Finding Local Experts for Dynamic Recommendations Using Lazy Random Walk

Puspitaningrum, Diyah; Fernando, Julio; Afriando, Edo; Utama, Ferzha Putra; Rahmadini, Rina; Pinata, Y.

doi:10.1109/citsm47753.2019.8965354

Cited by 2 publications

(6 citation statements)

References 5 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Paper Objective [12] Created an item-based hierarchical deep learning architecture based on a dynamic recommender system [13] Implemented an algorithm for dynamically predicting the next item in the recommendation list sequence [14] Introduced a neural recommendation technique to model the dynamic network interactions [15] Introduced a dynamic recommender model implementing the random lazy walk to predict the top-n locations to users [16] Predicted a 5-day recommendation for consuming fruits and vegetables [17] Analyzed the 5-day recommendation based on analyzing oxidative damages and anti-oxidant defense [19] Researched the availability trend of fruits and vegetables across 10 European nations [20] Generated shopping list predictions for every customer to use the Point-of-Sale (POS) data to help with the sales in any offline store [21] Studied the application of recommender systems in detail for offline POS systems for retail businesses…”

Section: Ta B L E 1 a Summarized Analysis Of The Related Workmentioning

confidence: 99%

See 1 more Smart Citation

A novel approach to alleviate data sparsity and generate dynamic fruit recommendations from point‐of‐sale data

Gupta

Katarya

2022

Concurrency and Computation

View full text Add to dashboard Cite

Summary Recommender systems have become a core part of the retail experience. Retailers often rely on recommender systems to help them drive more conversions through targeted communication and advertisements. However, recommender systems are not one size fits all. Specialized retailers require specialized recommender systems to consider various features, attributes, and dynamics about the product category. In this paper, we have proposed a novel fruit recommender system that generates dynamic recommendations while remediating the problem of data sparsity. We have developed a novel fruit recommender system that considers the temporal dynamics in the fruit market, like price fluctuations, fruit seasonality, and quality variations that occur throughout the year. To perform this task, we have used Recurrent Recommender Network (RRN), which uses the deep learning method Long Short‐Term Memory (LSTM) to implement the system model. To ensure that our work and results obtained are practical, we have worked in a real‐world setting, by tying up with a specialty fruit retailer based in New Delhi to get the real‐world Point‐of‐Sale (POS) data of consumers. The result of the study suggests our algorithm performs better than other benchmark algorithms along NDCG and RMSE metrics.

show abstract

Section: Ta B L E 1 a Summarized Analysis Of The Related Workmentioning

confidence: 99%

“…To carry out this task, we add the stationary profile vectors u i and f j with the time-varying profile vectors, that is, u it and f jt with respectively. In other words, we conceive that the rating is a function of both dynamic and static states, as shown in Equation (15).…”

Section: Rating Emissionsmentioning

confidence: 99%

A novel approach to alleviate data sparsity and generate dynamic fruit recommendations from point‐of‐sale data

Gupta

Katarya

2022

Concurrency and Computation

View full text Add to dashboard Cite

show abstract

“…The weighted HITS algorithm based on user friendships between a target user and other users was applied to generate recommendation lists. In 2019, a dynamic recommender system for suggesting shopping places on Foursquare was proposed [35]. This work searched for local experts in two aspects including experts in their location (called local authorities) and experts in query categories (called topic authorities).…”

Section: Location Recommendation Based On Local Experts or Other User...mentioning

confidence: 99%

“…The first user has higher entropy values than that of the latter, which may not be correct. In [2], [13], [35], they leverage social relationships to recommend locations. However, friends do not always share common preferences.…”

Section: Location Recommendation Based On Local Experts or Other User...mentioning

confidence: 99%

“…In [34], the work discovers interesting locations based on distances, which could take more computational time in the online phase to calculate distances between users and locations. In [35], positive reviews are used to filter interesting locations. However, fake reviews might exist, and might cause biased recommendation lists.…”

Section: Location Recommendation Based On Local Experts or Other User...mentioning

confidence: 99%

See 1 more Smart Citation

N-Most Interesting Location-based Recommender System

Darapisut

Amphawan

Rimcharoen

et al. 2022

ECTI-CIT Transactions

View full text Add to dashboard Cite

The popular and ubiquitous location-based social networks (LBSNs) ap-peal many users for sharing interesting locations with other users. As the collected data (such as users' profiles, location comment, and suggestion) become a lot larger in size, location-based recommender systems require more effective fillters to be able to suggest potentially preferable locations to users. Location recommendation is more difficult and challenging especially if users have few or no check-in histories as a new user. Therefore, instead of depending on users' check-in histories, previous works focused on creating recommended location lists by leveraging the information given by other users who check in locations in each area. However, previous studies took into account only the frequencies of users for creating recommendation lists and have su˙ered from the cold-start problem where new users have few or none histories. As a result, the recommenders hardly suggest any locations matching the users' preference. In this paper, we propose an enhanced location-recommendation approach called N-most interesting location-based recommender system (NILR) to recommend interesting lo-cations for new users. Our approach can be divided into three phases. First, the NILR discovers interesting locations by taking into account both the visiting frequencies and the preferences of users already in the system. Second, a ranking procedure is executed to create a final recommendation list based on two interestingness scores: one obtained from the HITS-based model (as adopted by [1] and [2]) and the other from our proposed method. Finally, we re-filter interesting locations based on the current location of the new user. Experimental results reveal the NIRL can reach better precision, recall, average ranking and NDGC than HITS by 6%, 6%, 30%and 8% for Tokyo and 24%, 30%, 43% and 15% for New York dataset, respectively.

show abstract

Finding Local Experts for Dynamic Recommendations Using Lazy Random Walk

Cited by 2 publications

References 5 publications

A novel approach to alleviate data sparsity and generate dynamic fruit recommendations from point‐of‐sale data

A novel approach to alleviate data sparsity and generate dynamic fruit recommendations from point‐of‐sale data

N-Most Interesting Location-based Recommender System

Contact Info

Product

Resources

About