out combined classification on features, and thus accuracy of advertisement click prediction is improved.

Advertisement click-through rate (CTR) prediction method based on factorization machine (FM)

The invention provides a semantic computing method for improving a word vector model. The method comprises the following steps of S1, preprocessing a corpus; S2, tagging parts of speech and tagging the parts of speech of words obtained through preprocessing of the corpus; S3, initializing vectors and vectoring the words obtained through tagging of the parts of speech and the parts of speech; S4, integrating context vectors and computing and integrating context word vectors and part of speed vectors of the words; S5, establishing a Huffman tree, training a network, optimizing a target function and judging whether an error reaches a threshold value or not; S6, obtaining the vectors and obtaining the word vectors and the part of speed vectors; and S7, applying the vectors and carrying out semantic computing through application of the word vectors and the part of speed vectors. Compared with the prior art, the method has the advantages that a part of speech factor is added to the vectors, an existing Word2vec model is improved, moreover, innovative application is carried out according to the improved model, and a function of carrying out semantic computing through the Word2vec is expanded.

Semantic computing method for improving word vector model

The invention discloses an advertisement putting processing method and device based on advertisement matching. The method comprises: receiving an advertisement putting request; according to an advertisement putting plan carried in the advertisement putting request, obtaining, from an advertisement database, candidate advertisements satisfying the advertisement putting plan; according to user labelinformation carried in the advertisement putting request, comparing the user label information with the target user information corresponding to each candidate advertisement; and screening out, fromall the candidate advertisements, the target advertisements with the target user information matched with the user label information and carrying out putting processing. According to the technical scheme provided by the invention, by matching the user label information and the advertisement target user information, the advertisements are delivered to users meeting the advertisement putting requirements, so that the advertisements delivered can better satisfy the requirements of the users, the problem that advertisement resources are wasted due to the fact that advertisements are delivered in anon-targeted manner is avoided, and the advertisement putting effect is guaranteed.

Advertisement putting processing method and device based on advertisement matching

The invention provides a news sentence clustering method based on semantic similarity. The method includes the following steps: preprocessing news sentences of a corpus, and extracting available words; utilizing the available words to train a continuous bag-of-words model to obtain an initial word vector of each available word; utilizing an initial sentence vector of each news sentence and the initial word vectors of the left and right adjoining available words of a certain available word in the news sentence to train the continuous bag-of-words model in an iterative manner to obtain a currentword vector of each available word in the news sentence and a final sentence vector of the news sentence; merging an average value of the word vectors of all the available words, one-hot vectors of high-frequency words and the final sentence vector of each news sentence to obtain a semantic vector of the news sentence; and calculating distances between the semantic vectors to obtain the semanticsimilarity between the different news sentences, and clustering the news sentences of the corpus in accordance therewith. The invention also provides an electronic device and a computer-readable storage medium.

News sentence clustering method based on semantic similarity, device and storage medium

The invention provides an advertisement information pushing method. The method comprises the steps that a server side acquires the advertisement pushing request of a client; the server side sends an advertisement request to an advertisement pool module, and the advertisement pool module acquires advertisements from a database; the server side receives the advertisements returned by the advertisement pool module; the server side calls an algorithm policy module to estimate the click rate of the advertisement data, and carries out filtering and sorting; and the server side sends filtered and sorted advertisements back to the client. Compared with the prior art, the method provided by the invention carries out personalized advertisement recommendation for mobile device users in view of the characteristics of a mobile terminal. According to the invention, more suitable advertisement data can be more accurately selected for the needs of users, and are accordingly pushed to users. The invention further provides an advertisement information pushing system for implementing the method.

Advertisement information pushing method and system

The invention relates to a message filtering method. A Bayes classifier model is used for calculating the probability that a message is a rogue message; the property of the message is judged according to the probability of the rogue message, wherein the property of the message includes the normal message, the rogue message and the suspicious message; the rogue message is directly filtered away, and the normal message and the rogue message are remained; then, the further classification is performed by using a semanticbased deep study model; and the message is determined to be the normal message or the rogue message. Compared with the prior art, the method and the device have the advantages that the automatic study can be realized; a training set is regularly updated; the synonym is recognized; the manual annotation cost is greatly reduced; and high accuracy, robustness and stability are realized. In addition, the invention also provides a message filtering realization device.

Liu Di

Papers

Advertisement information pushing method and system

Message filtering method and device

FP-CNNH: 一种基于深度卷积神经网络的快速图像哈希算法 (FP-CNNH: A Fast Image Hashing Algorithm Based on Deep Convolutional Neural Network).