ᴍᴀᴛʜᴇᴍᴀᴛɪᴄᴀʟ ᴍᴏᴅᴇʟʟɪɴɢ ᴏꜰ ᴏɴʟɪɴᴇ ᴘʀᴏᴅᴜᴄᴛ ᴏꜰꜰᴇʀɪɴɢꜱ ᴠɪᴀ ᴠᴀᴅᴇʀ ʟᴇxɪᴄᴏɴ-ʙᴀꜱᴇᴅ ɴᴀᴛᴜʀᴀʟ ʟᴀɴɢᴜᴀɢᴇ ᴘʀᴏᴄᴇꜱꜱɪɴɢ ᴀɴᴅ ᴀʀɪᴍᴀ ᴛɪᴍᴇ-ꜱᴇʀɪᴇꜱ ᴀɴᴀʟʏꜱɪꜱ
𝗣𝗨𝗥𝗣𝗢𝗦𝗘: This modelling analysis pertains on the joint use of lexicon-based Natural Language Processing and Time-Series Analysis to elucidate the mathematical models that aim to analyze the three product offerings: baby pacifier, hair dryer, and microwave oven of Sunshine Company. The analysis tried to identify, describe, and support their market offerings in Amazon through mathematical evidences and meaningful quantitative and qualitative data patterns.
𝗠𝗘𝗧𝗛𝗢𝗗: The methodology follows the framework of Valence Aware Dictionary for sEntiment Reasoning (VADER) in processing the textual and rating-based parameters of the data sets. Likewise, Autoregressive Integrated Moving Average (ARIMA) model was utilized for time-based inferences of the product purchases. Analyses were performed via Python programming and resulting mathematical models were established.
𝗥𝗘𝗦𝗨𝗟𝗧𝗦: Combinations of review sentiment and star rating measures of the analysis basically indicated that hair dryer product has the highest potential of market success. The star ratings from the mathematical models generated could either increase or decrease, depending on the frequency of the text-based quality descriptors from the review body. The mathematical model generated from the microwave oven data reflected negative effect for the product’s success due to having significant neutral and negative sentiment parameters. Baby pacifier is evident on having a large scale of predicted increase in the product purchase when offered from the Amazon platform. The hair dryer and microwave oven have generally low predictive successes in compare to the baby pacifier, but evident trend increase are also manifested.
𝗖𝗢𝗡𝗖𝗟𝗨𝗦𝗜𝗢𝗡: For microwave oven, Sunshine company should pay attention to the performance and price of the product, and make timely improvements. For the hair dryer, the company should focus on product performance and appearance issues to find out the factors of product success. For pacifier, the evaluation is generally high, and the company should pay attention to the product life. The sales volume of all products evidently increases through time. Meanwhile, the satisfaction of buyers to the products also increases. Analyses suggest that if Sunshine Company would want to further understand the product information accurately, it should take the star rating of the products as the measurement standard.
𝗘𝗩𝗔𝗟𝗨𝗔𝗧𝗜𝗢𝗡: Using the VADER framework in natural language processing of the review sentiments provided an ample quantification of sentences based on the proportion of the sentiment polarity (negative, neutral, or positive). This method is seen to be consistent with the actual situation. Furthermore, the ARIMA forecast is acceptably accurate and the mathematical models generated are very parsimonious and adequate. The time series prediction method was effective to predict the effect of changes in the short term.
𝗜𝗠𝗣𝗟𝗜𝗖𝗔𝗧𝗜𝗢𝗡𝗦: The online product review analysis can help us to reduce the cost of competition in the industry and understand the emotional needs and consumer preferences, so as to carry out fast and accurate commercial marketing.