Upload folder using huggingface_hub

custom_index/default__vector_store.json

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:4fb36718e1010f6a5649d954b95d8983b533b7f374ddbf98a06648777eaf707d
-size 28694879
+oid sha256:d17ed74c1649a438e518a8dc56a7772913dfe1ea7a7605bce069c63872431455
+size 72

custom_index/graph_store.json

@@ -1 +1 @@
-{"graph_dict": {}}
+{"graph_dict": {"Relative valuation": [["performed manually", "considers two factors: company and its affiliates"]], "CRV": [["is", "heterogeneous multi-modal graph neural network"], ["is", "used for corporate relative valuation"]], "Heterogeneous Multi-Modal Graph Network": [["for Corporate Relative Valuation", "Definition 2"]], "Data visualization": [["Fig. 3", "(a) is the number of each class in relative valuation level (CRV) and (b) is the number of each class in business category (BC)"]], "page\\_label": [["is", "3"], ["is", "100"]], "file\\_name": [["is", "Corporate Relative Valuation Using Heterogeneous MultiModal Graph Neural Network Yang et al.pdf"], ["is", "On the dynamics and adaptivity of mental processes"]], "file\\_path": [["is", "docs\\Corporate Relative Valuation Using Heterogeneous MultiModal Graph Neural Network Yang et al.pdf"], ["is", "docs\\On the dynamics and adaptivity of mental processes"]], "LSTM": [["is", "module"], ["based", "module"]], "Bi-LSTM": [["is", "employed"]], "1) (dimension": [["varies", "from 32 to 256"]], "2) (performance": [["degrades", "after 128 dimension"]], "EA model": [["consists of", "three types of elements"], ["contains", "architecture building blocks"], ["contains", "relations between architecture building blocks"], ["contains", "attributes of architecture building blocks"]], "transformation model": [["defined as", "{x|x\u2208 currentArchitecture \u00d7targetArchitecture }"]], "noSuccessor set": [["check", "current architecture"], ["check", "target architecture"]], "noPredecessor set": [["suggest", "target architecture"]], "Diefenthaler": [["is", "one of the authors of From Gaps to Transformation Paths in Enterprise Architecture Planning"]], "Bauer": [["is", "one of the authors of From Gaps to Transformation Paths in Enterprise Architecture Planning"]], "planningKnowledgeBase": [["is", "concrete actions"], ["is", "applicable actions"], ["is", "possible change"], ["is", "transformation path"], ["is", "sequence of concrete actions"], ["is", "state space"]], "MasterData": [["is successor of", "MasterData v1"], ["is successor of", "MasterData v2"]], "Semantic web technologies": [["formalize information sources", "advantages"]], "Quasar Enterprise": [["is", "app-roach"]], "Gap analysis": [["requires", "modeling current application landscape"]], "Tool": [["capable to generate", "list of actions to result in ideal landscape"]], "Heterogeneous Graph Transformer": [["architecture for modeling", "Web-scale heterogeneous graphs"], ["designed for homogeneous graphs", "infeasible to represent heterogeneous structures"], ["node- and edge-type dependent parameters", "characterize heterogeneous attention over each edge"], ["relative temporal encoding technique", "capture dynamic structural dependency with arbitrary durations"], ["heterogeneous mini-batch graph sampling algorithm", "efficient and scalable training"], ["architecture", "Figure 2"], ["parameterize weight matrices", "meta relations"], ["incorporate network dynamics", "relative temporal encoding mechanism"], ["is", "model for downstream tasks"], ["includes", "three components"], ["takes", "edges and their corresponding meta relations as input"], ["adopts", "Graph Attention Network as Attention mechanism"], ["model", "captures different semantic relations"], ["is", "baselines"], ["is", "designed for homogeneous graphs"], ["is", "designed for heterogeneous graphs"]], "WATT": [["matrix", "distinct edge-based"]], "\u03d5(e)": [["function", "edge type"]], "RTE": [["is inspired by", "Transformer's positional encoding method"], ["has been shown successful to capture", "sequential dependencies of words in long texts"], ["denoted the relative time gap as an index to get", "relative temporal encoding"]], "Graph Convolutional Networks": [["is", "GNN baseline for homogeneous graphs"]], "Graph Attention Networks": [["is", "GNN baseline for homogeneous graphs"]], "Relational Graph Convolutional Networks": [["is", "dedicated heterogeneous GNN baseline"]], "Heterogeneous Graph Neural Networks": [["is", "dedicated heterogeneous GNN baseline"]], "Heterogeneous Graph Attention Networks": [["is", "dedicated heterogeneous GNN baseline"]], "Heterogeneous weight parameterization": [["is", "major component of HGT"]], "Relative Temporal Encoding": [["is", "major component of HGT"]], "HGT-RT E": [["is", "ablation study for HGT without relative temporal encoding"]], "HGT+RT E": [["is", "ablation study for HGT with relative temporal encoding"]], "HGT-RT E-Heter": [["is", "ablation study for HGT without heterogeneous weight parameterization"]], "J. Treur": [["predefined", "basic combination functions"], ["is", "author"], ["wrote", "On the dynamics and adaptivity of mental processes"], ["wrote", "Relating adaptive dynamical systems and self-modeling network models by mathematical analysis"], ["is", "author of the book"]], "Continuous smooth dynamical system": [["can be described by", "temporal-causal network model"]], "Theorem 1": [["from dynamical system to network model", "any continuous smooth dynamical system model can be transformed into a temporal-causal network model"]], "X5": [["affects", "X1"], ["X1 + X1(X5 - \u03b1)", "X5"], ["X1 + X1(X5 - \u03b1)", "X5"], ["X1 + X1(X5 - \u03b1)", "X5"], ["X1 + X1(X5 - \u03b1)", "X5"], ["X1 + X1(X5 - \u03b1)", "X5"]], "X1": [["affects", "X3"], ["(X5 - \u03b1) + X1", "X1"], ["X1 + X1(X5 - \u03b1)", "X1"], ["X1 + X1(X5 - \u03b1)", "X1"], ["X1 + X1(X5 - \u03b1)", "X1"], ["X1 + X1(X5 - \u03b1)", "X1"]], "X2": [["affects", "X3"]], "X3": [["affects", "X4"]], "X4": [["affects", "X5"]], "combination functions": [["identified from", "differential equations (8)"]], "MIT Press": [["is", "publisher of the book"]], "pp. 1-43": [["is", "range of pages in the book"]]}}