ASSESSMENT OF THE REVIEURS QUALITY ON THE BASE OF QUALIMETRIC METHOD «THE VALUE OF OPINION»
Keywords:review, recommendation system, collaborate filtering, quality, qualimetrics, metrix
The paper considers relevance of the subject, dealing with the assessment of the reviews quality, based on the qualimetric method in 2022 within the frame of the market places development, where greater share of purchases takes place after reviews analysis. Quality assessment can be used for the improvement of the quality of recommendation systems.
For the investigation the largest reviews data base in Ukraine on the platform for the motor vehicles search in Ukraine – Automoto.ua was used.
Aim of the research is assessment and improvement the reviews quality, systematization of the quality parameters of reviews analysis for their application in the development of the recommendation systems on the base of the qualimetric method and facilities. Available scientific substantiation for the assessment of reviews quality on the base of the qualimetric approach of the reviews analysis, which is the base for the algorithm of motor vehicles recommendations on the base of users “opinion” is – “The Value of Opinion” and for the introduction of the qualimetric method «Cyclogram of reviews quality “The Value of Opinion”».
Scientific task includes: arrangement of the automated collection of the reviews about the motor vehicles in one data base (parsing), systematization of the reviews quality indices on the base of qualimetric method within the frame of parsing; development of new indices of reviews quality indices, based on qualimetric approach; assessment of reviews quality; conclusion, regarding the rationality of usage the index “The Value of Opinion” for the development of the recommendation systems of the collaborate filtering, which comprises the assessment of the users “opinion”.
The expedience of using qualimetric method for reviews quality assessment is characterized. The importance of the reviews of the motor vehicles is described. The usage of machine learning technology for reviews processing within the frame of the set task has been considered.
New qualimetric approach to the reviews quality measuring as a part of data, the recommendation algorithms of modern recommendation systems will be based on, has been considered in the given scientific work. The suggested system of reviews quality assessment in the format «Cyclogram of reviews quality “The Value of Opinion”» contains the set of eight quality indices of the reviews about the motor vehicles. The importance of the given study for business-tasks has been described. Conclusions and final opinions on the scientific work have been formulated.