Optimal Solution Generation from Reviews and Micro-Reviews using Greedy Algorithm
Sarika S. Hulyalkar, Harmeet Kaur Khanuja · 2017
The survey matter that is available, and the truth of reviews being exceedingly different and needlessly made of more words, clients habitually encounter the issue of choosing the suitable reviews on devour. Micro-reviews are rising similarly as another kind of web survey substance in the online networking. Micro-reviews are presented by clients in the form check-in benefits. They are brief and apt (about 200 characters long) and are profoundly concentrated, as opposed to what the long and wordy reviews recommend. Here, in this paper, a new problem for review mining is proposed, that obtains together these two different sources of review content. Specifically, the scope of micro-reviews concerning a particular destination is utilized for choosing a set of reviews that disguise proficiently those remarkable viewpoints for a substance or item. The approach comprises of a two-stage procedure: matching audit penalties to micro-reviews, what's more like picking a small set of reviews which disguise as a number of micro-reviews concerning the likely illustration, with couple penalties. The goal is defined similar to a combinatorial streamlining problem, and hint at how to infer an ideal result utilizing basic straight customizing. We additionally recommend a proficient heuristic calculation that approximates the optimal results. Finally, contributing an approach for implementing a system which determines query facets by combining frequent lists from the top results.