In the new technical background, more and more goods appear in front of the user. Unfortunately, users are increasingly easy to get lost in the massive commodity information. In order to improve the users experience o...
详细信息
ISBN:
(纸本)9781450366007
In the new technical background, more and more goods appear in front of the user. Unfortunately, users are increasingly easy to get lost in the massive commodity information. In order to improve the users experience of mobile e-commerce, combined with the advanced technology of large data age, there have been many recommendation algorithms for mobile e-commerce. All of this recommendation algorithms based on customer interests, sales and other different considerations, in order to achieve a purposeful and efficient screening recommendation information. In many ways of thinking, the user reputation as a new era of the spirit of the contract reflects the important value. In this paper, we focus on the current technical situation, designing a set of scoring algorithm through the analysis of the previous behavioral data modeling. After this algorithm, from different customer categories of different needs to start using the mainstream for the score of the business recommendation algorithmalgorithm to achieve. And the results are compared with the expected analysis in order to obtain the current algorithm in the credibility of the algorithm under the improvement and adjustment. Finally, we will design a set of recommendation algorithm based on the reputation standards.
Most traditional recommendation algorithms rely on common scoring items between users when making product recommenda-tions. The data is highly sparse and the recommendation effect is not good. Therefore, an improved c...
详细信息
ISBN:
(纸本)9781728181172
Most traditional recommendation algorithms rely on common scoring items between users when making product recommenda-tions. The data is highly sparse and the recommendation effect is not good. Therefore, an improved collaborative filtering recom-mendation algorithm is proposed. Based on user purchase rec-ords, the algorithm uses a representation learning method to construct a user product network, obtains the low-dimensional em-bedded semantic relationship between users and product no-des, and uses cosine similarity to measure the semantic similarity between products. Then, according to the hidden Dirichlet topic distribution model, the topic features of the products are obtained, and the cosine similarity is used to calculate the similarity of the topic features between the products. The linear fusion method is adopted to effectively alleviate the data sparse problem and improve the recommendation performance. Through Amazon product reviews Data set to verify the effectiveness of the recommendation algorithm.
Taking Weibo content as the research object and aiming at the disadvantages of only popular recommendation for Weibo, we propose personalized recommendation scheme. The improved user-based collaborative filtering algo...
详细信息
ISBN:
(纸本)9789881563903
Taking Weibo content as the research object and aiming at the disadvantages of only popular recommendation for Weibo, we propose personalized recommendation scheme. The improved user-based collaborative filtering algorithm is used to give a recommendation list. In this paper, the keywords extraction technology is used to extract the keywords of Weibo content as the project, and the results of the sentiment analysis algorithm are used as the scores to form the project score matrix. Then, the similarity of the item and the score are calculated by the cosine similarity and the Jaccard coefficient, respectively, the final similarity result is obtained by weighted fusion. The experimental results show that the improved algorithm proposed in this paper can effectively solve the problem that the microblog content has no scoring system, which helps to improve the performance of the recommendation system and obtain better recommendation results.
With the advent of the data age, the public has been facing the problem of information overload. recommendation algorithms are an effective way to solve this problem. At present, a large number of recommended algorith...
详细信息
ISBN:
(纸本)9781665440738
With the advent of the data age, the public has been facing the problem of information overload. recommendation algorithms are an effective way to solve this problem. At present, a large number of recommended algorithms adopt the following two ideas: content-based text similarity algorithm and user-based collaborative filtering algorithm. Researchers have developed a distributed collaborative recommendation protocol based on blockchain. However, these algorithms ignore the characteristics of the news industry itself. Just adopting the above ideas will inevitably lead to many internet public opinion problems. Therefore, this paper proposes an improved N-TF-IDF algorithm, which is more suitable for the news industry, and can control the outbreak of negative public opinion, and has a positive effect on stabilizing internet public opinion. Through the verification of the experimental data set, the algorithm is superior to the traditional information retrieval and text mining technology TF-IDF in both the time dimension and the emotional dimension, and this algorithm is not affected by citizens' privacy rights.
User listening to songs is sequential, while traditional music recommendation algorithms do not con-sider the relationship between listening behaviors and their context, so a music recommendation algo-rithm based on c...
详细信息
User listening to songs is sequential, while traditional music recommendation algorithms do not con-sider the relationship between listening behaviors and their context, so a music recommendation algo-rithm based on contextual semantics is proposed. Different from the traditional collaborative filtering recommendation algorithm based on user-scoring matrix, this algorithm abstracts the listening records of different users into time-series text sequences, and uses natural language models to capture songs in the process of song conversion. Context semantic relations. For quantized songs, it can be mapped to a high-dimensional space, and the similarity relationship can be obtained by calculating the distance between the songs. For the user, through the user's music listening records, the user can be portrayed to the same dimensional space as the song. Similarly, by calculating the distance between the user and the song, the similarity between the user and the song can be obtained, so as to achieve the purpose of recommendation. Compared with traditional recommendation algorithms, this algorithm can not only capture contextual semantic relationships that traditional recommendation algorithms cannot capture, but also has better overall top-n recommendation performance than traditional algorithms.
The traditional collaborative filtering recommendation algorithm has the problem of data sparsity and expansibility. Aiming at this problem, and improved bisecting k-means collaborative filtering algorithm *** algorit...
详细信息
ISBN:
(数字)9781510639690
ISBN:
(纸本)9781510639690
The traditional collaborative filtering recommendation algorithm has the problem of data sparsity and expansibility. Aiming at this problem, and improved bisecting k-means collaborative filtering algorithm *** algorithm first removes unrated items in the rating data matrix based on the Weighted Slope One algorithm preprocessing to reduce its sparsity. Then the preprocessed rating data is clustered based on the bisecting K-means algorithm, which reduces the nearest neighbor search space of the target user by assembling similar objects, thereby improving the algorithm's expansibility. Finally, use the recommendation algorithm to generate the final *** results show that the improved bisecting k-means algorithm improves the recommendation effect.
Traditional information recommendation system using only the user's score is calculated and recommended, although to a certain extent, can obtain the implied characteristics of users or resources, but the lack of ...
详细信息
Traditional information recommendation system using only the user's score is calculated and recommended, although to a certain extent, can obtain the implied characteristics of users or resources, but the lack of enough semantic interpretation, affecting the effects of recommendation. This article studied and analyzed the recommendation based on attribute coupled matrix decomposition algorithm in the application of Internet of things, on the foundation of the matrix decomposition model successively introduced global offset and time offset, in order to improve the prediction accuracy and the quality is recommended. In this paper, the algorithm is proved by experiment and the prediction accuracy of the algorithm is improved.
Since the beginning of the 20 th century, with the continuous development of computer technology, more and more people began to use the convenience of computer to improve efficiency, but just because a large number of...
详细信息
Since the beginning of the 20 th century, with the continuous development of computer technology, more and more people began to use the convenience of computer to improve efficiency, but just because a large number of Internet users continue to increase, there is also a situation of information overload, which will lead to some problems, such as the huge amount of data, the extraction and utilization of effective information will increase Difficulties. The second is the previous recommendation algorithm, most of which predict through the score, but if only through the score, it will not make full use of other data information in the data *** order to make the experimental results more convincing, this experiment uses a recommendation algorithm based on two-way attention model. First, the movie attributes and user attributes are processed by Convolutional Neural Network(CNN), and then through the full connection layer and the built attention model, effective information is obtained. Finally, the predicted score is generated and compared with the real score, and the final result is obtained *** experiment uses the movielens data set, by changing the parameters to affect the experimental results, so as to determine the final value of each parameter. This experiment is a comparative experiment, the recommendation algorithm with two-way attention model is compared with many previous algorithms, and the final conclusion is drawn. The experimental results are expressed by RMSE and MAE. According to the index results, the recommendation algorithm proposed in this experiment has better performance.
Due to the increasing popularity of multimedia social networks (MSNs), the ability to mine users' interests in different contexts on such networks is crucial in recommendation systems. It is, however, challenging ...
详细信息
Due to the increasing popularity of multimedia social networks (MSNs), the ability to mine users' interests in different contexts on such networks is crucial in recommendation systems. It is, however, challenging to mine users' current preferences based on session on MSNs. In this study, we propose a novel recommendation algorithm based on both social situation analytics and collaborative filtering for session-based recommendation. Specifically, the algorithm predicts the rating for target users based on their nearest neighbors and historical behaviors. First, for the purpose of mining users' current intentions, the session-level behavioral sequences of target user are analyzed based on SocialSitu (t). Then, the recommended contents are generated for target users based on their behaviors and perceived session-based intentions and identity. We evaluate the performance of the proposed algorithm using real-world social media dataset from Shareteches. Findings demonstrate that our algorithm outperforms two classical algorithms and a state-of-the-art method.
This paper describes a new collaborative filtering recommendation algorithm based on probability matrix factorization. The proposed algorithm decomposes the rating matrix into two nonnegative matrixes using a predicti...
详细信息
This paper describes a new collaborative filtering recommendation algorithm based on probability matrix factorization. The proposed algorithm decomposes the rating matrix into two nonnegative matrixes using a predictive rating model. After normalization processing, these two nonnegative matrixes provide useful probability semantics. The posterior distribution of the real part of the probability model is calculated by the variational inference method. Finally, the preferences for items that users have not rated can be predicted. The user-item rating matrix is supplemented by a preference prediction value, resulting in a dense rating matrix. Finally, time weighting is integrated into the rating matrix to construct the 3D user-item-time model, which gives the recommendation results. According to experiments using open Netflix, MovieLens, and Epinion datasets, the proposed algorithm is superior to several existing recommendation algorithms in terms of rating predictions and recommendation effects.
暂无评论