Spaces:
Sleeping
Sleeping
File size: 6,788 Bytes
71dd4c8 aae5063 71dd4c8 aae5063 71dd4c8 aae5063 71dd4c8 aae5063 71dd4c8 aae5063 71dd4c8 aae5063 71dd4c8 d67cdc6 445a2ed 71dd4c8 a7359a8 71dd4c8 a1e2d65 71dd4c8 df96f45 71dd4c8 |
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 |
import streamlit as st
import pandas as pd
from eventbrite_scrapper import Eventbrite
from sentence_transformers import SentenceTransformer
from sklearn.metrics.pairwise import cosine_similarity
import numpy as np
from datetime import datetime
from dataclasses import dataclass, field, replace
from typing import List, Any
# Dataclasses for event structure
@dataclass(frozen=True)
class EventAddress:
latitude: float = None
longitude: float = None
region: str = None
postal_code: str = None
address_1: str = None
@dataclass(frozen=True)
class EventVenue:
id: str = None
name: str = None
url: str = None
address: EventAddress = field(default_factory=lambda: EventAddress())
@dataclass(frozen=True)
class EventImage:
url: str = None
@dataclass(frozen=True)
class EventTag:
text: str = None
@dataclass(frozen=True)
class Event:
id: str = None
name: str = None
url: str = None
is_online_event: bool = False
short_description: str = None
published_datetime: datetime = None
start_datetime: datetime = None
end_datetime: datetime = None
timezone: str = None
hide_start_date: bool = False
hide_end_date: bool = False
parent_event_url: str = None
series_id: str = None
primary_venue: EventVenue = field(default_factory=lambda: EventVenue())
tickets_url: str = None
checkout_flow: str = None
language: str = None
image: EventImage = field(default_factory=lambda: EventImage())
tags_categories: tuple = field(default_factory=tuple)
tags_formats: tuple = field(default_factory=tuple)
tags_by_organizer: tuple = field(default_factory=tuple)
def __hash__(self):
return hash(self.id) if self.id else hash((self.name, self.is_online_event, self.start_datetime, self.primary_venue.name))
# Event Retrieval Pipeline
class EventbriteRAGPipeline:
def __init__(self, events: List[Event], embedding_model: str = 'all-MiniLM-L6-v2'):
self.events = [
replace(
event,
tags_categories=tuple(event.tags_categories),
tags_formats=tuple(event.tags_formats),
tags_by_organizer=tuple(event.tags_by_organizer),
)
for event in events
]
self.model = SentenceTransformer(embedding_model)
self.event_embeddings = self._compute_embeddings()
def _compute_embeddings(self) -> List[np.ndarray]:
def event_to_text(event: Event) -> str:
text_parts = [
event.name or '',
event.short_description or '',
' '.join(tag.text for tag in event.tags_categories),
' '.join(tag.text for tag in event.tags_formats),
' '.join(tag.text for tag in event.tags_by_organizer),
event.primary_venue.name or '',
event.primary_venue.address.region or '',
event.language or ''
]
return ' '.join(filter(bool, text_parts))
return self.model.encode([event_to_text(event) for event in self.events])
def query_events(self, query: str, top_k: int = 5) -> List[Event]:
# query_embedding = self.model.encode(query).reshape(1, -1)
# similarities = cosine_similarity(query_embedding, self.event_embeddings)[0]
# top_indices = similarities.argsort()[-top_k:][::-1]
# return [self.events[idx] for idx in top_indices]
query_embedding = self.model.encode(query).reshape(1, -1)
similarities = cosine_similarity(query_embedding, self.event_embeddings)[0]
top_indices = similarities.argsort()[-(top_k * 2):][::-1] # Get extra events to filter duplicates
unique_events = {}
for idx in top_indices:
event = self.events[idx]
if event.id not in unique_events:
unique_events[event.id] = event
if len(unique_events) == top_k:
break
return list(unique_events.values())
# Event Evaluator
class EventEvaluator:
def __init__(self, pipeline):
self.pipeline = pipeline
def evaluate_query(self, query):
"""Evaluate a single query and return results."""
# top_events = self.pipeline.query_events(query)
# results = []
# for event in top_events:
# result = {
# "Event Name": event.name,
# "Online Event": event.is_online_event,
# "Start Time": event.start_datetime,
# "Venue Address": event.primary_venue.address.address_1,
# "Venue Name": event.primary_venue.name,
# "Description": event.short_description,
# "Tickets URL": event.tickets_url,
# "Language": event.language,
# "Categories": [tag.text for tag in event.tags_categories],
# }
# results.append(result)
top_events = self.pipeline.query_events(query)
results = []
seen = set()
for event in top_events:
if event.id not in seen: # Ensure unique events
seen.add(event.id)
results.append({
"Event Name": event.name,
"Online Event": event.is_online_event,
"Start Time": event.start_datetime,
"Venue Address": event.primary_venue.address.address_1,
"Venue Name": event.primary_venue.name,
"Description": event.short_description,
"Tickets URL": event.tickets_url,
"Language": event.language,
"Categories": [tag.text for tag in event.tags_categories],
})
return results
# Fetch events from Eventbrite API
client = Eventbrite()
events = client.search_events.get_results(
region="ca--los-angeles",
dt_start="2025-02-26",
dt_end="2025-02-28",
max_pages=6,
)
# Initialize pipeline and evaluator
rag_pipeline = EventbriteRAGPipeline(events)
evaluator = EventEvaluator(rag_pipeline)
# Streamlit UI
st.title("🎟️ Event Search App")
st.write("Find events based on your interests!")
query = st.text_input("🔎 Enter your search query:")
# if query:
# results = evaluator.evaluate_query(query)
# if results:
# df = pd.DataFrame(results)
# st.dataframe(df) # Display results as a formatted table
# else:
# st.warning("No results found.")
if query:
print(f"🔍 Processing query: {query} ") # Debugging query input
results = evaluator.evaluate_query(query)
if results:
df = pd.DataFrame(results)
st.dataframe(df) # Display results as a formatted table
else:
st.warning("No results found.")
|