def compareAndMerge(grp_id):
pre_work()
grp2_id = None
grp2_keywords = None
max_match = 0
#print("compareAndMerge: ", grp_id)
group1_keywords = read_list(
f"data/group_profiles/group_keywords/{grp_id}_group_keywords.npy")
for g_id in GRP_IDS:
if g_id != grp_id:
group2_keywords = read_list(
f"data/group_profiles/group_keywords/{g_id}_group_keywords.npy"
)
similarity = findSimilarity(group1_keywords, group2_keywords)
if similarity >= THRESHOLD:
print("CompareAndMerge: similarity>=THRESHOLD: ", similarity,
"Group1 id: ", grp_id, "Group2 id: ", g_id)
if similarity > max_match:
max_match = similarity
grp2_id = g_id
grp2_keywords = group2_keywords
else:
print("CompareAndMerge: similarity<THRESHOLD: ", similarity,
"Group1 id: ", grp_id, "Group2 id: ", g_id)
if grp2_id is not None:
group1_data = read_dictionary(
f"data/group_profiles/group_information/{grp_id}_group_data.npy")
group2_data = read_dictionary(
f"data/group_profiles/group_information/{grp2_id}_group_data.npy")
mergeTwoGroups(group1_data, group1_keywords, group2_data,
grp2_keywords)
def groupPersonalisation(results, user_id):
if os.path.exists("data/user_to_group.npy"):
user_to_grp = read_dictionary("data/user_to_group.npy")
grp_id = user_to_grp[user_id]
if os.path.exists(f"data/group_profiles/group_keywords/{grp_id}_group_keywords.npy"):
grp_keywords = read_list(f"data/group_profiles/group_keywords/{grp_id}_group_keywords.npy")
grp_results = []
for res in results:
page_keywords = get_page_keywords(res[0])
similarity = findSimilarity(page_keywords, grp_keywords)
grp_results.append((res[0], similarity))
grp_results = Sort(grp_results)
return grp_results
def pre_work(user_id):
global DOCID_TO_CLICKS, URL_TO_DOCIDS, USER_KEYWORDS
if os.path.exists(
f"data/user_profiles/docid_to_clicks/{user_id}_docid_to_clicks.npy"
):
DOCID_TO_CLICKS = read_dictionary(
f"data/user_profiles/docid_to_clicks/{user_id}_docid_to_clicks.npy"
)
if os.path.exists("data/already_visited_urls.npy"):
URL_TO_DOCIDS = read_dictionary("data/already_visited_urls.npy")
if os.path.exists(
f"data/user_profiles/user_keywords/{user_id}_user_keywords.npy"):
USER_KEYWORDS = read_list(
f"data/user_profiles/user_keywords/{user_id}_user_keywords.npy")
def pre_work():
global GRP_IDS
if os.path.exists("data/GRP_IDS.npy"):
GRP_IDS = read_list("data/GRP_IDS.npy")
def get_grp_keywords(grp_id):
if os.path.exists(f"data/group_profiles/group_keywords/{grp_id}_group_keywords.npy"):
return read_list(f"data/group_profiles/group_keywords/{grp_id}_group_keywords.npy")
def get_user_keywords(user_id):
if os.path.exists(f"data/user_profiles/user_keywords/{user_id}_user_keywords.npy"):
return read_list(f"data/user_profiles/user_keywords/{user_id}_user_keywords.npy")
def get_doc_keywords(doc_id):
keywords = []
if os.path.exists(f"data/doc_pages_keywords/{doc_id}.npy"):
keywords = read_list(f"data/doc_pages_keywords/{doc_id}.npy")
return keywords
def get_page_keywords(docID):
doc_keywords = []
if os.path.exists(f"data/doc_pages_keywords/{docID}.npy"):
doc_keywords = read_list(f"data/doc_pages_keywords/{docID}.npy")
return doc_keywords
