PsyProxy
Datasets·Customer reviews·12__consumer_reviews__amazon_video_games__star_rating

Amazon video-game reviews: star rating (ordinal)

Amazon Video-Games Reviews is a public Kaggle corpus of consumer reviews scraped from Amazon’s video-games category, each with a star rating, verified-purchase flag, and free-text body. The texts in the sampled corpus predominantly discuss a variety of experiences related to gameplay mechanics , installation issues , and product quality . Many reviews express disappointment with specific games due to perceived shortcomings, such as poor graphics , limited multiplayer options , and authentication problems . Positive sentiments often highlight enjoyable gameplay and immersive storylines , while some reviews focus on the functionality and aesthetics of gaming accessories like controllers and headsets. Additionally, there are mentions of customer service experiences and the impact of game content on different age groups, reflecting a broad spectrum of user engagement with video games and related products. [Summary on 50 random texts by ChatGPT 4o Mini].

Distribution of 1–5 star rating
1
2
3
4
5
36,488 at floor173,880 at ceiling
299,990
items
53,609
holdout n
1–5 star rating
target
Ordinal
kind
25
systems compared
Criterion validity

Reported holdout systems from the verified card

Ordinal prediction uses Quad. κ as the task-primary metric. Secondary columns keep the companion metrics visible so binary, ordinal, regression, and multiclass cards are not compared through one flattened score.

Source podium · Quad. κ · 10 families
Gold
PsyProxy
0.767
Silver
OpenAI (Rathje)
0.548
Bronze
VADER
0.452
Model-family mix
PsyProxy · 4OpenAI / LLM · 3Lexicon · 2Topic model · 1Baseline · 15
SystemFamilyVariantQuad. κWithin-oneMAEPrimary scale
psyproxyPsyProxy — Social Economics Lens v0.5 · 1000d
PsyProxystrict0.7670.8840.50
psyproxyPsyProxy — Behavioral Sciences Lens v0.5 · 1000d
PsyProxystrict0.7480.8760.52
psyproxyPsyProxy — Technology Lens v0.5 · 800d
PsyProxystrict0.7460.8750.52
psyproxyPsyProxy — Health Lens v0.9 · 1100d
PsyProxystrict0.7430.8730.53
llmOpenAI Model gpt-4.1-nano
OpenAI / LLMstrict0.5480.8090.74
llmOpenAI Model gpt-5-nano
OpenAI / LLMstrict0.4880.7970.78
llmOpenAI Model gpt-4o-mini
OpenAI / LLMstrict0.4520.7830.83
lexValence Aware Dictionary and sEntiment Reasoner (VADER)
Lexiconstrict0.4520.7890.80
lexLinguistic Inquiry and Word Count (LIWC)
Lexiconstrict0.4230.7790.83
topicHierarchical Dirichlet Process (tomotopy HDP)
Topic modelstrict0.2070.7500.94
baselineTextDescriptives
Baselinestrict0.1310.7500.94
baselineEmpath
Baselinestrict0.0730.7430.97
baselineTool for the Automatic Analysis of Syntactic Sophistication and Complexity (TAASSC)
Baselinestrict0.0470.7430.97
baselineTool for the Automatic Analysis of Lexical Sophistication (TAALES)
Baselinestrict0.0380.7390.98
baselineTool for the Automatic Analysis of Cohesion (TAACO)
Baselinestrict0.0020.7390.98
baselineAmazon Video-Games reviews continuous · via best lens
Baselinestrict
baselineDisneyland TripAdvisor reviews continuous · via best lens
Baselinestrict
baselineIMDB movie reviews (ACL) binary · via best lens
Baselinestrict
baselineDouban movie reviews (Chinese) ordinal · via best lens
Baselinestrict
baselineSentiment140 tweets binary · via best lens
Baselinestrict
baselineDruglib drug reviews regression · via best lens
Baselinestrict
baselineDruglib drug reviews ordinal · via best lens
Baselinestrict
baselineDisneyland TripAdvisor reviews binary · via best lens
Baselinestrict
baselineLIAR fact-check statements ordinal · via best lens
Baselinestrict
baselineAmazon Video-Games reviews binary · via best lens
Baselinestrict